Analytic Football

Beyond 'Moneyball':

2011-09-08T00:04:00.000-07:00

The first a a three part series I am writing was published in Analytics Magazine.

Beyond ‘Moneyball’:
The rapidly evolving world of sports analytics, Part I

By Benjamin Alamar and Vijay Mehrotra

Over the past few years, the world of sports has experienced an explosion in the use of analytics. In this three-part series, we reflect on the current state of sports analytics and consider what the future of sports analytics may look like.
We define sports analytics as “the management of structured historical data, the application of predictive analytic models that utilize that data, and the use of information systems to inform decision makers and enable them to help their organizations in gaining a competitive advantage on the field of play.”

(http://www.analytics-magazine.org/special-articles/391-beyond-moneyball-the-rapidly-evolving-world-of-sports-analytics-part-i.html)

Can Alex Smith Save the 49ers?

2011-07-08T15:01:00.000-07:00

I recently broke the news to my son that Alex Smith appears to be slotted in as the starting QB for his beloved 49ers. I took the dejected look on his face as a challenge to provide him some hope. Going to the data on QB's is always a iffy proposition as QB stats are the product of the efforts of an entire team, not just one player and given the woeful play of the 49ers over the last few years, I was expecting that this would be a rather fruitless effort in restoring hope.

I pulled up Smith's page on Pro-Football Reference and checked on his performance. I then took his age, years in the league, Yds/Att and interception rate for last season and plugged them into PFR's useful Player Season Finder. I frankly did not believe what I found. Besides Smith, the Player Season Finder found three other players who had a similar performance in Yds/Att and interception rate during their 5th season, at age 26: Drew Brees, Peyton Manning, and Ben Roethlisberger. Looking at a broader range of stats, Smith actually fits in with this group fairly well.

On Yds/Att, QB rating, TD% (indexed by PFR), Int% (indexed by PFR), and Completion Percentage, Smith fits right in. He is better than Big Ben in TD%, QB rating, and Compl%, better than Brees in Int%, and just behind the three on Yds/Att. The other three of course all of Super Bowl rings and multiple All-Star game appearances on their resumes. So what is going on? How can the much maligned Alex Smith, playing for multiple head coaches and offensive coordinators, have delivered a similar performance as these shining examples of QB play?

To try and answer that question, I expanded my search. I removed all the filters on performance, and just looked for QB's in their 5th season while they were 26 years old. Since 1980, there are 33 QB's not named Alex Smith who played their 5th season when they were 26. After eliminating all the QB's with fewer than 8 games played, I loaded their advanced passing stats into two different clustering analyses to find groupings of players that had similar performances across all of the stats. The results of the two analysis were the essentially the same and put Smith in with the following group:

Tommy Kramer
Steve McNair
Rich Gannon
Bernie Kosar
Neil Lomax
Gary Hogeboom
Trent Dilfer
David Whitehurst
Randall Cunningham
Paul McDonald
Kyle Boller
Tim Couch
Ben Roethlisberger

So this list seems rather more satisfying. Some players who have had success (Gannon, McNair, Boller, Roethlisberger) and some that haven't (Hogeboom, Couch). For the most part this list should provide 49er fans with some hope. There are no spectacular QB's here, but plenty that have played good efficient football and several that achieved more later in their careers (Gannon, Cunningham, McNair).

For some context, the rest of the QB's fall into the following groups:

Stars	Flashes	No Impact
Dan Marino	Vince Evans	Dave M. Brown
Peyton Manning	Mike McMahon	Mike Moroski
Chris Miller
Drew Brees
Dave Krieg
Brett Favre
Daunte Culpepper
Aaron Rodgers

Ok, maybe Chris Miller is not really a star, but for the most part, these groupings seem fairly reasonable, so 49er fans, have some hope, Alex Smith may just be Trent Dilfer.

Initial Poll Results

2011-07-02T09:05:00.000-07:00

As data starts to arrive from the entirely unscientific poll posted below, one result jumps out immediately. The very first question of the poll asks which area within analytics will have the biggest impact on teams. The choices offered were: Collecting new data, better data management, new metrics, integration of statistical analysis in decision making, and better information systems for decision makers. In writing that question I had imagined that the "new metrics" answer would be fairly popular. That is the area that draws most practitioners initially, as it is the use of statistical tools to measure and learn about a sport. The results did not at all match that expectation.

The chart above indicates that integration of analysis and better information systems are the areas that the respondents felt would have the biggest impact, while none of them selected new metrics. What this result suggests (and again, this was hardly a rigorous poll) is that the area of analytics that is perceived to be the biggest area of growth is finding ways to help decision makers use the analysis that has already been done. This can take the form of better reporting, more in depth conversation, better explanations, and/or better information systems (that put the decision makers in control of the analysis to some degree), to name a few.

Still time to take the poll if you have not already.

Future of Sports Analytics

2011-06-29T12:12:00.000-07:00

Five questions on the future of sports analytics. What are your thoughts?

Create your free online surveys with SurveyMonkey, the world's leading questionnaire tool.

Additcted to Math?

2011-06-28T12:21:00.000-07:00

My first reaction after reading Jonah Lehrer's piece on Grantland was to try and ignore it. It's the same debate we've been having since Lewis published Moneyball right? But there is something different in this latest railing against sports analytics (and please can we stop referring to statistics applied to any sport as Sabrmetrics? That is a baseball term, nobody in basketball or any other sport refers to their work that way.) The difference here is that instead of being accused of not being relevant, us geeks are being accused of being too relevant. Much as my grandfather accused me of ruining sports, Lehrer is accusing decision makers in sports of ruing the game because they only care about the numbers.

This is a rather bizarre charge, as any of who have worked for teams not named the Houston Rockets can tell you. Decision makers (GM's, coaches, etc) are not exactly waiting breathlessly for the latest pronouncements from their resident geeks. Do some teams factor good analysis into their decision making? Sure they do. By my count 11 of the 16 NBA playoff teams employed analysts, at least to some degree, but all of those teams also have serious scouting departments that employ a lot more people, and the scouts are not ignored. Personnel and coaching decisions are looked at from every angle imaginable, and it is rather silly to suggest that somehow math has put all of these other sources of information in the back seat.

I do want to address two specifics of the article though. The first is the use of the Mavericks as the counter example and the second is the phrase "The underlying assumption is that a team is just the sum of its players, and that the real world works a lot like a fantasy league." and both points actually tied back to the same idea.

First the Mavericks: The Mavericks are one of the most innovative teams when it comes to analytics, employing the first statistician who actually travels with the coaching staff and works with them on a daily basis. Roland Beech is one of the best statisticians in the world when it comes to basketball, and while I do not know what Roland's numbers looked like for Barea, I can bet that he had input on the decision. Data may not have made the decision, but, as it should be, was a factor. To continue on the Mavericks example, Lehrer notes that "According to one statistical analysis, the Los Angeles Lakers had four of the top five players in the series. The Miami Heat had three of the top four." (cleverly linking back to Beech's own site 82games.com to make his point). The problem is that not all analysis is created equal. What is published on the internet may not always be the cutting edge of basketball analysis (or football or baseball or soccer ....). Wayne Winston once famously said that, because of his analysis, we would advise a GM not to sign a young Kevin Durant at any price. Just because some one puts a number down does not mean it is a valuable number.

On the second point regarding the assumption that a team is somehow the sum of its parts, I would suggest that most GMs, Coaches, and even statisticians are sophisticated enough to know that this is simply not the case. I have never heard a serious statistician argue (outside of baseball at least) that you can simply add metrics together and get a result that predicts the outcome of adding a specific player.

What this comes down to, I believe, is a general misunderstanding about how sports analytics is both practiced and utilized. Sports analytics is, at times, a set of sophisticated tools that can help provide insight into the games we love. It can even be applied, as Dean Oliver has, to issues like team fit, but any good statistician will also be the first to explain to a decision maker the limits on the analysis. Are their people that take their analysis too far and draw conclusions that are not supported by their own work? Of course, their are irresponsible people in every profession. That does not mean, however that the tool is being over used or is in any way shifting a decision maker's focus away from the important variables.

Let us not have this debate any longer. Live and let live. Statisticians, scouts, fans, coaches and general managers can choose how much stock they put into various types of analysis, but lets not dismiss an entire field that is, honestly, still in its infancy.

Sports Analytics: Should Fans Care?

2011-06-20T16:42:00.000-07:00

I have not posted for while but the passing of Father's Day yesterday has inspired me. The main reason that Father's Day could have this effect on me is that I am occasionally and forcefully reminded that most of the world does not care one lick about sports analytics, and can even be offended by sports analytics. This reminder was given to me by, of all people, my own grandfather. My grandfather is a rather opinionated man (as any 91 year old has a right to be) and likes to occasionally make proclamations about the world, and how I am personally functioning in it.

His most recent proclamation was that people like me are ruining sports. The normal argument about how advanced statistical analysis is bad for sports rests squarely on the idea that it is not an effective tool. As my grandfather is never one to take the road well traveled, he instead took the opposite tact - sports statistics is too effective a tool. Yes, sports statistics is ruing sports, because the analysis takes all of the uncertainty out of sports. We are too good at what we do.

I thanked him for the complement, but it reminded me of some work I had done with Matt Futterman at the Wall Street Journal. In this analysis, I looked out how predictive baseball standings were on June 1 of whether a team would make the playoffs. Turns out the answer is very predictive (teams below .500 on June 1 has only a 9% chance of making the playoffs). This seems to be in line with my grandfather's argument, fans of teams that are below .500 in MLB now might as well give up. Which leaves me with the question: Did a simple correlation calculation suck all of the fun out of the baseball season for a large chunk of fans?

In my defense, I would like to offer two ideas that, regardless of how well statisticians can predict outcomes, should inspire fans to want more, not less analysis.

The first idea is that all leagues collect and report statistics. These statistics are used all of the time by writers and announcers to tell stories about a player or a team or a season. The problem is that many of these statistics are misleading, incomplete, or just plain wrong. For example, one of the most convoluted statistics in sports is the NFL QB Rating. If a QB completes a 4 yard pass on 3rd and 3, he has made a good play - keeping his team's drive alive, but his QB rating is likely to go down (I say likely as the exact calculation depends upon the QB's performance up to that point in the game), while if a QB completes an 8 yard pass on 3rd and 10, the team has to give up the ball, but the QB's rating probably went up. Having watched the two plays you may know that one QB made a good play, while the other didn't, but if you didn't watch the plays, and just looked at the reported QB ratings, you might get the wrong idea about the performance of the two players. Since we are going to be given numbers to look at and they are going to be used to tell stories, better to have the right numbers.

The second idea has to do with one of the many reasons we love sports: the incredible. Whether it be an incredible play, and incredible game, or an incredible season, the truly spectacular and unexpected moment is a unique aspect of sports. We rarely, if ever, have the truly spectacularly unexpected in any other form of entertainment. What good analysis does, is allows us to recognize again and again, how truly uncommon a moment was. The statistics can give us context for understanding how rare something we just saw, really was. Returning to baseball, the Pittsburgh Pirates are currently 2 games below .500, which means, at best, they have a 9% chance of making the playoffs. But what if they went on a run, maybe picked someone up at the trade deadline, and made the playoffs. The analysis allows us to understand how unlikely and special that would really be, and to appreciate it 4 years from now and remember how special that Pirate's team that beat the odds really was.

I would not try and argue these points with my grandfather, we have plenty of other battles to have, but I would suggest that most fans can appreciate, and in fact desire more and better analysis. When the writers and broadcasters embrace new statistics and analysis, then fans can understand the games they love better, and identity and appreciate the truly spectacular in a deeper and longer lasting manner.

Risk and Drafting a QB

2011-04-27T23:49:00.000-07:00

What kind of football related blog would this be without some kind of post related to the upcoming NFL draft? It was not difficult to know that a post was necessary, finding an angle that has not been done to death by the football media that, outside of some truly exciting court rulings, has had nothing else to talk about recently.

After surveying the landscape, and seeing yet another discussion of how many first round busts there have been (particularly at the QB position) but no QB in current crop of potential first rounders being labeled as the potential bust, I thought I would take up the mantle of uncertainty in the NFL draft.

As a believer in on field performance above all else (pro day work outs, combine performances etc) I dug into the on field performances of all the QBs that received a draft grade of 2 or higher from Sports Illustrated in order to see which player's performance might have the highest probability of being a mirage. The most direct numerical way to look at this is to build the 95% confidence interval around a QB's completion percentage. Using a player's number of attempts with their completion percentage, I calculated a range of values for each QB's "true" completion percentage. The wider the range, the riskier the pick is.

Based on this admittedly crude measure we can see that Cam Newton has the highest level of uncertainty around his college completion percentage. In his career, Cam completed 65.4% of his passes, but due to the small sample of passes attempted, his "true" completion percentage is somewhere between 59.9% and 70.9%. This is a fairly wide range of potential values, which suggest that we likely do not have enough information to properly evaluate Cam at this point. To be clear, I do not mean that statisticians do not have enough information, but that talent evaluators of all types, do not have enough information to determine Cam's true level of performance.

Cam has the highest standard error in this group at +/- 5.5% while the standard error of the next highest QB to receive a draft grade higher than 3 is Ryan Mallet with a standard error of +/- 3.1%. This suggests that the evaluations of players like Mallett and Gabbert are more likely to be accurate that the evaluation of Newton. This is not to suggest that Cam will not be a successful NFL AB, but that NFL teams should be careful to factor in the high level of uncertainty when drafting him relative to the other highly regarded QBs.

