Best of Advanced NFL Stats 2010

This past year was a big one for this site. For the first time Advanced NFL Stats became a full-featured destination for advanced football analysis. There was always the blog, then came the win probability graphs, and this year marked a major leap forward with advanced statistics for players and teams, plus new features like the match-up pages and live comments during games.

I get the feeling some readers think I run this site full time, but it's really just a small hobby. (Ok, really big hobby.) I truly do it for fun and for the satisfaction I get from learning new things. In addition to running the site, 2010 was a busy year for me. I got married, bought a new house, moved (twice), and worked full time which included travel out of the country for almost a quarter of the football season. (Bet you couldn't even tell I was gone). I also trained for and completed my second triathlon.

Last year I did a fun post that highlighted all the best articles and features of 2009, and this post will do the same thing for 2010. There are lots of new readers, and I realize it's difficult to comb through the archives looking for interesting nuggets, so this will hopefully collect them all in one place. I'll start with last January and go in chronological order. Keep in mind this is only a fraction of the 243 posts of 2010.

The Raiders' Shane Lechler wins every punting award there is. We found out whether he really deserves them.

An historical analysis of the relative success of running and passing in the NFL explains many of the inefficiencies of modern offenses.

A look at fumbles answered some basic questions about one of the most random but important plays in the sport.

Late in the playoffs last year, I added a live comment feature as part of the WP graphs.

Kurt Warner retired, and the Hall of Fame discussion began. WPA weighed in.

NFL overtime rules were a hot topic, and in February I re-entered the debate with some hard numbers. There was a lot of bad information out there, not just about what the real numbers are, but what's the right way to look at them.

Is Adrian Peterson really any good? Do all his fumbles cancel out his big yardage gains? WPA and EPA tell us. (The 'official' WPA and EPA numbers have been updated/corrected since.)

WPA and EPA were used to break down the true MVP of Super Bowl XLIV.

At the end of February, I launched the initial player stats pages. Offensive skill players came first.

By March, the NFL's new overtime rules for the playoffs were leaked, and backward induction helped  show how the effect might be counter-intuitive.

Pages for advanced player stats by team went live. A couple weeks later, player career pages were launched.

I spent a lot of time thinking about how improve statistics for defensive players. Tackle Factor came to me in a dream. Actually, it came to me reading about Bill James' Range Factor for baseball fielding. TF still needs to be improved, and like everything here, it's not perfect but might be useful and interesting.

Bill Polian, GM of the Colts, is not a fan of statistics, except when they're pulled out of his butt.

In April, I was finally able to crack the code on how to scientifically measure coaching performance. The results were very surprising.

The Winner's Curse helps explain why free-agents turn out disappointing so often, (and why the Redskins stink).

Football Island, an article about a possible statistical reason why Samoans are so drastically over-represented in the NFL, was one of my favorites of 2010, not least because of some hilarious hate mail I received.

Rookies probably aren't overpaid the way everyone thinks, and there's a really good reason why both the players' union and the league want people to think they are.

I weighed in on the debate about whether top QB draft picks are really any better than later round picks.

In May, I introduced +WPA and +EPA, statistics that for the first time could measure a defenders' play-making impact on a game. Individual stats pages for defenders were also launched.

Here are two short and straightforward briefs I made that explain why offenses should be going for it on 4th down more often.

June was officially "other sport month" at Advanced NFL Stats. We saw the other kind of football take center-stage with the World Cup. I did some fun stuff with World Cup win probabilities.

Game theory has been a favorite topic of mine. Here is a sampling of some game theory research in other sports.

I released full play-by-play data for every NFL game of the past several years. For two years I tried and tried to figure out a way to first get access to, and then efficiently parse, play-by-play data in a meaningful way. When I was finally able to do it, it unlocked all kinds of possibilities for research. I hope there are other guys out there who can put the data to good use.

In July, I figured out that all the media reports about how Los Angeles has been struggling to build a stadium for an NFL team were completely wrong. Los Angeles has actually helped build several stadiums for NFL teams. It's just that they're not in LA.

One application of WPA is measuring how much officiating errors affect the outcomes of games. In August, the referee of Super Bowl XL publicly admitted significant errors in the infamous game, and WPA told us how much they affected the outcome.

Another application was measuring the biggest plays in Super Bowl history.

In September, I redid the 4th down analysis for my alma matter and was surprised by the results. It was an example of how EP analyses can be tailored to particular teams or opponents, and even different leagues.

September was an auspicious month because it marked the arrival of Carson Cistulli at Advanced NFL Stats. On loan from the awesome Fangraphs, Carson is a breath of fresh air, as he is not only a smart guy with a keen eye for numbers, but he's also one of the most entertaining sports writers out there. I've looked forward to his Weekly League and Notes column each week.

Football game theory is all zero-sum: I win, you lose. But that's only true if you think of each team as a completely unified whole. Players sometimes have their own interests, and they don't always overlap with their team's. The Prisoners' Dilemma shows us how a team can be, quite literally and mathematically, greater than the sum of its parts.

