Team Efficiency Rankings - Week 13

The toughest schedules so far this year belong to AFC East teams, Miami and New England in particular. Jacksonville has also faced a tough slate of opponents to date. Aside from the NFC West teams that get to play each other twice, the softest schedule so far belongs to San Diego.

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

New Offensive Line Page

Aboutt a month ago, I introduced a method for valuing offensive line play using advanced statistics. The full explanation can be found in the original post, but the basic concepts goes like this:

The offensive line's job is, ironically, defensive. Lineman protect the quarterback or the rusher from tacklers. The best an offensive line can do is prevent their opposition from making plays, either by hitting or sacking the QB, by tackling a rusher for a loss or short gain, or by deflecting a pass. The impact that an opposing front seven defenders make can be measured by their +EPA or +WPA. These stats count only plays in which the defense scores a 'victory' by forcing a setback on the offense. An offensive line can be measured by how few 'victories' they allow, measured by how little +EPA or +WPA they forfeit to their opposing front seven.

In this implementation, the +WPA/+EPA of opposing defensive lineman are counted for all plays, and the +WPA/+EPA of linebackers are counted in only runs and pass rushes. In other words, pass defense plays by linebackers are excluded because the offensive line has no influence on linebackers in pass defense roles.

For example, the Patriots' offensive line leads the league with opposing defensive front sevens totaling the least +EPA with 84.6. The league average allowed +EPA for offensive lines is 127.6. That makes the NE offensive line 43.0 EPA better than average.

New Matchup Pages

You've probably noticed the win probability scoreboard at the top of the page. Compared to the old drop-down menu, it should be a much easier way to navigate from game to game to check out the graphs or leave comments. On Sundays it's a great way to stay on top of all the simultaneous games.

Clicking on a live or final game will take you directly to the WP graph for the game, but until now, there was nowhere to go by clicking on an upcoming game. Starting today, there is a matchup page for all forthcoming games, complete with one-stop shopping for a comparison of each team's stats. There's team efficiency stats, advanced team stats, plus conventional and advanced player stats for each team. I plan to add more features in the near future.

The matchup page started as something I thought might help Carson with his Weekly Notes feature or me with my contributions at the Post or Times, but I quickly realized it would be something everyone would like. It's a bit of an eye chart, but there's lots of good stuff in there. The efficiency table, which is the first one, might be the simplest and most useful.

For example, looking at Sunday's game between MIA and OAK, I can quickly see that MIA has the edge in passing, but OAK has the edge in running. On defense however, OAK has a hard time stopping the run. Both team have slightly better than average pass defenses. The difference in this game might come down to penalties, as OAK has one of the worst penalty rates in the league.

Roundup 11/27/10

Reader Gautham Venugopalan alerted me to this article that describes how Packers defensive coordinator Dom Capers grades out his defense. It sounds similar to a simplified success rate stat.

Behind the net on the randomness of NHL win-loss records.

What are the most prominent "wrong football beliefs?" Punting all the time, maybe. Running back overuse? The importance of running?

Bruce D shares his breakdown of lucky plays at the Community Site. He tracked each team's "lucky" plays, defined as non-repeatable things such as kick returns for TDs or fumble recoveries. He scores out each team in terms of these plays. Bruce then demonstrates that these plays do not correlate from one half of the season to the next, at least for 2009, suggesting they really are lucky plays. Awsome. I'd love to see more from Bruce and more from all you guys crunching numbers on your own.

This site appears to have some sort of EI algorithm for telling you which games on your DVR you should watch and which ones you should delete. Interesting idea, but how many people would use this? Helmet-knock: Tech Crunch and premium subscriber Borat. (For Christmas this year, I'm renewing Borat's premium subscription for half-price. I was raised to be very generous like that.)

Manning vs. Brady using EPA and WPA.

There's a great new ESPN stats blog, simply called Stats & Info. According to the site, it started as an in-house source of content for producers and analysts, but now it's publicly available. Nice blend of conventional and advanced statistical insight. Here are some example posts. Look for more advanced stats, including for football, from ESPN in the future. It's a niche they're looking to fill, and they have some very bright guys working on their stats team.

The Weekly League: Notes and Ideas for Week Twelve

This week's edition of The Weekly League features:

1. Previews of the Green Bay-Atlanta, Philadelphia-Chicago, San Diego-Indianapolis, and San Francisco-Arizona games.

2. A photo of a player relevant to the article -- always a good technique to attract the attention of readers.

and

3. Over 100 points of joie de vivre.

The Four Factors represent each team's raw performance in four important categories (pass and rush efficiency, pass and rush efficiency against) relative to league average (where 100 is league average and anything above is good).