Player	School	Grade	Compl%	High	Low	Std Err.
Cam Newton	Auburn	3.12	65.4%	70.9%	59.9%	5.5%
Jeff Van Camp	Florida Atlantic	2.14	57.2%	61.4%	53.0%	4.2%
Scott Tolzien	Wisconsin	2.26	68.1%	71.8%	64.4%	3.7%
Ryan Colburn	Fresno State	2.11	62.1%	65.8%	58.4%	3.7%
Greg McElroy	Alabama	2.34	66.3%	69.9%	62.7%	3.6%
Jordan La Secla	San Jose State	2.1	59.1%	62.7%	55.5%	3.6%
Tyrod Taylor	Virginia Tech	2.36	57.2%	60.5%	53.9%	3.3%
Ricky Stanzi	Iowa	2.5	59.8%	63.0%	56.6%	3.2%
Ryan Mallett	Arkansas	3.02	57.8%	60.9%	54.7%	3.1%
Blaine Gabbert	Missouri	3.14	60.9%	64.0%	57.8%	3.1%
Christian Ponder	Florida State	2.83	61.8%	64.9%	58.7%	3.1%
Ben Chappell	Indiana	2.27	61.1%	64.0%	58.2%	2.9%
Jerrod Johnson	Texas A&M	2.33	58.6%	61.5%	55.7%	2.9%
Jake Locker	Washington	2.92	53.9%	56.8%	51.0%	2.9%
Taylor Potts	Texas Tech	2.22	66.3%	69.1%	63.5%	2.8%
Colin Kaepernick	Nevada	2.9	58.2%	60.9%	55.5%	2.7%
TJ Yates	North Carolina	2.38	62.3%	65.0%	59.6%	2.7%
Andy Dalton	TCU	2.88	61.7%	64.3%	59.1%	2.6%
Nathan Enderle	Idaho	2.41	54.6%	57.2%	52.0%	2.6%

Accepting an 18 Game Schedule

2011-04-19T11:46:00.000-07:00

The NFL owners have made it clear that they want an 18 game regular season and will push the players hard to get that concession. The primary objection to the 18 game schedule from the players point of view is of course the increased risk of debilitating injury which will likely shorten careers and lower the total earning potential of the individual players. While the incidence of injury within a season will most certainly rise with more games, the focus of much of the discussion undoubtedly also turns on the issue of concussions which is not well understood. After much public pressure and increasing evidence on the effect of repeated brain injury on football players, the NFL is taking steps to at least better handle concussions on the field. The question still remains though about how damaging concussions and other more minor repeated brain injuries really are which allows the NFL to be less than fully aggressive in minimizing not only the impact of specific injuries, but on monitoring and understanding the long term effects of these injuries.

If the players are serious about understanding the impact of repeated brain injury, as I suspect as least some of them are, then they should take a lesson from State Attorney General's from across the country when they sued the tobacco companies and insist on disclosure. Disclosure has been a part of many settlements with the tobacco industry in which the tobacco companies have agreed to make their internal documents available to anyone, including researchers in a broad array of fields such as chemistry, biology, marketing, economics, and political science. These documents have allowed these scientists to learn not only about the behavior of the companies, but gain a greater understanding of the broader health and policy issues related to tobacco.

The deal for the players should actually require a lot less from the NFL than these settlements required of the tobacco companies. Instead of forcing the disclosure of previously secret memos and financial information, the players should request the collection and disclosure of new information. In exchange for the 18 game regular season, players should request the following three items:

Brain scans at the begining and end of the season for every NFL player.
Insertion of a chip into every player's helmet that tracks the impact that the player's brain is exposed to both on a play by play basis and on a cumulative basis.
Disclosure at the end of the season of all the data (anonymized) to any and all interested parties.

This data would very quickly help researchers understand the full impact that repeated brain injury has on the brain throughout the course of an NFL season, as well as allow all interested parties to see the impact of various policies and equipment on reducing this impact. This data would go a long way from moving the debate from its current stage of arguing about how serious an issue it is, to figuring out the best way to protect players from these types of injuries.

Blind Side Project Help

2011-04-13T22:39:00.000-07:00

Last season I ran a pilot study on valuing offensive linemen. Some of the results are described here. The results are promising, but are clearly lacking in usefulness as only a sample of games of a few teams are covered. In order to move the process forward, I am looking for a few reliable volunteers who would be willing to assist in collecting the data. If you are willing to spend hours watching and re-watching plays to collect the needed data, I want to hear from you. All consistently contributing volunteers will have access to the complete data set.

The data collection requires some basic understanding of offensive line play as well as some training on a specific game to insure high quality and consistent data. If you are interested, please send an email to: quantsports at gmail dot com. In the email, please rate you knowledge of offensive line play and explain your rating. Also, please indicate which team(s) you would want be able to cover (you must have your own video access that can rewind and pause the games you chart) and how many games you would be able to do in a week (games typically take 2 to 3 hours to complete for experienced trackers).

Thank you to all willing volunteers.

Business Intelligence on YouTube

2011-03-27T19:24:00.000-07:00

After reading Freedarko exhort the wonders of YouTube and what it has done for basketball, I thought it might make sense to see what YouTube has done for analytics/business intelligence. What I learned is that a lot of companies like to advertise their products on YouTube and that some folks have no idea how to relate their quality audio content to meaningful video content. I also learned that there are some interesting BI related videos available on the web. Here are a few:

The Undisputed Guide to FREEDARKO

2011-03-26T17:17:00.000-07:00

Sitting on my sofa with a beer next to me and the battle for a seat in the Final Four of college basketball unfolds seems to be as near as an ideal setting for writing a review of Freedarko's Undisputed Guide to Pro Basketball History as it gets. Before discussing the book though, I must explain why a review of the book is appropriate on a blog devoted to analytics, I mean no where in their 223 pages do they even give so much as a shout out to Dean Oliver and the only advanced metric they utilize is the Jazz-O-Meter. The motivation is to remind the analytic minded among us, that there are elements of the sports we analyze that we can't get at with our metrics. There are stories that lie outside of our experiences and a richness to the sport that can be forgotten when we focus solely on the data.

My adventure with the Undisputed Guide began on my trip to the Sloan Sports Analytics Conference earlier this month. As I was packing, I look at the ever growing stack of books that I need/want to read and whittled it down to two choices: The Book of Basketball: The NBA According to The Sports Guy or FreeDarko Presents: The Undisputed Guide to Pro Basketball History. Deciding which hardcover book to slip into my computer bag was an easy choice, as Simmon's book would have required me to pack my back brace along with my computer, while FD's slim 223 page tome was significantly more portable. Once the decision was made and I found myself taxing for takeoff, I opened it up to soak in the knowledge.

What I got was an entirely unique view into the history of professional basketball. Starting with the inception of the sport, the FD team brings the reader from the barnstorming years (as a side note, some day there has to be a movie that features a game between the All American Red Heads triumphing of over the Terrible Swedes, has to happen) to LeBron with a unique point insight, humor, and totally original visualizations.

The visualizations are truly the key that separates FD from the rest, not just because of what each individual graphic communicates (ex. 20 years of draft history color coded for college experience and post-draft value) but how their inclusion drives home the point that what FD provides is the color. I am an analyst and I live in the data. FD brings the game outside of the data and reminds me that it is not just about delving deeper and deeper into the data to find the right answer, but that what we often refer to as "noise" is actually full of great stories, if not definitive answers.

One near perfect example from the book of this is the chapter on the ABA that is subtitled "What the Hell Was the ABA?". This is a provocative question, and it largely goes unanswered, mostly because there is no definitive answer. The ABA had a profound effect on the NBA in terms of salary and style of play, but it was also fueled expressly by business men looking to cash in. Large contracts were mirages and while the league was often at the cutting edge of marketing, they could also be as hokey as it can be. As an analyst I am used to asking questions and delivering the best answer that I can. FD presents a host of interesting questions, and reminds us that some of them are so mutli-dimensional, they really don't have answers.

After reading a serious of analytics and related books, settling back into the wonders of sport is a welcomed reminder of my original motivation for being involved in sports: the fun of it all.

Differentiating Sports Analytics

2011-03-26T16:19:00.000-07:00

I had a conversation recently with USF analytics professor Vijay Mehrotra that resulted in Prof Mehrotra's column in January issue of Analytics Magazine. As part of this conversation, I discussed with Vijay the general issues that arise when trying to implement an analytics program in a sports organization. After sharing my experiences, Vijay shared his wealth of experience in logistics and analytics in a variety of business sectors and we were both struck by the similarities. A very common story in sports and business in general, is for leaders in a business with a wealth of domain knowledge or faced with being told by people who have not have near the experience or "feel" for the industry how to change the way they operate. Understandably, domain experts are hesitant to embrace insights and new management styles from those that have not had the same deep experiences in the industry that they have.

This story in sports of course is a coach or general manager being told how to change their strategy or which players to add to their roster based on a series of calculations that they may or may not understand. For a business leader, the story is a CEO being told that their past "gut" level decisions have often been wrong and that they should change how they view their industry by looking more closely at information that is the result of a series of calculations that they may or may not understand. Clearly the data is different and the techniques used may be different, but there are many overlapping principles in how to both technically and culturally implement a strong analytics program in sports and in other industries.

The three elements of analytic systems (data management, predictive analysis, and information systems) are the same in sports and other industries. Many of the cultural hurdles are also the same, so is sports analytics a specialty within the general field of analytics or is it a field unto itself that draws on a similar set of tools? The answer is yes. Sports analytics is simultaneously a specialty within the broader field and a unique field that draws on a similar set of tools.

Explaining how sports analytics can occupy that space, lets explore why it is important to understand the uniqueness of sports analytics as a field. An analytics program has the highest probability of success when, going in, all parties involved have an understanding of the possibilities and value of sports analytics. If sports analytics is viewed as either a specialty or as completely distinct, some of the potential value of the program can be reduced.

Sports Analytics as a Sub-Field
When sports analytics is viewed as a specialty within the field of analytics, there is a tendency to seek out experts in analytics and ask them to build a product for a sports organization. The basic problem with this approach, is that consultancies that specialize in building analytic systems such as Accenture, have built their businesses by building systems that help managers maximize revenues. This can take many different forms and different industries have their own components, divisions, and peculiar industry structure, but they are all focused on profits. While I would never argue that professional sports teams are not trying to maximize profits, the sport side of the organization tends to be evaluated not on a profit basis, but on a wins basis.

Coaches that win keep their jobs regardless of how profitable the larger organization is, general managers who build consistent winners become presidents of the organization, regardless of their ability to manage the business side of the team. Therefore, the analytic systems that get produced for the sport side of the organization need to be focused on maximizing wins, and the difference between objectives (wins and profits) is not trivial.

Wins vs. Profits
Wins are a fixed resource. In the NFL for example, there are only 256 games and each team plays only 16 games. A win for one team means a loss for another. Each game is also a binary event, you either win or you lose. This creates a very different set of of objectives and manner of evaluation. In business, a project can be evaluated on a clear and consistent basis: does the investment in the project have a expected return on the invested resources that is high enough. On team, the question is not what kind of return on invested assets does an activity have, but rather can this activity increase the odds that the team wins one more game? The binary nature of the payoff has a significant impact on how information is used, the types of questions that get asked, and the impact that information can have.

Type of Information
The other major difference between sports and most businesses is the type of information that is used. Even the least analytic CEO has read a P&L, they have seen and used quantitative information on a regular basis through their careers. While coaches and GM's have certainly seen box scores their entire lives, the information that is far more important to them is the film. Coaches review film for tendencies and strategic information and play personnel folks watch hours and hours of film to evaluate players. This creates a unique situation in which the best analytic systems need to incorporate film as an additional type of data and that any new metrics that are developed need to jibe with what the decision makers see on the film. In once respect the analytic systems serve as a mechanism to help the decision makers more efficiently process the film that they see (ie point them to the most important and descriptive film).
Sports Analytics as Separate Field
The thoughts above may suggest that sports analytics is its own unique field and is only superficially connected to the application of analytic tools in business. This of course overlooks the commonalities of the fields and how these commonalities can be utilized to increase the effectiveness of both fields. The danger for the fields is that they, like information in many organizations, become siloed and the advances in one field never get leveraged to advance the other. While the data and objectives may be different, there is significant overlap in many of the tools and techniques used throughout the process.

Data visualization, for example, is a rapidly developing field and there is significant experimentation going on in business, academia, and sports. Feedback from end users will be gathered on the effectiveness and utility of various techniques in all areas, and only if the fields are connected, will these advances be rapidly deployed.

So in the end sports analytics is the same as regular analytics, but different.

Impact of defensive pressure, distance on pass completion

2011-03-23T20:06:00.000-07:00

The data from 2010 Brazil Serie A provides data on passing that has not been previously available. The first rule of good analysis is that when you get new data, the first thing to do is to look at the data. So I started by generating a historgram of distance of all 118,445 passes in the data set. The result (below) is much as expected with the mean distance of pass being 21.6 yards, with more passes below 20 yards than above 30 yards.

Click here for the rest of the article.

Who's Your Analyst?

2011-03-17T20:24:00.000-07:00

Teams, like all organizations, can easily become a group of silos. Each group is so focused on their tasks that is very difficult to make time to interact with other groups. This is usually not a good structure because one hand then rarely knows what the other is doing and instead of one team focused on complementing each other, the groups become independent actors focused only on doing what they do best. To put this in basketball terms, most organizations have a bunch of Allen Iversons - great at what they do, but not focused on the overall team goal. From an analytics perspective this kind of structure can lead not just to inefficient management, but poor analysis and often wasted time and money.

As teams look to expand their analytic capability, it is tempting to have a draft analyst, a pro player analyst, a game strategy analyst etc. and have each of these analysts sit under a manager in those departments. This is a initially seems like the proper solution and analyts become submerged in a functional area that they're supposed to become expert in and on a daily basis assist those that are trying to utilize and learn the information generated by analyst. This structure leads to two types of inefficiencies as mentioned above. The best organizations have, instead of putting an analyst function within a department, created a department of analytics that acts principally as a consulting group for internal clients.

Businesses that have structure themselves this way have generally found that analysts are able to communicate with each other on a more regular basis, allowing for more sharing of techniques and creative brainstorming around challenging analytic problems. Additionally, within the context of a sports team, having a central analytic consulting group allows for consistency in the language and style in which the analysis is presented.

The consistency of the message is particularly important in an organization that is trying to incorporate a type of information that it has not utilized in the past. If every analyst has their own style and manner of presenting data, than instead of one core institutional language, each function within the team will have their own language. On team where, eventually the scouts, coaches, trainers etc all have to get on the same page, having one core analytic language means that no one has to interpret between the groups.

An additional benefit of this structure is the ability to rapidly and efficiently deploy the analytic capacity within the organization. Different, and predicatable times during the season, different departments may require more analytic capability than their normal baseline needs. For example, as the draft approaches, the amateur personnel department may have a greater need for analysts than they do in the months after the draft or coaches may want extra analytic fire power if the team makes the playoffs.