Comparing the success of 3-4 and 4-3 defenses.

The Advanced NFL Stats Community Site was re-launched, this time with much greater success. Longtime reader Ed Anthony kindly volunteered his services as the editor, overseeing some really solid articles from many contributors.

In September I added two new features to the site: a mobile version of the live win probability graphs and team-level advanced stats pages.

In addition to continuing contributions to the Fifth Down, this season I began contributing a weekly article for the Washington Post's Redskins Insider site. The Redskins had a disappointing and exceptionally uninteresting year on the field. The team has become all about internal drama, which doesn't show up on a graph very well. The most interesting thing about the experience was the quality of comments the posts received, which may say more about the local school systems than about much else.

In the 2010 season, I continued to do a lot of in-game strategy analysis, such as 4th downs, onside kicks, going for the 2-point conversion. (Just do a search for "blunder" and a bunch will come up in the results.) But the one that was most interesting to me, and the one that got the most attention, was the question of whether the Packers would have been better off allowing the Bears to score a touchdown unopposed in the final minutes of their week 3 match-up.

Another analysis from October did not get as much attention, but I think it was just as interesting as the Packers' dilemma. Up by 4 points in the final minutes and pinned against their own goal line, the Steelers may have been better off intentionally taking the safety and cutting their lead to 2 points rather than punting. Here's why. It's all about the mythical yet seductive concept of "field goal range."

What are NFL coaches optimizing with their strategies? It's clearly not winning. It's clearly not maximizing net point differential. This was one of the great mysteries of football research (to me) until it became apparent that they are optimizing individual play success. In other words, coaches are maximizing the simple count of plays they consider successful rather than fully weighing plays by the magnitude of their success. To a significant extent, they are still living by the old adage “Three things can happen when you throw the football, and two of them are bad.” This was one of my favorite articles since I started researching football, and whenever I mention it I must thank Chase Stuart, who contributes at the PFR blog and at the NYT's Fifth Down. Chase wrote to me suggesting that yards per carry might not be a good way to measure team strength in the running game because it is too sensitive to a few random break-away runs. It turns out he was right, and his suggestion unlocked the path to some really interesting stuff. It was a good lesson for me personally--sometimes we learn the most when question prior assumptions and discover we were wrong.

Marking another new application of advanced statistics, the win probability model was used to analyze a replay challenge decision.

Raheem Morris announced that stats are for losers. I know what he meant, and I agree to some extent. But his comment did give me the opportunity to write this

Like it or not, football is a game of numbers. Coaches like Morris are fond of pointing to the scoreboard as the only measure of a team. But take a good look at the scoreboard and what do you see? Numbers, and lots of 'em. Now look down at the field. What do you see there? Oh, look at that, big white numbers painted all over the field. The question is not whether or not numbers matter, because they certainly do. The only question is which numbers matter most.
In November I was able to add another feature to the site, opponent-adjusted team success rate.

Offensive linemen are an under-appreciated bunch, partly because they have no stats to highlight exceptional performance. It's also because their names are only mentioned for failures--giving up a sack, committing penalties, or allowing a stuffed run. So how do we statistically measure offensive line success when we can only measure failure? How can we....ahah! We measure the absence of failure. As one commenter put it, we're measuring the white space on the page, or in Holmesian terms, it's the dog that didn't bark.

Later that month, I added an offensive line statistics page.

Here's a look at which stats are the best and most consistent predictors of future team success.

How random are NFL team records? Here's one way to estimate it.

Michael Vick's incredible Monday night performance against Washington sparked another fun application of the WPA and EPA models. Here are the best single-game performances of the past 10 years.

What if interceptions were completely random? How different would it be from what we actually observe?

The Falcons continue to be a statistical anomaly.

I added a live win probability scoreboard feature to the site. But I wanted somewhere for the links to go before the games started, so I created match-up pages for each game, complete with comparisons of every team stat and top players.

December started off with an essay about how motivation ≠ money.

Short passes are, on average, no better than running. The advantage of passing comes when teams are able to throw deep.

This just in...Jamaal Charles is good. Really good.

Game theory has a lot to say about 2-point conversion attempts. It appears that teams should be running a lot more often, and it's possible that the equilibrium success rate is over 50%. This would mean that going for 2 is the value play rather than kicking the extra point. Wow.

How much of the game is passing, running, kicking, or punting? How much of football are interceptions? Fumbles? How much do penalties impact the outcomes of games? Here is one way to look at it.

Another new feature: advanced player stats by week. This is something I had on the back burner for a while and finally got around to getting it done.

Who are the most "clutch" QBs?

And finally, 2010 ends with a look at this season's All-WPA teams for offense and defense. Selectees to the all-WPA teams get to travel at their own expense to my house for a pick-up football game in my back yard...all wearing Wrangler jeans in front of American flags and pickup trucks, of course, just like NFL players always do.

Happy New Year!

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

0 Response to “Best of Advanced NFL Stats 2010”