To be fair though, most sports teams have yet to reach the point that this is an issue. The analytic capabilities are at this point mostly one or two individuals who do often work as an internal consulting group. But as teams begin to invest more in their capabilities in this area, these issues become inevitable and must be carefully thought through. Happily for those starting up now though, much of the research in this area has been done and the results are clear.

Rethinking the NFL

2011-03-14T20:46:00.000-07:00

The current labor dispute in the NFL will be negotiated and litigated over the coming months. At its core though it is not just players against owners, but rather a three way negotiation featuring the players, against the small market owners, against the large market owners, against the players.

This of course is not an original insight, but a point that is important to keep in mind as we think about what the NFL could look like. If the litigation all actually goes to judges and juries for final decisions and the appeals are all exhausted, one possible, out come, is that the NFL is declared, by law, to be acting as an illegal monopoly in violation of anti-trust law. If this occurs, then what have become standard labor practices, such as the draft and the salary cap, will essentially be banished forever. It is worth considering what the NFL becomes in that situation, if for no other purpose, than for all parties to understand what is at stake so they can get back to the bargaining table and put a new CBA together.

Once all 32 NFL teams are prohibited from working together to set the labor market for professional football players, there will be teams who go high (imagine Jerry Jones with no salary cap) and some will go low (imagine Mike Brown with no salary floor). Economic theory and common sense both actually agree on the outcome here - a league with a few very good teams that dominate competition and a few awful teams that do not even have a reasonable chance to win 4 or 5 games a season.

One this situation is reality, then the league will either become unwatchable (how many times do you really need to see the Globetrotters beat the Generals) or they can move to a radically different league structure. There is of course a workable league structure for a league with this type of financial structure that is in practice in many parts of the world: hello relegation!

In an NFL with relegation, we could have three divisions, with the top 12 teams in the first division, the next 12 in the the second division, and whoever is left, or wants to start a team in the third division.

In a relegation/promotion league, teams can play anyone and earn points based on the quality of teams they beat (a win against a first division opponent is worth more than a win against a third division opponent). Teams that earn the most points either stay in the first division or get promoted up to the first division and teams that do not maintain a high enough point total over a season get relegated down to lower divisions. Each division can have its own playoff system so even the Bengals can have a shot at a playoff game every once in a while.

Then there is what I'll refer to here as the Ellison Effect. Imagine if anyone could just start a professional football team and start competing in the Third division. WOuld Larry Ellison (or any of the other billionaires laying about) be interested in perhaps starting a team or two or three in Las Angeles? Perhaps an extra team in Chicago? Sure they would take some losses as they started up and tried to move up the division ladder, but the financial promise of the First division would be more than enough to tempt a few wealthy business folks to give it a try.

I could not begin to put a probability on this scenario, but it is one that is intriguing to me, because I think the drama of teams starting up and recruiting players as well as seeing teams compete to keep their place in the higher divisions would be exhilarating.

From the perspective of most of the owners though this is probably a less than enticing scenario. They have grown to enjoy their multi billion dollar TV contracts and packed stadiums. The prospect of having to hope to fill a 20,000 seat stadium in the third division (I'm looking at you Buffalo) is probably not something that too many owners really want. Players too, at least as a collective, probably don't love the idea that half of the high paying (and high minimum contract) jobs could be gone. So with that incentive, I invite the various groups of owners and players to continue their negotiations, and not make the judges decide the future of the NFL.

Analytics and Communication

2011-03-12T07:09:00.000-08:00

In his excellent book on the research around the theory of Deliberate Practice, Geoff Colvin distills volumes of often complicated academic research into a clear understandable prose that is easily accessible to a wide audience. Colvin uses a series of carefully chosen anecdotes to explain the various dimensions of this complicated theory (yes, there is more to it than 10,000 hours by the way). What is remarkable about the book, is that it accurately reflects the messages and strength of the research that has been done. There are varying levels of evidence from the research for different aspects of the theory, and those points are clearly made. The goal of this post is not to sell more copies of Talent is Overrated (though I have provided a link in case you are interested), but rather to make the point that clearly communicating research is at least as important as the work itself. This is just as true for statistical analysis as it is for scholarly research.

At the 2010 Sloan Sports Analytics Conference, I was sitting next to a high level NBA executive at a research presentation. The work being presented was interesting if not revolutionary. When the presentation was over, I went to the front of the room to ask a few questions of the presenter, but was beaten to the punch by the exec I was sitting with who said (and I paraphrase here a bit) "oh my God, you can talk". And it was true, the presenter had distilled some very complicated analysis down to the core message and accurately conveyed the strengths, potential, and limits of the work in such a way that audience could clearly understand it. If the presenter had not been able to do communicate his research to a non-geek audience (ok, it was SSAC 2010 non-geek is overstating) he would not have been speaking to a full room by the end of his presentation, never mind having extended conversations with NBA execs about it.

Communicating statistical analysis is a careful balance between the strength and limits of the analysis. The story of the analysis has to be conveyed in such a way that a non-geek user of the information can use the analysis properly. Your projection may show that a player is going to improve their rebounding by 20% over the next 3 years, but you also need to convey the risk associated with that analysis. What are the range of likely outcomes? What are the risks?

It is tempting when when working with team executives to make it all too simple and speak in absolutes, especially when others are making similar statements about their point of view. It is incumbent on the analyst not to fall into this trap though, because our analysis does contain variance and we will be wrong. When we are wrong, it becomes easy to dismiss the analysis if we have spoken in absolutes, while if we have strongly communicated (and accepted ourselves) what are research actually says, then, while we may not win every argument, we will win more over the long term.

It is also possible to believe that we are so clever in the techniques we have used to solve a problem, we lose sight of the problem we were trying to solve. It is rare that you will run into an exec who understands or truly cares about how cutting edge the techniques are. They want to know that they are getting good information that they can have confidence in, not that you used some slick new neural network algorithm in R to get the slickest results. This is one of the reasons the communication piece can be so tricky. We have confidence in our results because of the techniques used, and while you may want to have the one sentence description of what you actually did ready in case they ask, management will only have confidence by seeing the results.

After spending hours and hours carefully constructing your analysis, be sure to put a significant amount of time into deciding how to present it. Think like your audience, and what will help them use the analysis properly. If you don't communicate the analysis effectively, then the analysis will be wasted.

Players: One Definition

2011-03-09T20:11:00.000-08:00

Ask a business man to define a customer, their definition will depend largely on their function within the business. If they are in sales, they will talk about points of contact and sales histories, but if they are in product development, they will talk about demand for new features and usability. Finance, marketing, and H&R would also likely have different definitions for a customer. It is not hard to see however, how a business could benefit from having one comprehensive definition of what a customer is, that brings all of that information together in one place. R&D could then look at sales records to see if the new features under development match the needs and wants of the most profitable customers.

A sports team is no different. Every function within the sport side of a professional team has their own definition for a player. Coaches are often focused on current performance, while the personnel side is often focused on scouting information, and trainers are often focused on health related information. As each of these groups has interest in the information that the other groups have, that is not their focus, and it is often difficult for the various types of information to be synthesized.

Coaches collect, process, and analyze a wide variety of information. Game data includes quantitative data such as performance metrics related to play calls, video, and qualitative grades. Practice data often includes video, specific measurements, and observations. Classroom information includes information on preparedness, and the ability to understand and process game plans.

The personnel operation collects information from a variety of sources. Intelligence is generally qualitative information on a player's background and any current personal issues. Personality information may include quantitative or qualitative psychological assessments as well as anecdotal information gathered from other layers/coaches/friends. Specific skill data can be quantitative or qualitative information on a player's strengths and weaknesses.

The trainers and medical staff focus on a rich set of qualitative and quantitative information regarding a player's injury history including type, treatment, and recovery times. THey often use information on how players train and the frequency with which they train, as well as their pre and post training routines. They also monitor nutrition and hydration.

These are all of course gross simplifications of the broad classes of information utilized by different functions on a team. What is important is not that specific types of information, but rather the synergistic value of keeping one definition of a player that is updated and analyzed by all functions within an organization.

Once all the types of raw data and information on a player are collected consistently collected in one place, then the coach who is wondering how to better motivate a player can see from intelligence information what other coaches have done in the past, or the general manager who is wondering why a high potential prospect is not developing can see that they are not processing information in the classroom well and has poor post-training habits.

While all of this already occurs, it usually occurs when a decision maker in one function asks for information from another function. This may take 5 minutes for a response, it may take 3 days - by which time the original thought that led to the request is gone. An organization that has one definition of a player, has a system that allows thorough and creative analysis to flow freely and not be constrained by the response time of other members of the organization.

The bottom line is that having one definition of what a player is, focuses an organization on what it believes is most important about a player, and drives the resources, strategic thinking, and tactical analysis of the organization through that definition. That process gives the organizations long term strategy the best probability to succeed.

Why Sports Analytics?

2011-03-08T21:01:00.000-08:00

The "Why" of sports analytics is actually fairly obvious: good sports analytics can help a team win more frequently and more consistently. The real question with sports analytics is not "why?", but "how?" By how I do not mean how do we set up a good sports analytics program - this is a vital question, but not what I am tackling today. Before you determine how to implement a sports analytics program, you first have to understand how sports analytics can help your team win.

Sports analytics can help a team win more frequently and consistently through three general functions: New Information, Synthesis of Information, and Efficient Information. Each team will emphasize a different function of sports analytics based on their core competencies, but a comprehensive sports analytics program will involve all three.

Functions of Sports Analytics

New Information
The classic use of sports analytics is to provide new information - usually through statistical analysis of performance or game data. This function uses raw, quantitative data to create new information. Team's in all sports have done this privately, creating metrics and projections that help them better measure current and future performance of their players as well as how to best utilize the players they currently have. The power of new information is obvious when compared to the investment community. Investors that have information that other investors do not about the future earning power of a company have a competitive advantage, this why there are laws governing insider trading (hello Martha Stewart).

New Information & the Draft
In sports, if the draft is akin to acquiring an asset in the investment market, than knowing more than about the future potential of the draft eligible players puts your team at a competitive advantage, and more likely to out perform your draft slot - that is you are more likely to draft a player who will perform above the average player selected at that position. One example of this is Chase Budinger and the Houston Rockets. The Rockets are know to have the most advanced analytics program in the NBA and in the 2009 draft they put that program to use. The Rockets traded a future 2nd round pick to the Detroit Pistons for the 44th pick in the draft and selected Chase Budinger. In less than two seasons, Chase has logged over 2700 minutes, hits 35% of his 3s, and has become a regular starter. For an idea on how significantly Chase has out performed the 44th pick, previous players drafted 44th overall include: Reyshawn Terry (yes, that Reyshawn Terry), Tim Pickett, and Lonny Baxter. Houston used their analytic draft system based on college data to provide them with better information on Chase's probable impact on the NBA.

Synthesis of Information
Sports teams, like all organizations, are filled with a many different types of information. There are vasts amounts of quantitative data such as player and team performance metrics, qualitative information such as scouting reports, and hyrbid information such as medical reports. Typically these all reside in various hard drives, spreadsheets, disjoint databases, and word documents. Synergies between the various types of data often exist. That is, putting financial information together with performance data does more than just bring to two types of information together, it creates a new type of information on the value of the player. Good analytic systems bring these various types of information together not just to house them in one place, but to combine them and extract all of the synergistic value out of them.

Efficient Information
Top decision makers with pro sports teams may actually be the busiest people on earth. They are constantly being bombarded with information and requests for their time. Agents, players, fans, owners, and other related parties are always looking to get some time with the top guys and each game, workout, practice, news event etc create more information for them to consider. As Dean Oliver said "stats can see all of the games" which means that decision makers, if they have the time, can consider all of the games. These executives have deep knowledge about their sport, and no one is in a better position to leverage the information that exists into a competitive edge than they are, the limiting factor on them is generally not resources or information, but time. Good information systems give them back some of their time, by creating a more efficient process for them to interact with the information. If an efficient analytic system can easily save a decision maker 10 to 20 minutes a day, that translates into 5 to 10 extra hours a month. That is extra time to spend watching more film, talking through more trade options, exploring additional strategies, and processing even more information. These decision makers will gain a competitive advantage by being able to accomplish more than their adversaries in the same amount of time.

New, better, and efficient information are paths by which a careful investment in sports can yield more wins. It has been estimated for non-sports teams, that an investment in an analytic system with a statistical component is approximately 145% (Davenport & Harris, 2006). While there is no comparable estimate for sports teams, it is not hard to see how analytic systems can help create more wins and keep winning teams on the winning path.

Advice for the Aspiring Sports Analytic Professional

2011-03-05T14:07:00.000-08:00

After two very interesting days at the Sloan Sports Analytics Conference I am more convinced than ever that the most significant improvement teams can make, the area in which they can gain a huge competitive advantage, is in data management. Lets be clear, data management is not exciting. Creating great information systems is not going to get you in the draft room arguing about who to take with the 4th pick in the draft. It is however the skill that is most likely to get an analytic minded person employed with a sports franchise.

Students often ask what they can do to get a job with a team, and more and more, my answer is to acquire good data management skills. Front offices do not lack for people offering them the next great statistic that will tell them who they should be adding to their team. Some of these metrics may even be useful, but teams already have plenty of people telling them, so they probably won't know if yours is actually informative, and they are already bombarded with so much information, thinking about integrating one more type of information is painful to think about.

And that is exactly why data management is the key to the kingdom. Right now sports teams have more data than they know what to do with. The data comes in many forms, from scouting reports, contract parameters, performance data, practice data, injury reports, the list goes on and on... The problem teams have is not that they don't have enough data to consider, but that they don't have a mechanism to efficiently consider all of the data that they do have.

For most of the teams that I have spoken with, data sits in a variety of places. This may include a centralized database that has even a good percentage of the team's data in it, but there are still multiple data silos within the organization. They have their geek box who keeps reams of useful data on excel spreadsheets, they have a trainer who has a ton of health/training related data on their computer, a psychologist who has a host of data (both quantitative and qualitative) on their computer, the cap manager may have their detailed league wide cap model on their computer, and none of this data is collected and processed together. For a decision maker to truly consider all of the different sources of data that they ALREADY USE, even if they have access to it all from their computer, would likely have to have seven or eight windows open, and then scroll from window to window trying to connect the dots between the various data sources. This is a major time waster for people that place a premium on time.

Good data management which includes data structuring, processing, and front end information systems could easily save a group of top level decision makers for a sports team multiple hours a day. Any analytical professional that can demonstrate to a team that they can pull all the various pieces of information within an organization together, and turn it into information that a decision maker can access from one open window, or one app on their phone, would enable that decision maker to spend more time actually strategizing, processing information, and making the team better, instead of having to find and synthesize the data.

So all of you brilliant students desperate to work in sports, give up on finding the next PER or adjusted +/-, and get the skills necessary to help give decision makers back some of what they value the most: time.

Get Your Geek Out of the Box

2011-03-03T14:37:00.000-08:00

Organizations that have taken use statistical analysis often do so with a “toe in the water” approach. They hire a well regarded blogger to do a project to see if the information might prove useful. This usually takes the form of some sort of player projection system. The front office then huddles around the results and decide if the analysis points out anything useful to them. If so, they may even bring the analyst in on a full time basis to provide them with information. The analyst then gets a desk in the team offices and, equipped with a high powered computer and the newest version of Excel (and maybe R or Stata) they get to work mining data for useful insights.

This is often the end of the story though. The analyst works diligently and adds value to the organization, providing useful information and everyone is happy. But the organization is not maximizing the value of the analyst, because all the useful insights into player value and game strategy that the analyst may bring to the table are put into a special geek box. All of the other coaches and personnel folks open the box on occasion, look at the information, and even utilize, but then they put it back in the box and do not see it again until the next time they choose to open the geek box.

This leads to a situation in which, when the non analyst wants some information, they either have to go find the geek box or find an alternative geek box. It is often, particularly for coaches out on the road, easier to find information that looks like the information from the geek box online from sources like ESPN or HoopData. So when a coach wants to know something about their opponent, they may turn to alternative sources – outside of the organization.

The reason the geek box is a problem is that it creates multiple versions of the truth within the organization. Consider the situation in which an analyst has, through careful analysis, developed a metric that isolated a RBs contribution from their offensive line’s efforts. Occasionally the general manager opens the geek box and sees that, while his team has a strong running attack, their RB ranks near the bottom of the advanced metric. The head coach has been too busy to open the geek box and instead saw quickly on ESPN that his RB is in the top five in the league in yards per rush. There are now two very different impressions of the team’s RB floating around. Now, instead of the GM and coach meeting to discuss how to solve their RB problem, they are headed for conflict as to whether to offer a big extension to the RB or not.

The solution to the geek box problem is an enterprise wide approach to analytics. This is the opposite of the toe in the water strategy, it is the canon ball off the high dive approach. It puts all of the various forms of information that sit within the organization – scouting, statistical, medical etc in one place, integrates it all, for anyone within the organization to easily access and utilize. Now, when a scout is on the road and wants to have a more complete view of the player they are about watch, they instantly have the all of the relevant information on the player, and it is the same information that the GM and coach are looking at, so when the three individuals are doing their analysis, they are starting from the same place.

Making information more accessible to all, and keeping everyone on the same page has significant benefits to the organization. My bias is of course to point out the benefit of having the statistical information easily available to all, but the truth is it goes both ways. As analyst, we look to inform the rest of the organization, but we can also learn a great deal from seeing the information that everyone else in the organization is looking at, so when we have conversations, we are starting from the same place as well.

It is time to take the geek out of the box and have a full enterprise wide approach to information in sports organizations.

NFLPA's Secret Leverage

2011-03-03T11:49:00.000-08:00

Judge Doty’s recent decision to relieve the NFL owners of their $4billion extortion fee from the networks garnered the predictable reaction from the NFL that the decision would have no impact on the ongoing labor negotiations. That reaction is obviously as phony as a tight rope walker suggesting that removing his net would have no impact on his decision to have a shot and a beer before attempting a high wire walk. The question is not whether this decision would have an impact on negotiations, but rather how significant the impact will be.

Mark May of ESPN was recently quoted by Wilbon on PTI that while the owners would certainly like to have the cash, their pockets are so deep that it really does not affect their calculus on the labor front dramatically. This comment is born of May’s personal experience as a player negotiating against many of these owners. May’s experience though was at a different time, and the NFL and the US economy have gone through some major transformations since the last time there was a work stoppage in the NFL.

During the first six or seven years of the last decade, while the NFL continued to solidify its position as the most dominant for in professional sports, there was a burst of economic growth and activity across virtually every industry in the US. CEO’s across the country were pushing to realize higher and higher rates of return for their investors and many leveraged their assets to produce those returns. The debt that many of these companies took on finally proved to be too extensive as the economic crisis of 2008 hit. Many previously very successful business were forced either into bankruptcy or to sell out at pennies on the dollar to rivals.

NFL owners were operating in the same economic world as Washington Mutual, General Motors, and hundreds of small businesses that no longer exist. It occurs to me that it may be a bit naïve to believe that none of the NFL owners got themselves over extended and perhaps are carrying far more debt than their fellow owners realize. If any owners are privately facing this situation, they have been able to delay the reckoning that other businesses have gone through because the money tree that is NFL ownership has continued to flower during the economic downturn.

Without the $4billion from the networks though, for the first time, there may be some owners out there that are looking at the money tree and wondering if getting a labor deal done quickly isn’t the best way to get the tree to flower. The alternative may be rather uncomfortable for some of them.

Sports Analytics is Not a Strategy

2011-03-02T09:46:00.000-08:00

As I travel towards the Sloan Sports Analytics Conference I am pondering what sports analtyics actually is. Part of that thinking is spurred by the book Competing on Analytics by Davenport and Harris. One of the first points that Davenport and Harris make is that analytics is not a strategy in and of itself, but rather a tool that should be used to support the "core competency" of an organization. This point is, I believe, one of the core misunderstandings of what it means for a team to start using analytics. There seems to be an impression that statistical analysis (one tool of sports analytics) is a strategy that dictates a certain style of play and/or a certain type of player.

The legendary book Moneyball by Michael Lewis served sports analytics extremely well in detailing how a team that competes with analytics can prosper. The book also, however, reinforced the impression that analytics had a unique purpose - unearth undervalued players and find areas of in

game strategy for which the excepted wisdom was not entirely correct. While this is certainly a viable use of analytics, it is hardly the unique or, for some teams, even the most productive use. The A's employed analytics in this manner because it fit with their core competency: winning with a low payroll. It was the strategic goal of the organization that determined how to best use analytics, instead of analytics dictating a strategy to the team.

Teams have their own unique core competencies and areas in which they choose to have as strengths. In the corporate world, Walmart competes on price so they utilize analytics to better manage their inventory. In the NFL, the Ravens have traditionally competed on elite level defense and evaluation of collegiate talent. Should the Ravens choose to use analytics to help provide a competetive edge, then their initial focus should principally be in support of these functions, which is fundamentally different from how the Saints or Colts might choose to best deploy their analytic assets.

At this stage in the growth of sports analytics, it is incumbent on the analysts to communicate to team executives, that analytics is not about identifying the "best" strategy or players, but rather to work within the core competency of the team and help push that forward. For the most part however, we as analysts have done little to demonstrate the value of analytics beyond identifying new metrics to rank teams and/or players. While I am certainly guilty of this, and to a large degree it has been necessary, I believe that we need to start pushing the discipline beyond this and demonstrate to teams how analytics can be used to help make better decisions, develop a culture of evidence and uniquely support the core competency of each team.

Analysts working for teams that place a high value on their player development function should have a very different focus than one that focuses on finding value in free agent markets. Having had the privilege to work with a variety of talented executives, every analyst should understand that these leaders understand their sport at a truly elite level and have developed long term strategies that they believe in. When an analyst can demonstrate how a strong analytics program can function as a tool to insure that the tactics used to support those strategies are the ones with the highest probability of success, then they will maximize their value to the team.

Week 10 Picks

2010-11-10T20:54:00.000-08:00

I am feeling pretty good about my return to publishing my weekly picks after last week's 85% accuracy on winners. I won't pretend that that was typical week, but it sure is nice to have that number to brag about. Since I have a jump on getting the picks up early this week, I might as well focus on the Thursday night game.

The QB darlings of the 2008 draft meet this week as 3rd overall pick Matt Ryan plays host to 18th overall pick Joe Flacco and his Baltimore Ravens. These two QBs have certainly earned their position as the top 2 QBs taken in their draft as they have 47 starts between them and the rest of the QB class of '08 has a combined 11 starts (with Henne logging 8 of those). Both teams sit atop their divisions with 6 wins and are coming off of important home wins. To further drive the point home about the evenness of the matchup, the point spread is 0.

The fact that neither team is favored though tells us something about how the bettors see these teams. Since home field advantage is a well established concept, if these teams were even, we would expect the Falcons to be favored by two to three points. Since they are not favored, it seems that those with money on the line, believe the Ravens are the better team - and the data agrees.

My model has the Ravens winning 21 to 18, and while the Ravens are known for their defense, it is the offense that is going to make the difference in this one. Flacco is picking up 7.3 yards per attempt and he will be facing a defense that is giving up 7.9 yards per attempt. Add to that that the Falcons are giving up over 5 yards per rush attempt on first down, and you have a recipe for the Ravens to control this game from the beginning. On the other side of the ball, Ryan is picking up 6.7 yds/att and the Ravens are giving 6.8 yds/att - not too much mystery here.

And the rest of the picks:
Tampa Bay (21) vs. Carolina (20)
Indianapolis (22) vs. Cincinnati (20)
Buffalo (16) vs. Detroit (17)
Jacksonville (14) vs. Houston (13)
Denver (10) vs. Kansas City (13)
Chicago (16) vs. Minnesota (10)
Pittsburgh (21) vs. New England (10)
Washington (17) vs. Philadelphia (20)
Arizona (17) vs. Seattle (16)
San Francisco (7) vs. St Louis (12)
Miami (20) vs. Tennessee (22)

Week 9 Picks

2010-11-05T22:35:00.000-07:00

After a few weeks off, I finally have my act together enough to post my weekly picks before the games are played (trust me, I was great in the weeks I didn't publish, only missed two games the entire time). All the picks are below, but in going through the data, I wanted to focus on one pick in particular. It should come as not shock that the model picks the Steelers to beat the Bengals in Cincinnati. The Steelers are 5 and 2 and at the top of the division while the Bengals are 2 and 5 and in the cellar of the division. The data though has two interesting bits of information, one about the Steelers, and one about the game.

The Steelers defense is in line to be epically good against the run. On 158 attempts, the Steelers' opponents have logged a total of 412 yards or 2.6 yards per carry. The next closest average is the Jets which give up 3.4 yards per carry. Since 2000, only 3 defenses have finished the season allowing fewer than 3 yards per carry ('00 Ravens 2.7, '07 Ravens 2.8, & '06 Vikings 2.8). Even more impressive perhaps, is that they allow only 1.9 yards per carry on 1st down. Since none of their remaining opponents have truly dominant run games (depending upon how you feel about the Jets and Raiders) they could finish the season allowing less than 2.5 yards per carry. Historic defense really is sexy.

Now for the game, the model suggests that while there is a 66% chance the Steelers will win, there is a 70% chance that they won't cover the 6 point spread. You have no reason to believe this, particularly after I just got done explaining how epically good the Steelers defense is, but give me a chance to explain. The Bengals passing attack is about average and so is the Steelers pass defense, particularly on first down. What the Bengals are doing well though is taking advantage of the fact that they gain 50% more yards on first down when they pass than when they run. They have the fourth highest first down passing rate in the league, which will allow them to continue to do what they do well, most frequently. The Steelers on the other hand get more than double the yards passing than running on first down yet have the lowest first down pass rate in the league (32%). The Bengals are using their limited resources in a far more efficient manner. This effective use of resources may not be enough to win the game, but it will likely be enough to keep it close.

On to the rest of the picks:

Minnesota (28) vs Arizona (20)
Buffalo (14) vs. Chicago (17)
Green Bay (26) vs Dallas (19)
Philadelphia (24) vs. Indianapolis (21)
Oakland (18) vs Kansas City (21)
Baltimore (21) vs. Miami (16)
Cleveland (20) vs. New England (26)
Carolina (14) vs. New Orleans (21)
Houston (18) vs. San Diego (21)
Atlanta (24) vs Tampa Bay (23)

Note: Giants and Jets predictions are available in the New York edition of the Wall Street Journal.

How 18 Games Could Change the Game

2010-10-13T20:48:00.000-07:00

The "Enhanced Season" is all the rage in the NFL owner circles as they see the large increase in revenue associated with 2 more games and think that is the kind of enhancement they like. As a football fan, two more glorious Sundays of NFL action sounds pretty good to me too. The players have a rather obvious and valid objection to this lengthening of the season though; they are more likely to get injured in an 18 game season than a 16 game season. If we can assume though that players and owners can come to an agreement so that the players feel fairly compensated for the added risk of injury, the issue then shifts to the effect that a longer season would have on the product that is put on the field every week.

The first pass answer is that the product will suffer, because the inevitable increase in injuries will cost teams more missed games by their best players which means more Sundays watching Todd Collins going toe to toe with Jimmy Clausen and frankly, I'm not sure how much more of that modern quarterbacking can handle. This assumes though that teams will not change they way they build rosters and play the game with this shift in the season.

While not all teams will change immediately, I find it hard to believe that GMs and Coaches will not be able to take this fundamental change to the game (more player games lost due to injury) into account when building rosters and playbooks. The question then is not if the product on the field changes, but how it changes. I put my normally data obssessed brain to work on this issue and tried to think through what changes we might see in player personeel and strategy as the league enhances their season. The biggest shift I could come up with was the use of QBs and the type of QB that would be in demand.

Specifically I see more use of the "3rd down specialist" QBs - Do you remember the tail end of Ron Jaworski's career as it conincided with the beginning of Randall Cunningham's? The Eagles dabbled with platooning this QBs so that Cunningham played in certain situations, so that he gained experience in certain limited situations. He was a rookie so he couldn't master the entire playbook, but he could get up to speed on a portion of it, so when they needed to run a play of the type that Cunningham knew, he could go in a run it.

The Eagles did this because they knew that Jaws was soon to be done and they needed to see if the exciting but inexperience Cunningham could develop into a true NFL QB. Teams may like to go more to this model, not just for developmental purposes, but so that there is a built in reason for them to have two QBs taking significant reps during the practice week. Teams will not have pure backup QBs, more QB 1a and QB 1b who more or less split up the practice reps and are both used (while healthy) during the game.

This could take the shape of the "3rd down specialist" ala Cunningham, or it could become more like the baseball model with relievers/closers that come in and either manage a game that their team is winning or start bombing away if their team is losing.

This kind of shift in the perspective of the role of the QB gives coaches significantly more flexibility to set up game plans to maximize the talents of two different types of QBs, forcing defenses to be prepared for both Kolb and Vick or both Young and Montana. In order to get the most out of this shift, then, teams would want their 1b QB to actually be as different as possible from the 1a QB. Pur passing QBs would be paired with non-traditional running QBs, and those running QBs would no longer be taught stop running and look to throw, but to look for those running lanes. There would be less interest in teaching the running QBs to be traditional because teams would not be dependent on one QB whose injury could undermine an entire season. If one of the QBs goes down, you still have someone who is essentially a starting QB - they will just now have to play the entire game.

If this scenario plays out, does it enhance the product? Does the longer season actually create a more engaging product or just more of the same (or more of something that is lesser than it currently is)? I would argue that this could actually be a renaissance in offensive football, forcing/allowing coaches to be more creative and flexible and trying to do things on the field that have not been done previously.

So this little brain exercise has left me hoping that the season is lengthen because it may truly enhance the product.

Week 5 Picks

2010-10-09T23:26:00.000-07:00

I had a very poor 39% showing straight up last week, but a rather profitable 62% against the spread (would have been even more profitable with proper risk adjustments on any theoretical wager). Picks (less the Giants and Jets which are only available in the New York WSJ) are below.

Cleveland (14) vs. Atlanta (22) Confidence 55%
Carolina (17) vs. Chicago (20) Confidence 51%
Baltimore (23) vs. Denver (21) Confidence 88%
Washington (22) vs. Green Bay (15) Confidence 56%
Buffalo (20) vs. Jacksonville (18) Confidence 71%
Indianapolis (24) vs. Kansas City (27) Confidence 54%
Arizona (18) vs. New Orleans (20) Confidence 61%
San Francisco 18 vs. Philadelphia (20) Confidence 67%
Oakland (17) vs. San Diego (21) Confidence 88%
Detroit (10) vs. St. Louis (21) Confidence 75%
Cincinnati (22) vs. Tampa Bay (18) Confidence 88%
Dallas (17) vs. Tennessee (26) Confidence 64%

Northern California Symposium on Statistics and Operations Research in Sports

2010-10-06T15:48:00.000-07:00

One more plug for my upcoming conference. Tickets can be bought here. And the schedule for the day is:

Schedule

8am – Registration Opens, Continental Breakfast available

9am – Welcome

9:10am – “The intra-match home advantage in Australian Rules football”

By Richard Ryall & Anthony Bedford

9:45am - “Pitcher Accuracy through Catcher Spotting: Assessing Rater Reliability”

By Andrew C. Thomas

10:20am – “Optimal Dynamic Clustering Through Relegation and Promotion: How to Design a Competitive Sports League”

By Martin L. Puterman and Qingchen Wang

10:55am – “Dynamic Effort, Sustainability, Myopia, and 110% Effort”

by Stephen Shmanske

11:30am – Featured Speaker: Sig Mejdal of the St. Louis Cardinals

12:15pm Lunch Provided and Poster Session

1:45pm – “Drafts versus Auctions in the Indian Premier League”

by Tim Swartz

2:20pm – “Reconsideration of the best batting order in baseball: Is the order to maximize the expected number of runs really the best?”

by Nobuyoshi Hirotsu

2:55pm –“Uncovering Football’s Best Goalscorers in the English Premier League, La Liga, Serie A, Bundesliga, Eredivisie, and Ligue 1 for the 2009-2010 Season”

By Joel Oberstone

3:30pm – “An Empirical NFL Draft Value Chart”

By Michael Schuckers

4:05pm – Featured Speaker: Roland Beech of Dallas Mavericks

5:00pm – Concluding Remarks

For further information on each presentation, see the presentation descriptions.

Blind Side Project : How Good is D'Brickashaw Really

2010-10-05T13:26:00.000-07:00

D'Brickashaw Ferguson is the left tackle for the New York Jets. He provides an interesting example to begin the ranking of offensive linemen because he has been in the league for four years, was the fourth pick in the 2006 draft, was selected to his first Pro Bowl last season after starting every game of his NFL career. Expectations for D'Brickashaw were high coming out of college and he appears to be realizing them. He also protects the blind side of a second year QB who many feel is below average and believe that his gaudy 8.5 adjusted yards per attempt and 0% interception rate is due more to short quick passes and loads of YAC than his true QB skill. This suggests that D'Brickashaw is not serving an insurance role as much as he is creating value for the Jets offense.

To get an idea of how effective a LT D'Brickashaw really is I used the data set that I have collected with the tremendous assistance of Keith Goldner. We now have 9 teams in the data set and enough pass attempts to make a first pass at evaluation. In order to evaluate linemen we will borrow from those that came before us at footballoutsiders.com (as well as many others) and compare performance to average. In this case though, what we are going to measure as performance, is the probability that, for a certain length of time, the lineman will be able to successfully hold their block.

Warning: Technical Stats Jargon in Use: Trying to estimate the average probability that a LT can hold a block for 2.5 seconds is tricky because the data is censored data. In this case, the phrase "censored data" means that on any play that a lineman holds his block successfully and the QB throws the ball after 2.5 seconds, we do not know how long the lineman could have held that block. The lineman may have been about to lose the battle, or maybe he could have continued to protect the QB for 2 or 3 more seconds...we will never know. This censoring of the data can lead to a biased estimate of the probability of success. Happily statisticians have built tools to deal with this problem, namely Cox Regression which is used in this first pass. Estimates are done by position.

Ok onto the results. Remember the linemen score reflects the probability that the player adds or subtracts from the estimated average probability of success. The results below are for the Jets offensive line as this is the team that I have the most data for and thus the best estimates. I have also included a consistency score which is the standard deviation of a player's performance on a play-by-play basis (player predicted performance - average performance). Lower consistency scores mean the player is more consistent.

Player	Ferguson	Slauson	Mangold	Moore	Woody
Grade	2.2%	-1.8%	0.0%	2.0%	-1.0%
Consistency	0.07	0.09	0.15	0.07	0.07

The 2.2% grade for D'Brickashaw is the highest of any linemen in the data set and his consistency score of 0.07 is tied for the best consistency score in the data set. Overall, the interpretation here is that Ferguson (LT) and Moore (RG) are consistently outstanding, Mangold's performance has been about average though highly variable and Slauson and Woody have performed well below average for their positions.

As the data set continues to get bigger, these estimates will be more precise and the effects will be more clear, and more teams and players will be ranked. This analysis as is though suggests that Sanchez gets Blind Side pressure on 2.2% fewer plays than the average QB. I think Jay Cutler is probably jealous.

Week 4 Picks

2010-10-02T09:10:00.000-07:00

I am in full conference planning mode and the Blind Side Project is in full data gathering mode so I have no time for anything as frivolous as an explanation for each pick. So since the model hit 68% of the games last week (including the Giants and Jets which appeared in the Wall Street Journal) and a confidence weighted 62% against the spread, I'll let the picks speak for themselves.

San Diego (30) vs. Arizona (18) Confidence: 74%

Pittsburgh (26) vs. Baltimore (10) Confidence: 80%

New Orleans (26) vs. Carolina (14) Confidence: 82%

Cleveland (13) vs. Cincinnati (18) Confidence: 81%

Tennessee (23) vs. Denver (21) Confidence: 84%

Green Bay (29) vs. Detroit (18) Confidence: 85%

Oakland (21) vs. Houston (27) Confidence: 63%

Jacksonville (20) vs. Indianapolis (24) Confidence: 56%

Miami (23) vs. New England (21) Confidence: 63%

St. Louis (18) vs. Seattle (21) Confidence: 57%

Atlanta (18) vs. San Francisco (21) Confidence: 56%

Philadelphia (28) vs. Washington (27) Confidence: 81%

The Blindeside Project: Some Jets Questions

2010-09-24T10:59:00.000-07:00

As the fresh data pours in, I wanted to start asking some questions about both the Jets offense and defense (we'll focus on the Giants in another posting). This posting will have one question for each side of the ball, and most of these results will be fairly raw and rather stream of conciousness style thinking so any comments or thoughts on areas that seem the most fruitful are welcome.

The first question is about the inherent aggressiveness of the defense. They get pressure quicker (2.3 seconds vs 2.45 average) than other teams. Do they get this pressure faster because they have better rushers or because they send more rushers?

The answer seems to be both. Jets are nearly twice as likely to send a 5th or 6th or even 7th rusher than any other team currently in the data set (see the graphs below for a comparison between the Jets distribution of # of player rushing the QB vs the entire data set).

This would suggest that the shorter time is due to higher volume of bodies flying at the opposing team's offensive line, but that is not the entire story. With a basic regression of time in the pocket on number of rushers, while controlling for the QB's dropback and whether they were in shotgun formation or not, I find that the Jets get more kick from their additional rushers than other defenses. For example, when the Jets go from four to five rushers, their opponents average time in the pocket drops by 0.22 seconds, while the average team only sees a drop in time in the pocket of abou 0.14 seconds when they send an additional rusher. This suggests that it is not just volume of rushers (though that certainly helps) but also what those rushers are doing. The skill of the rushers and how they are sent are effects that we can't unpack just yet, but they seem to have a big effect for the Jets defense.

On the other side of the ball, Sanchez seems to be getting some credit for his cool demeanor, and I wonder if he really is cool under pressure, or is something else going on.

Looking at the data, the Jets face the fewest rushers on average of any team in the data set and have the longest average pressure free time in the pocket. They also have teh shortest average length of a throw (7 yds vs 9+ for everyone else. So, while none of this is truly conclusive, Sanchez gets more time to complete shorter throws, so his completion percentage better look good.

NFL Week 3 Picks

2010-09-23T23:54:00.000-07:00

Last week I had the Jets over the Patriots and nearly the 49ers over the Saints (both right against the spread). Since I'd rather not dwell too much the Vikes or Cowboys picks, lets get right into week 3. As always, home team is listed first, projected winner in bold and confidence level is the probability that the pick is correct.

New Orleans (26) vs. Atlanta (14) Confidence: 80%
Atlanta has not been very good on offense, particularly on first down and they have been aweful stopping the run on first down. The Saints on the other hand have been very good running the ball on first and generally playing very good defense. The match-up signals bad news for the Falcons this week.

New England (25) vs. Buffalo (15) Confidence: 77%
The Patriots bounce back from not being able to stop the efficiency of Mark Sanchez by shutting down the UFL level Bills offense. The BIlls have had some success defending the pass, but that will not be enough to pull out a win on the road.

Carolina (13 ) vs. Cincinnati (15) Confidence: 92%
Highest confidence on the board despite the closeness of the score. Even the Bengals shouldn't be able to blow this as they face the 30th ranked offense in the league that is paired with one of the worst pass defenses in the league.

Baltimore (21) vs. Cleveland (20) Confidence: 78%
The model finally takes its talents and gets out of Cleveland. The Ravens good if not great offense goes to battle against a sub par defense on all fronts and as Cleveland has one of the worst passing attacks in the league, the Ravens should be winners at home.

Houston (26) vs Dallas (18) Confidence: 76%
Finally going against Dallas and strongly against them. The Texans have the 2nd best passing attack in the league and they are not afraid to use it, while the Cowboys have struggled mightily defending the pass (particularly on 3rd down). The Cowboys offense will continue to look good though as they will put up some yards against a porous pass defense, but it won't be enough.

Minnesota (28) vs. Detroit (20) Confidence: 88%
Detroit is the hot pick to start putting up some wins due to their gaudy sack numbers. They won't pull it off against the Vikes though as Favre is still leading a good offense and the Lions 27th ranked offense will not be able to break through a very strong Minnesota defense.

Kansas City (17) vs. San Francisco (18) Confidence: 62%
Chicago (23) vs. Green Bay (15) Confidence: 59%
Denver (24) vs. Indianapolis (16) Confidence: 77%
Arizona (26) vs. Oakland (20) Confidence: 97%

Jacksonville (21) vs. Philadelphia (24) Confidence: 61%

Tampa Bay (18) vs. Pittsburgh (19) Confidence: 73%

Seattle (21) vs. San Diego (20) Confidence: 55%
St. Louis (19) vs. Washington (30) Confidence: 71%

Note: Giants and Jets predictions will appear in the New York edition of the Wall Street Journal.

Time in the Pocket and the Effects on a QB

2010-09-22T11:52:00.000-07:00

The stake in the ground of the Blind Side Project is that time in the pocket, which is increased by good offensive linemen, has a significant effect on a QBs ability to complete passes. The natural first pass at trying to establish this would be to look for a direct correlation between completion percentage and time in the pocket while adjusting for the distance of the throw. While a small positive correlation can be found, it does not seem to be of the magnitude one might expect, or at least I would not expect. This could be a function of the size of the data set, and that more data will demonstrate a stronger correlation, but I think that the complexities of the game of football, are more the culprit. There are a host of factors that could effect the relationship between time in the pocket (even adjusted for distance of the throw) and completion percentage. One area that is clear however, is that when a QB is under pressure, their completion percentage (adjusted for the distance of the throw) drops. In our 2007 data, when Peyton Manning is under pressure, his adjusted completion percentage drops from 86% to 46% and when Tom Brady is under pressure, his adjusted completion percentage drops from 71% to 39%. There is also a very direct connection between how long a QB holds the ball and the probability that they are under pressure. The graph below demonstrates the "survival function" or the estimated probability that a QB will not be under pressure, for any given length of time in the pocket.

This graph looks about like we could expect it to. As time in the pocket increases, the probability that the QB is not under pressure drops and the probability drops the fastest between 2 and 3 seconds. This suggests that the average offensive line has a fairly high probability of keeping their QB free of harassing defensive players for about two seconds, but there is only about a 40% chance that the QB will have 3 seconds without pressure.

This helps provide a baseline for the overall quality of an offensive line. Lines (or individual linemen) that can help increase this probability, can help increase the overall efficiency of the offense. Additionally, linemen that can increase the No Pressure probability, also provide the playcallers with increased flexibility in areas such as depth of routes and number of receivers running routes (as opposed to staying in to help block).

Northern California Symposium on Statistics and Operations Research in Sports

2010-09-20T13:30:00.000-07:00

Just a quick reminder that the Northern California Symposium on Statistics and Operations Research in Sports will be held on October 16th at Menlo College. Featured speakers this year are Roland Beech of the Dallas Mavericks and Sig Mejdal of the St. Louis Cardinals.

The Blindside Project

2010-09-20T11:57:00.000-07:00

I have been fascinated with the offensive line since the heyday of the Hogs in Washington DC. During the 1991-92 season, the Redskins offensive line (Schlereth, McKenzie, Jacoby, Bostic and Lachey) turned Mark Rypien into not only a Super Bowl winning QB, but into one of the most efficient QBs of all time. That year Rypien amassed over 3500 yards and 28 TDs with only 11 interceptions. He was able to pull that off because he was sacked only 7 times the entire season. To put that into context, the Bills, Texans, Steelers, Panthers and Eagles have all allowed 7 or more sacks after just two games, and every team is on pace to allow at least 8 sacks this season. And to further explain the excellence the 'skins line exhibited that season, remember that they played both the Jerome Brown, Reggie White and Clyde Simmons's Eagles and Lawrence Taylor's Giants twice that season.

The issue with offensive linemen is that beyond the proverbial "I know it when I see it" there is very little known about how to quantify their performance and understand their true value to a team. Clearly the 'skins would not have won the the Super Bowl, or even many games with Rypien at QB without the exceptional line play (the next season with a banged up offensive line Rypien tossed 13 TDs and 17 interceptions while being sacked a more normal 23 times). But what contribution did each of these players make individually to the team?

I have begun to tackle this question before and would like to dig deeper into the data that I have now, and expand this data set, so that we can be as confident in the results as possible. Below I will begin to further probe the data set that exists to try and understand the right directions to take, but the data collection provides the stumbling block.

One of the most important jobs of the offensive line is to provide the QB with a safe pocket in which he has plenty of time to make the right decisions. Clearly the longer a QB has to watch his receivers develop their routes, without being harassed by 350 lbs defensive ends trying to take his head off, the better his decisions are going to be, and more likely it is that the receivers will get open. I based that not on any data analysis (though the analysis does support it) but rather the knowledge that I make better decisions when I do not have 350lbs of DE rushing towards me.

This implies that to understand the value of each offensive linemen, we need to know how well they provide this safe haven for their QB. Which is exactly what I set out to do in my original stab at this problem. At the time I (with the help of my then intern Jesse Weinstein-Gould, put a stop watch on every pass play for a small group of teams, to get an idea of how long linemen could hold their blocks - how long did QBs have in their safe pocket. While this is a very time consuming project, in certainly yielded some clear results on how variability in the quality of an offensive line could effect the completion percentage of a teams' passing attack. This was done for 5 teams for the first four weeks of the 2007-08 season.

Now, looking to expand the data set, I will again (with the help my current intern Keith Goldner) put a stopwatch on every pass play (both offense and defense this time) for two teams over the course of the entire season. I will be charting the Jets and Giants all year and gathering all of the data that was originally tracked, plus some additional variables on formation and play type that will help further inform the analysis. Some of that work will appear here and some will be over at the Wall Street Journal but the goal of all of it is to enter into a massive study of the effect of individual linemen on a teams' passing attack.

As I have discussed here previously the dominant thought in the NFL is that LTs are by far the most valuable of offensive linemen (as implied by their significantly higher salaries) and therefore must have the biggest impact on a team. According to the Blind Side Hypothesis, this takes the form mostly of insurance. The LT protects the blind side of the QB (assuming he is right handed) and so keeps the QB safe (or not) from what he cannot see. As the story goes, Joe Theisman had not chance to avoid Lawrence Taylor because he could not see him coming. Thus, the data should indicate that when a LT fails to hold his assigned block, it results in a sack more often than when any other offensive linemen does.

Once we control for actual time in the pocket, this is exactly what the data shows. When a Left Tackle fails, the odds that a QB is sacked increases by over 20%. In contrast, when a Right Guard fails, the odds that a QB is sacked increase by only 9.2%.

	Probability of a Sack
Left Tackle	21.6%
Left Guard	16.0%
Center	4.4%
Right Guard	9.2%
Right Tackle	14.6%
Other	13.7%

These results suggest that there is a significant difference between failures and sack totals, but does that translate into an effect on completion percentage? There the data appears to suggest that no, failure is failure when it comes to completing passes, so the higher salaries may be attributable solely to the insurance aspect: we don't want our QB getting hit too much. The differences in salary for LTs relative to other linemen and other players on the team though seem too great to be explained just by insurance. Many teams, for example, pay their left tackles more than their QB, you would not pay $1mil to insure a $500,000 asset - so it seems like their needs to be some team performance element to the salary differences.

These are just preliminary results though, as the data set currently includes only 478 plays from five different teams. As the data set grows and number of teams covered expands, the data set will be more robust which should lead to clearer results. Once we can fully evaluate the contribution of each linemen, then hopefully we will be able to at least better understand the salary structure for linemen in the NFL.

Singletary was right

2010-09-19T08:01:00.000-07:00

Sure I love a good coach meltdown as much as the next guy, but the thing about Mike Singletary's nuanced answers regarding his team's upcoming match-up with the Saints is that he was right.

Asked how his offense will try to move the ball against the Saints' defense, coach Singletary replied "We won't try to move the ball against the Saints' defense, we will move the ball against the Saints defense and we will score points. Next question." His skill in directly answering the question with confidence and annoyance, plus sticking the "Next Question" dismount earns him a perfect 10. Allow me to just read between the lines a little bit for you though so you can see how those points will get scored. Once you use more than just last weeks results (win for Saints, embarassing loss for the 49ers) to make projections, you can see some interesting matchups.

The 49ers 1st down passing attack was not overly impressive, except that they really seem committed to throwing on first down. They get 5.3 yards per pass play on first down which is a full yard better than what they get when they run the ball. As the Saints run defense is excellent (allowing less than 4 yards a carry) and their pass defense is suspect (allowing 6.3 yards per attempt), the 49ers commitment to throwing on first down is a major strength. This is a good math-up for Alex Smith to demonstrate why the 49ers believe in him.
Particularly if the 49ers can employ the same tactics that the Ravens did last week against the very aggressive Jets defense: lots of shotgun with quick passes, limiting the effects that their incredible pass rush could have on their young QB.

Asked how his defense will try to stop the high power Saints' offense, Singletary again stuck the dismount: "We will not try to stop the Saints' offense. We will stop the Saints' offense. Next Question!"

The 49ers will stop the Saints' offense because the Saints actually like to run on first down which is the great strength of the 49ers defense. They have a good pass defense, but the best rush defense in the league on first down (again, this is not just based on last weeks data). If New Orleans utilizes a similar game plan to last week, they will find themselves in 3rd and long on a regular basis against a defense that will be able to slow them down enough to get the win.

Next question.

NFL Week 2 Picks

2010-09-15T20:26:00.000-07:00

Week one was an interesting one for the model. Of all the picks made, it got 53% correct (against the spread as well). Once the confidence in the pick is factored in (estimated accuracy of the pick) the accuracy shot up to 72% overall and 67% against the spread. This included picking the Steelers and the Texans. I did pick the Cowboys and they blew it by holding on the last play of the game, but that is what I get for picking the Cowboys.

This week we final have some current data to feed into the model so we are looking for slightly better overall performance. Here goes nothing. Home teams are listed first, each team's projected score is in parentheses and the confidence level reflects the odds the team projected to win will actually win the game:

San Francisco (20) vs. New Orleans (19) Confidence: 51%
The 49ers ran into so....technical difficulties last week, but we assume that Singletary and Co. will be able to get the play calling under control and get Alex Smith in position to have the break out year I have projected him to have. Meanwhile, New Orleans is not running the ball well nor (strangely enough) passing enough.

Washington (17) vs. Houston (21) Confidence: 52%
The Texans look to follow up their big upset of the Colts with a win on the road over the incredibly lucky Redskins. Houston averaged over 7 yards per play on first down against a good Colts defense, hard to imagine what they will be able to do to a 'Skins defense that gave up nearly 6 yards a play on first against a good but not that good Cowboy offense.

Carolina (20) vs. Tampa Bay (21) Confidence: 58%
These were two of the worst third down defenses last week, both giving up close to 10 yards per pass play on 3rd and long. Not sure that either team has the firepower to take advantage of the others weakness though so Tampa should be just good enough on offense to squeak out a win.

Cincinnati (21) vs. Baltimore (16) Confidence: 59%
The Ravens must be riding high after their big Monday night upset of the suddenly cowering Jets, but it will not be enough to beat a division rival on the road. The Bengals are passing well and often which should more than over come a leaky first down defense to score them the win at home.

Cleveland (24) vs. Kansas City (21) Confidence: 61.3%
The Chiefs are getting a ton of buzz and my good friends over at Football Outsiders are feeling good about their pick of the Chiefs winning the division. I however see there 1.4 yards per pass attempt on first down against the Chargers and believe that everything is not quite up to date in KC. Cleveland's strong defense will let them win at home, regardless of their starting QB.

Denver (22) vs. Seattle (19) Confidence: 65.8%
Tebow Fever, catch it.

Green Bay (23) vs. Buffalo (15) Confidence: 64%
Buffalo looks to have one of the worst offenses in the league, averaging approximately 3 yards per play on first down. Their defense is stout, but not enough to contain Rodgers and company (even without Grant) in Lambeau.

Tennessee (22) vs. Pittsburgh (19) Confidence: 70.4%
I had the Steelers win pegged last week, and feel equally confident about a loss for them this week. The Titan's D is looking dominant in all phases and defenses are so scared of Chris Johnson, Vince Young is becoming an efficient option.

Detroit (9) vs. Philadelphia (10) Confidence: 71.2%
The Vick experiment continues in the backup QB bowl. It looks to be close, but since the Detroit Defense gave up 11.3 yards per 1st down pass attempt and 10 yards per 2nd down pass attempt, even Vick should be able to move the chains.

Oakland (12) vs. St. Louis (17) Confidence: 71.4%
St Louis is apparently not afraid to let their rookie QB throw the ball (60% of first down plays), and the Oakland D seems amenable to allowing him to actually complete a few this week.

Atlanta (12) vs. Arizona (21) Confidence: 75.8%
The Cards take their balanced and efficient offense on the road against an Atlanta team that appears to have weakened on the defensive end and have little to no rushing attack on offense. The Cards should have little problem taking this one on the road.

Minnesota (23) vs Miami (14) Confidence: 80%
Minnesota bounces back at home after being beat up by the Saints in the opener. They have a dominant run defense and Miami has almost no aerial attack. Peterson may have trouble finding some holes in a defense that looks to be just as good against the run as the Vikes, but at home in their very own dome, old man Favre should be able to find a few TDs to keep the Purple alive.

San Diego (28) vs. Jacksonville (18) Confidence: 85.7%
My one and only caveat to this pick is that the Chargers have to throw the ball more on first down. They are a very good passing team, yet Norv chose to throw on only 40% of first down plays against the Chiefs. That is 21st in the league, behind such stellar passing attacks as the Redskins, Rams and 49ers. If they throw, they'll win going away, otherwise...

Dallas (25) vs. Chicago (16) Confidence: 86.3%
They deeply disappointed me last week, but they also outplayed the Redskins in virtually every aspect of the game. As Chicago was on the other side of lucky last week, I see Dallas bouncing back and actually letting their very efficient offense score some points.

NFL Week 1

2010-09-08T21:46:00.000-07:00

Football can be viewed as a game of situations and match-ups, so my statistical model that predicts game outcomes analyzes situations and match-ups. Additionally, on a very macro level, the model analyzes how well coaches utilize their tools. For example, it does not matter how well a team passes on first down and how poorly their opponent defends the pass on first down, if the coach does not call many first down passing plays (I'm looking at you Norv).

The model was put into action during the playoffs last year and picked 9 of 11 games accurately, including the Jets upset win over the Chargers and the Saints upset win over the Colts (the model was also 67% accurate against the spread...not that I'm interested in that sort of thing).

The additional complication for the model at this point in the season though is that teams have changed from last year, so we do not have a clear set of estimates for exactly how well a team will be able to pass on 3rd and long. The starting point is obviously last seasons data, but I then go through a projection process that incorporates how much turnover each teams has had on each side of the ball, while also assuming that to some degree, each team's performance will move closer to the mean.

So without further ado, below are my picks for week one with the exception of the Giants/Panthers and Jets/Ravens. For those picks you'll have to check the Wall Street Journal. Home teams are listed first, each team's projected score is in parentheses and the confidence level reflects the odds the team projected to win will actually win the game:

Tampa Bay (23) vs. Cleveland (21)   Confidence 50.0%
When the 24th and 32nd projected offenses get together on the same field, the only real losers are the fans. The Bucs should be able to squeak this one out at home, but there is certainly no guarantee.

New England (24) vs. Cincinnati (21) Confidence 51.0%
The Pats face a tough home opener against a retooled Bengals team that could pull off the upset based on their excellent pass defense. If the Pats can throw effectively though their average running game won't be an issue and they should have just enough to pull it off.

Washington (18) vs. Dallas (21)    Confidence 51.7%
The Cowboys will storm into DC and likely steal one from their hated rival on the road. The Cowboys high powered offense will be too much for the 'skins defense and a hobbled McNabb (or worse, a healthy Grossman) won't be able to take advantage of the spotty Cowboys pass defense.

Seattle (10) vs. San Francisco (14) Confidence 53.4%
My hometown 49ers will run up the coast to take the season opener from division rival Seattle. They will bring a projected 5th ranked defense along with a much improved Alex Smith against a Seattle team that just received a vote of no confidence from the very experienced Alex Gibbs.

Jacksonville (14) vs. Denver (21) Confidence 53.6%
If Tebow doesn't play will anyone watch? Ok, they'll watch in case he does play I guess. Otherwise you have two mediocre teams playing outside the division…Someone has to win and odds are it is the Broncos by 7 due to the Jags inept first down pass defense.

New Orleans (28) vs. Minnesota (18) Confidence 61.9%
Full comments are available below but what it comes down to is Brees is a better QB than Favre ever was and taking away Bush's Heisman takes nothing away from a dominant offense and growing defense.

Buffalo (22) vs. Miami (21) Confidence 64.2%
With the Tuna now out in Miami, maybe they will be allowed to throw the ball a little bit, but then with their combination of RBs and QBs, maybe they are the exception to the rule that teams should pass more. Only Oakland and Miami are projected to actually gain more yards per play when they run than when they pass.

Philadelphia (21) vs. Green Bay (18) Confidence 66.0%
Rogers begins his quest to lead my fantasy league in points again by disappointing many against a top 10 defense in Philadelphia. The anemic rushing attack and inexperienced passing attack of the Eagles will still prove too much for the visiting Packers and their 15th ranked pass defense.

Kansas City (14) vs. San Diego (21) Confidence 67.8%
To quote a wise friend "How can Norv Turner hold the same position that Sid Gillman and Don Coryell did?" Norv's conservative ways aside, this Chargers team looks to have NO trouble on the road with the truly abysmal Chiefs and their 30th ranked offense and 22nd ranked defense.

Chicago (24) vs. Detroit (21)         68.3%
Stafford's sophomore season begins on the road against a top 10 defense. Not a real recipe for success when you will be relying on Nate Burleson to be easing some pressure off of Calvin Johnson. The Bears may not be able to run the ball against anyone, but with Martz calling the plays and one of the worst pass defenses in the league coming to town, they may not even try.

Houston (28) vs. Indianapolis (17) Confidence 73.0%
The Colts have not lost to the Texans since 2006, but this is the year of Schaub. Houston will have the best passing first down offense this year and Kubiak will not be shy about utilizing it. Expect the Texans to pass early and often against a good but not dominant Colts defense. While Schaub is still on the upside of his Career, Manning will begin (if just slightly) to show his age and any decline in Peyton's performance will prove devastating for the Colts.

St. Louis (15) vs. Arizona (18)      Confidence 75.8%
I feel pretty confident picking the Cardinals to win on the road, but I'm not very interested in it.

Tennessee (25) vs. Oakland (21) Confidence 80.3%
The Titans maybe adding Albert Haynesworth to a defense that is projected to be the 3rd best 3rd down defense this season. Couple that with the fact that they are playing the Raiders at home, and this is a pretty straightforward pick.

Pittsburgh (26) vs. Atlanta (14) Confidence 89.6%
Complete analysis of the Steelers and Dennis Dixon is below, but the Steelers could spot the Falcons and their "struggling" defense a touchdown and still win going away.

The Dixon Effect - What the Steelers Lose Without Big Ben

2010-09-06T22:34:00.000-07:00

I have the Steelers projected to win a boat load of games this year, 12 to be exact, which is the highest
projected total on my list. It has been suggested that I am nuts to have the rates so highly with Big Ben sitting out the first four games of the season and Byron Leftwich out with an injury. This has left Dennis Dixon as the starter and Charlie Batch as the primary alternative (just a note, they probably could have added Leinart to the mix here, yet by all indications no serious thought was given to this option).

So what exactly will the Dixon Effect be and how significant will it be on the Steelers' prospecting for winning in the first four weeks. The real problem with trying to make this determination is that Dixon has essentially no NFL experience, so nothing to concretely gauge what we should expect from him. With no real data to work from, I went to the facts that we do know: he was a 5th round pick, he is 25, he will be starting the season and likely playing 4 games and going back to the bench and he has almost no prior experience. This information actually describes a reasonable number of NFl QBs that can help put some parameters on Dixon's likely performance. Since 1980 20 QBs have fit this rough description and while some will give the Steeler faithful hope (Marc Bulger, ok maybe Bulger is the only serious hope on the list) most are in the Koy Detmer/Rusty Hilger model.

Player	Year	Cmp	Att	Cmp%
Eric Hipple	1982	36	86	0.419
Paul McDonald	1983	32	68	0.471
Mark Herrmann	1984	29	56	0.518
Steve Bono	1987	34	74	0.459
Steve Pelluer	1987	55	101	0.545
Rusty Hilger	1987	55	106	0.519
Kevin Sweeney	1988	33	78	0.423
Mark Vlasic	1988	25	52	0.481
Stan Humphries	1990	91	156	0.583
Scott Zolak	1992	52	100	0.52
Scott Mitchell	1993	133	233	0.571
Cary Conklin	1993	46	87	0.529
Rob Johnson	1998	67	107	0.626
Koy Detmer	1998	97	181	0.536
Moses Moreno	2000	27	53	0.509
Marc Bulger	2002	138	214	0.645
A.J. Feeley	2002	86	154	0.558
Jesse Palmer	2003	60	116	0.517
Cody Pickett	2005	14	35	0.4
Dan Orlovsky	2008	143	255	0.561

What this means for the Steelers is that they can reasonable expect Dixon to complete about 54% of his passes (Big Ben has a career average of 63%), turn 3.6% of his attempts into TDs (5.3% for Big Ben) and be intercepted on 3.9% of his attempts (3.4% for Big Ben) with an average of 6.3 yards per attempt (8 Y/Att for Big Ben).

The yards per attempt difference is large and is likely directly connected to the difference in TD%. As the int% is roughly equal, the biggest impact then will be on the offense's ability to score points. Assuming the Steelers throw on 35 to 40 plays against their week one opponent (Falcons), they will probably score about 4 fewer points due to Big Ben's suspension. As I have the Steelers winning that game rather decisively (26 to 14 with nearly 90% confidence) a drop in offensive output of even 5 points still has them winning by a touchdown, even if Dixon turns out to be Moses Moreno or Cody Pickett.

Vikings/Saints Preview

2010-09-06T10:59:00.000-07:00

The Saints open at home on Thursday night, kicking off their quest to repeat as NFL champions against the Favre-Vikings. The only downside to the NFL season starting is that with Favre in the season opener we will certainly hear more about the legendary Favre and how he just loves the game etc. We may not hear from the broadcasters that Brees's average yardage total for the four seasons he has been in New Orleans (4575) is 161 yards higher than Favre's best season (4413, in 1995, whne he was 26 and Brees was 16).

We can get past the Favre-fest though and enjoy the opener and watch knowing that there is a 62% chance the Saints win with a likely final score of 28 to 18. Yes the vaunted Saints offense will not be slowed at the start of the season as Brees an Co put up big numbers against the over-hyped Vikings defense. The Vikes have a great first down defense, but they struggle on 2nd and 3rd downs while the Saints Offense (projected 2nd overall coming into the season) is dominant in all offensive situations. They run well and pass even better. Consider that they have the 10th best projected yards per rush on 1st and 10, and 6th best projected yards per pass. The rankings get better as they move to second and third down scenarios.

On the other side of the ball, the Vikings do not maximize their assets very well as they have the 10th best projected yards per pass on first and 10, but only throw it 44% of the time. Yes Peterson is a great back, but on first and 10 they are projected to be 13th in the league in rushing. In yardage terms, the Vikings get 2.5 more yards per play when they pass on 1st and 10, but they still choose to run more often than not. By giving up those yards on 1st down, the Vikes will be playing into the strength of the Saints defense which is their 3rd down packages. They move from an average at best defense on first and second down to a top 10 defense on 3rd down.

So while the great Favrernator may exhibit is legendary love for the game Thursday night, it will be Brees and the defending champion Saints loving the victory.

Saints: 28, Vikings 18, Confidence level: 62%

The NFL is Making Less Money????

2010-09-02T23:44:00.000-07:00

I was browsing my favorite NFL propaganda website today when I saw the devastating news: USA Today Reports NFL Attendance for Third Consecutive Season. This was incredible because it suggests that the NFL may actually function like other businesses and see lower revenues during a very weak economy. Or worse, maybe the NFL is on shaky economic ground all by itself. If that were the case, then NFL owners would have a real case in the ongoing negotiations for a new collective bargaining agreement that they should be paying players less and have an "enhanced" season.

I needed to learn more about this dire business climate that the NFL owners are facing, so I clicked through to the actual USA Today story. This is when it starts to get interesting.

Reading the story sheds some light on what it is the USA Today is actually reporting. The headline on the NFLlabor site hints that the USA Today had done some economic analysis or at least interviewed some independent analysts. Not so. The USA Today article was an interview with NFL executive Eric Grubman. So the only source of the expected decline in ticket sales that the NFL is screaming about on its' propaganda site, is the NFL itself.

That would be just a case of a sloppy pr stunt, except that the USA Today did do some actual reporting that helps put that decline into context. During the 2009-10 season, ticket sales were apparently down 2.4%, double what was projected for the upcoming season. The NFL also saw a total increase in ticket prices of 3.9%. So in the face of the worst economy the NFL has ever seen, they raised ticket prices by 3.9% which translates into a total revenue increase of 1.4%.

So I find myself totally agreeing with Michael Wilbon as I watch PTI "The NFL is King"..."The NFL has nothing to worry about" Except perhaps that their attempts to cry poor entering the CBA negotiations will likely fall flat.

Season Predictions

2010-09-02T13:03:00.000-07:00

With a week to go before the season begins, for your commentary, I present my the win totals from my season simulator for the NFL 2010-11 season. I have also included each team's win total from last season, along with their expected change in wins.

Team	2010-11 Wins	2009-10 Wins	Change
Pittsburgh	12	9	4
NY Jets	12	9	3
San Diego	12	13	-1
Arizona	11	10	1
New England	11	10	1
New Orleans	11	13	-2
Philadelphia	11	11	0
Baltimore	10	9	1
Cincinnati	10	10	0
Green Bay	10	11	-1
Houston	10	9	1
Tennessee	10	8	2
Chicago	9	7	2
Dallas	9	11	-2
Denver	9	8	1
Indianapolis	9	14	-5
Minnesota	9	12	-3
San Francisco	9	8	1
Washington	9	4	5
Carolina	8	8	0
Seattle	7	5	3
Buffalo	7	6	1
Atlanta	6	9	-3
Jacksonville	6	7	-1
NY Giants	5	8	-2
Miami	5	7	-2
Detroit	4	2	2
Kansas City	4	4	0
Oakland	4	5	-1
Tampa Bay	4	3	1
St. Louis	2	1	1
Cleveland	1	5	-4

Not much of a surprise to see Cleveland and St Louis at the bottom (though winning four fewer games than last year for the Browns is pretty startling). Beyond the Colts dropping to 9 wins, the biggest surprise to me was the Redskins adding 5 wins. They have upgraded their offensive line (it would actually be difficult to not have improved that unit) and made a marginal improvement at QB (while McNabb is healthy...I think I would prefer Campbell over 2nd stringer Rex Grossman) but 9 wins seems awfully ambitious for that group.

A funny thing happened on the way back to the Super Bowl...

2010-09-01T23:26:00.000-07:00

I have been fine tuning my projections for actual win totals for the upcoming NFL season (ONLY 7 DAYS TO GO!!). To produce the win projections, I input the entire season into a simulator, and project the score of every game along with the probability that the winner is picked accurately. These projections are based on the model I used with the Wall Street Journal during the playoffs last year and explained roughly in previous posts. Using those projections I simulate the entire season and take average win totals.

The results are always interesting and occasionally produce a result that draws my attention. I like to compare the projected win totals with last season's win totals for each team to see who the system says the big risers and fallers will be. After running the simulation, I found that the team with the largest drop in wins from last season to this season is the Indianapolis Colts. That's right, Peyton and Co. are projected to go from 14 wins last season to only 9 this season (I say only 9...I think Rams/Browns/Raiders' fans would just about kill for 9 wins).

The Colts being the owner of the largest projected drop in wins is cause for further investigation. What causes the Colts to go from 14 wins and clinching home field advantage throughout the playoffs to 9 wins and maybe not making the playoffs at all?

Part of the answer of course is Peyton himself. Peyton will continue to be one of the best QBs in the league, but he will be 34 during the 2010-11 season and is projected to have his first sub 4000 yard season since he was 29. A 700 yard drop in total passing yards is fairly significant, particularly since his interception total has been trending up since he was 30, but perhaps not enough to explain a 5 win drop.

The second part of the explanation has to do with schedule. Last season, the Colts' opponents had an average win total of 7.5. The Colts' are faced with a much tougher schedule in the upcoming season as their opponents' have an average win total of 8.5. One extra game in average win total may not seem like a big deal, but consider that last season, only 37.5% of the Colts' opponents had more than 8 wins while 68.75% of their opponents in the upcoming season have projected win totals of more than 8 games.

Peyton's decline along with a significantly tougher schedule make the Colts a prime candidate for a post-Super Bowl slump to only 9 wins (again, sorry Rams/Browns/Raiders' fans for the "only"). A repeat Super Bowl appearance seems fairly unlikely as they are not lock to even make the playoffs.

Complete QB Rankings

2010-08-30T16:10:00.000-07:00

I powered through the rest of the QBs and below you have my projected yardage, interception and TD projections for each QB in the NFL this season, assuming they play a full 16 games. We know that Big Ben is not going to get a full 16, Donovan will likely miss at least 4 games in DC, and some of these guys will miss games dues to either injury or just straightforward benching (I'm looking at you Mr Edwards and Mr Hasslebeck)

Of note, the projections have Orton continuing to accumulate high yardage totals (better than Cutler) and Favre showing his age by slipping to 22nd in the league in yardage. RYan, Cutler and Flacco all look to have big yardage numbers, but Cutler will continue to struggle with interceptions which could land him on the bench before the season is out. The biggest surprise for many (though not those that read this blog regularly) would be Alex Smith in a break out season at 13th in the league in yardage while throwing 27 TDs and only 11 pics.

All of the projections take into account the player's age, experience and prior performance except for Bradford. With no past NFL experience, I was stuck projecting Bradford based on all other first round rookie's who started a significant number of games. This procedure put him near the bottom in yardage and with more Int's than TDs (only 29% of rookie QBs throw more TDs than pics).

Player	Team	Yds	Ints	TDs
Matt Schaub	Texans	4920	14	31
Drew Brees	Saints	4820	15	36
Tony Romo	Cowboys	4785	11	27
Aaron Rodgers	Packers	4510	11	32
Philip Rivers	Chargers	4364	10	29
Ben Roethlisberger	Steelers	4321	14	24
Tom Brady	Patriots	4200	26	14
Kyle Orton	Broncos	3850	11	24
Peyton Manning	Colts	3836	16	30
Matt Ryan	Falcons	3600	16	27
Jay Cutler	Bears	3600	30	36
Joe Flacco	Ravens	3600	12	22
Alex Smith	49ers	3540	11	27
Eli Manning	Giants	3500	13	25
David Garrard	Jaguars	3450	15	14
Jason Campbell	Raiders	3315	14	16
Josh Freeman	Buccaneers	3315	18	12
Donovan McNabb	Redskins	3312	12	19
Jake Delhomme	Browns	3121	12	13
Matt Cassel	Chiefs	3103	17	18
Vince Young	Titans	3000	10	20
Bret Favre	Vikings	3000	18	25
Matt Moore	Panthers	2990	13	18
Chad Henne	Dolphins	2932	14	14
Carson Palmer	Bengals	2900	15	19
Mark Sanchez	Jets	2850	15	13
Kevin Kolb	Eagles	2828	18	19
Matt Hasslebeck	Seahawks	2675	18	23
Sam Bradford	Rams	2570	20	17
Trent Edwards	Bills	2520	14	13
Matthew Stafford	Lions	2480	17	20
Derek Anderson	Cardinals	2175	18	18

The Blindside Hypothesis: Part 2

2010-08-26T15:24:00.000-07:00

I previously started to investigate the Blindside hypothesis by wondering why the LT is the second highest player on an NFL team and why they are paid 64% more than their line mates. Additionally, I wonder whether the difference can really be attributed to some sort of insurance policy on the QBs health or whether there are other factors involved.

In order to continue this investigation, I went back to a data set that I collected during the first four games of the 2007 NFL season. For seven teams I, along with the assistance of Jesse Weinstein-Gould, charted four games, focused entirely on the offensive line. The question we were trying to answer at the time was "how can we begin to measure the impact an individual offensive linemen has on a team's passing game". In order to begin to answer this question we look at each pass play for those games, recorded the time to throw (time from the snap of the ball until the QB starts the throwing motion), whether each offensive linemen successfully held their blocks during that time, whether there was a sack, the distance the throw traveled in the air, yards the receiver ran after the catch and several situation variables such as down, distance to go etc.

The table below summarizes the average Time in the Pocket, Completion Percentage, Sack Rate, Distance of Throw, the percentage of plays in which the QB had no pressure, and the percentage of plays that each position successfully held their blocks (TEs, RBs and WRs were grouped together in Other). Before reviewing the data though, refresh your memory on the win/loss records of the teams involved in 2007: Bengals (7-9), Bills (7-9), Colts (13-3), Dolphins (1-15), Jets (4-12), Patriots (16-0) and Redskins (9-7). This of course was the magic perfect season for the Patriots and the record setting year for Tom Brady.

Column1	Bengals	Bills	Colts	Dolphins	Jets	Patriots	Redskins	Max	Min
Time in the Pocket	2.12	2.47	2.41	2.36	2.25	2.34	2.53	2.53	2.12
Complete	68.3%	54.5%	64.7%	59.0%	56.6%	73.5%	50.0%	73.5%	50.0%
Sack	3.2%	11.4%	2.4%	4.8%	11.7%	4.3%	7.9%	11.7%	2.4%
Distance of Throw	8.29	6.61	10.41	9.09	6.08	7.86	11.95	11.95	6.08
No Pressure	68.3%	54.5%	69.4%	79.0%	62.3%	84.1%	73.7%	84.1%	54.5%
Left Tackle	93.7%	88.6%	84.7%	96.2%	90.9%	91.3%	92.1%	96.2%	84.7%
Left Guard	92.1%	90.9%	96.5%	95.2%	90.9%	95.7%	97.4%	97.4%	90.9%
Center	96.8%	93.2%	100.0%	96.2%	92.2%	97.1%	94.7%	100.0%	92.2%
Right Guard	90.5%	90.9%	95.3%	98.1%	92.2%	98.6%	94.7%	98.6%	90.5%
Right Tackle	98.4%	95.5%	92.9%	98.1%	90.9%	97.1%	92.1%	98.4%	90.9%
Other	92.1%	81.8%	95.3%	93.3%	80.5%	91.3%	89.5%	95.3%	80.5%

The first of these stats that grab my attention is the No Pressure rate. In this limited sample, the overall average No Pressure rate was 70%, while Tom Brady had no pressure on 84.1% of his pass attempts. This appears to be exceptional line play including near perfection from the Right Guard position (only Colts' Center Jeff Saturday was better, and he actually was perfect in the sample). This exceptional line play assisted Brady in completing 73.5% of his passes in the sample, but only the 3rd best sack rate in the sample (the Bengals and Colts both had lower sack rates). This means that on roughly 25% of plays in which Brady faced pressure, he was sacked, while Manning was sacked on only 8% of the plays on which he was pressured. This suggests that the line as a whole may have had a significant impact on Brady's ability to put up the record setting numbers that he did that season.

But what about the Left Tackle? Patriots LT Matt Light was successful on only 91% of plays in the sample (Redskins, Dolphins and Bengals all had superior play from the LT), so can one linemen have that much more of an impact? In my next post I will start to explore this with the play by play charted data, to try and determine if in fact the LT has a bigger impact than the rest of the line.

The Blindside Hypothesis: Part 1

2010-08-26T00:45:00.000-07:00

Michael Lewis' wonderful book (and I don't say that just because he footnoted me in the book, pg 100 if anyone cares) about the life of Michael Oher notes that the left tackle is more often than not, the second highest paid player on an NFL team. What is particularly striking about that fact is that many of the NFL decision makers that Lewis interviewed did not know it and could not explain why that was the case. This thought is further advanced in the film made of the book when Sandra Bullock likens paying left tackles to buying insurance.

While linemen are vital to any offense, due largely to the fact that no publicly available statistics are available for them, they are often an after thought of the fan and apparently of some professionals. If professionals are not aware of the salary standing of left tackles relative to the rest of the team, they probably not well informed on the relative salaries of the rest of the line, and why large disparities exist between line mates. In order to more closely examine linemen pay, I combined the salary data from USA Today with the games started data from PFR. This allowed me to calculated the average salary and cap number of starting linemen by position.

	Salary	Cap Space
LT	$ 3,575,265	$ 3,923,691
LG	$ 2,431,490	$ 2,868,801
C	$ 2,862,943	$ 2,694,972
RG	$ 1,395,010	$ 1,609,293
RT	$ 2,824,596	$ 2,176,198

The data on linemen salary show that, on average, left tackles are paid 64% more than their line mates and teams use 80% more cap space on left tackles than on other linemen. As of last year, the closest position to LT's in pay was RT and they were paid on average, $700,000 less than LTs. Teams use more than $1mil more on cap space than the next closest position on the line (LGs). These very large disparities beg the question, is the left tackle worth it or alternatively, why are their line mats not worth more?

In order to begin to answer these questions, we need to begin by answering some more basic questions. For example, is there a link between money spent on offensive linemen, and offensive performance. Once we calculate the average salary and cap space that teams spent on their starting line last season (see table below) we can begin to answer this by looking at correlations between those salaries and the performance of the offense.

Team	Total Salary	Cap Value
49ers	$ 3,066,833	$ 3,632,999
Bears	$ 4,159,756	$ 3,419,117
Bengals	$ 1,439,993	$ 1,792,931
Bills	$ 1,836,777	$ 1,022,468
Broncos	$ 1,591,193	$ 1,938,308
Browns	$ 2,349,000	$ 3,444,333
Buccaneers	$ 1,449,848	$ 2,787,048
Cardinals	$ 2,581,334	$ 2,804,084
Chargers	$ 1,427,108	$ 1,707,240
Chiefs	$ 2,079,990	$ 1,943,164
Colts	$ 2,872,784	$ 2,606,066
Cowboys	$ 2,787,103	$ 3,043,357
Dolphins	$ 5,167,467	$ 3,157,894
Eagles	$ 2,623,413	$ 3,280,240
Falcons	$ 1,774,780	$ 1,363,322
Giants	$ 3,305,000	$ 5,037,584
Jaguars	$ 2,532,326	$ 1,970,106
Jets	$ 4,191,848	$ 5,007,598
Lions	$ 3,534,920	$ 2,963,993
Packers	$ 1,924,336	$ 2,378,670
Panthers	$ 3,043,496	$ 2,966,222
Patriots	$ 2,807,733	$ 3,150,845
Raiders	$ 2,164,503	$ 2,036,698
Rams	$ 883,730	$ 1,017,063
Ravens	$ 1,551,713	$ 1,443,419
Redskins	$ 2,132,200	$ 2,511,252
Saints	$ 2,480,208	$ 1,895,449
Seahawks	$ 1,807,744	$ 2,282,744
Steelers	$ 4,848,176	$ 2,781,576
Texans	$ 1,941,447	$ 1,777,930
Titans	$ 3,050,770	$ 3,925,770
Vikings	$ 2,070,066	$ 3,261,142

Looking for correlations between salary/cap and offensive performance suggests that there are three areas that in which increased spending on the line are related to better offensive performance: sacks, interceptions and yards per catch. The effect on sacks seems fairly straightforward and something we should expect if the market for linemen is at all related to the insurance part of the Blindside hypothesis. The relationship between linemen and salary could be related as better linemen (ie higher cost linemen) may provide the QB more time and thus are less likely to make poor throws, which is similar to the story we could tell about the relationship between linemen salary and yards per catch.

As I continue to explore the value of linemen in the next few posts, the goal is to rigorously test the Blindside hypothesis and determine if LT's salaries are particularly high due, in part, to concerns on insurance and whether that fully explains the large disparities between LT's and the rest of the line.

QBs of the NFC East

2010-08-21T07:33:00.000-07:00

Traveling east from the land of Favre, I find that it is time to project the QBs of the NFC East. Here we have three established QBs with significant track records that can be used to make data-based projections, then we have Kolb. Kolb presents an interesting problem as he has been in the league for three seasons, played in 12 games, but completed fewer than 80 passes.

Kolb's career completion percentage of 60% should give Eagles fans some hope as Donocan only broke 60% in one season (at that was when he hit 61.5% in 2007). The problem is, for the Eagle fan, that if you take the 60% number seriously, then you probably have to take his TD% (3.1%) and INT% (5.4%) seriously as well, and when you do that, you notice that he throws 74% more interceptions than touchdowns.

This is where, as an analyst, I have to remember two important things: 1) a small sample is a small sample is a small sample...BE CAREFUL, 2) Andy Reid has a lot more information about Kevin Kolb than I do. So with those to ideas clear in my head I present to you NFC East QB projections:

Player	Team	Yds	Ints	TDs
Donovan McNabb	Redskins	3312	12	19
Tony Romo	Cowboys	4785	11	27
Eli Manning	Giants	3500	13	25
Kevin Kolb	Eagles	2828	18	19

Notice that Kolb is projected to throw the most interceptions in this group and ties Donovan fro fewest TDs, but no where near as many as his career numbers would suggest. The data do suggest that Kolb will struggle, but will show some improvement as well. The big news in these projections is Romo, and not just for his gaudy yardage total. The data suggest that he will have two more TDs than Eli for tops in the division and two fewer interceptions than Eli again for the best mark in the division. Romo has been a very efficient QB never posting a yards per attempt below 7.7 (a mark Donovan has only beaten twice in his career and Eli only once) and had a ridiculously low interception rate last season (1.6%). While the interception rate will likely come up a little, he will still be the top QB in the division by good margin.

Analytic Football

Beyond 'Moneyball':

Beyond ‘Moneyball’: The rapidly evolving world of sports analytics, Part I

By Benjamin Alamar and Vijay Mehrotra

Can Alex Smith Save the 49ers?

Initial Poll Results

Future of Sports Analytics

Additcted to Math?

Sports Analytics: Should Fans Care?

Risk and Drafting a QB

Accepting an 18 Game Schedule

Blind Side Project Help

Business Intelligence on YouTube

The Undisputed Guide to FREEDARKO

Differentiating Sports Analytics

Impact of defensive pressure, distance on pass completion

Who's Your Analyst?

Rethinking the NFL

Analytics and Communication

Players: One Definition

Why Sports Analytics?

Advice for the Aspiring Sports Analytic Professional

Get Your Geek Out of the Box

NFLPA's Secret Leverage

Sports Analytics is Not a Strategy

Week 10 Picks

Week 9 Picks

How 18 Games Could Change the Game

Week 5 Picks

Northern California Symposium on Statistics and Operations Research in Sports

8am – Registration Opens, Continental Breakfast available

9am – Welcome

9:10am – “The intra-match home advantage in Australian Rules football”

9:45am - “Pitcher Accuracy through Catcher Spotting: Assessing Rater Reliability”

10:20am – “Optimal Dynamic Clustering Through Relegation and Promotion: How to Design a Competitive Sports League”

10:55am – “Dynamic Effort, Sustainability, Myopia, and 110% Effort”

11:30am – Featured Speaker: Sig Mejdal of the St. Louis Cardinals

12:15pm Lunch Provided and Poster Session

1:45pm – “Drafts versus Auctions in the Indian Premier League”

2:20pm – “Reconsideration of the best batting order in baseball: Is the order to maximize the expected number of runs really the best?”

2:55pm –“Uncovering Football’s Best Goalscorers in the English Premier League, La Liga, Serie A, Bundesliga, Eredivisie, and Ligue 1 for the 2009-2010 Season”

3:30pm – “An Empirical NFL Draft Value Chart”

4:05pm – Featured Speaker: Roland Beech of Dallas Mavericks

5:00pm – Concluding Remarks

For further information on each presentation, see the presentation descriptions.

Blind Side Project : How Good is D'Brickashaw Really

Week 4 Picks

The Blindeside Project: Some Jets Questions

NFL Week 3 Picks

Time in the Pocket and the Effects on a QB

Northern California Symposium on Statistics and Operations Research in Sports

The Blindside Project

Singletary was right

NFL Week 2 Picks

NFL Week 1

The Dixon Effect - What the Steelers Lose Without Big Ben

Vikings/Saints Preview

The NFL is Making Less Money????

Season Predictions

A funny thing happened on the way back to the Super Bowl...

Complete QB Rankings

The Blindside Hypothesis: Part 2

The Blindside Hypothesis: Part 1

QBs of the NFC East

Beyond ‘Moneyball’:
The rapidly evolving world of sports analytics, Part I