Generic Win Probability (GWP) is the probability a team would beat the league-average team at a neutral site. It can be found for all teams here. Game Probability (PROB) is each respective team's chance of winning this particular contest. Those numbers (along with methodology) can be found here.

Finally, a glossary of all unfamiliar terms can be found here.

Green Bay at Atlanta | Sunday, November 28 | 1:00pm ET
Four Factors


Notes
• Pop quiz, America.
• What's more surprising:
• That (a) Green Bay is third in the NFL, with a 0.74 GWP, despite a rushing attack fronted by Brandon Jackson, or
• That (b) Atlanta is a very pedestrian 19th overall, with a 0.47 GWP, despite a glistening 8-2 record?
• In any case, as the Increasingly Renowned Brian Burke noted mid-week, the latter shouldn't be very surprising at all.

Roethlisberger's 2010

In several aspects, Ben Roethlisberger may be having a career year. 6.5 AYPA, +0.25 EPA/Play, and 2.3 sacks per game (which is good for him). Six games is not a big sample, but the Steelers offense has been firing on all cylinders since his return.

What's the Deal with the Falcons?

The Falcons are 8-2, tied for the best record in the NFL, but in terms of efficiency they are below average. The efficiency model has them ranked 19th out of 32 teams. So what's the story?

I often hear the refrain that Matt Ryan had a down year last year, but he's back as one of the best QBs in 2010. But he's throwing at a 6.2 net YPA pace, exactly average this season. The defense is giving up 6.7 net YPA, over a standard deviation worse than average. Their running game is nothing special either, gaining an average 4.1 YPC, while the defense gives up 4.3 YPC.

Atlanta hasn't faced a terribly tough schedule either. Their opponent GWP is slightly below average at 0.48.

So could it be a consistency thing? Do their SR numbers indicate they are deadly consistent, inching the ball up the field on offense and forcing frequent 3-and-outs? Nope, their opponent-adjusted team SR is the same as their efficiency ranking--19th.

Washington Post: Another Game Comes Down to the Final Play

Today's post at the Washington Post's Redskins Insider looks at how the Redskins were able to pull out another close one. Yet another Redskins game comes down to the final play.

-Where to the Redskins' games rank in terms of excitement this year?
-How is Donovan McNabb performing compared to his career averages?
-How big a role did penalties play in overtime?
-The 'Skins are 5-5. What are their chances at making the playoffs?

NYT: Weekly Game Probabilities

Weekly game probabilities are available now at the nytimes.com Fifth Down. This week I also put a number on Santonio Holmes' 2010 heroics.

Team Efficiency Rankings - Week 12

There's a ranking I never thought I'd see. Peyton Manning's offense is ranked 21st in the league in efficiency. Last week, I got a few questions about how that could be , and the answer is opponent strength. The Colts have put up some slightly above average passing numbers, but they've been against some very weak defenses, including HOU twice, JAC, WAS, and DEN. Even NE's defense is ranked 23rd, giving up 7.0 net yards per pass attempt. 

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

Advanced Team and Player Stats Updated

The database has been successfully repaired. Advanced team and individual player statistics are updated following each wave of games and currently include results through Monday night's game.

Advanced Team Stats
Opponent Adjusted Team Success Rate

Quarterbacks
Running Backs
Wide Receivers
Tight Ends
Players By Team


Defensive Ends
Defensive Tackles
Linebackers
Cornerbacks
Safeties
Defenders By Team

How Random Are Interceptions?

One of the most random events in football is the interception. We can say that not just because of tipped balls or freak gusts of wind, but because we know that interception rates for teams and for individual quarterbacks varies widely from season to season. They also vary greatly from one half of the season to the next, which suggests that more is at play than player ability. In a recent post I estimated the proportion of team win-loss records that can be attributed to sample error in an effort to demonstrate the technique.

I looked at the interception rates (int per attempt) for all ‘qualified’ NFL quarterbacks since from 2002-2009. To qualify, a QB must have thrown a certain proportion of his team’s pass attempts. The overall average interception rate is 2.9%. The standard deviation, from QB to QB, is 0.94%, which means about 2/3 of the QBs will have interception rates that fall somewhere inside 2% and 4%.

To calculate the variance due to randomness, I couldn’t directly use the formula from the binomial distribution. Unlike team seasons, which all have 16 games, QBs have varying numbers of attempts during a season. Instead, I used a random simulation to estimate the random variance. I started with the premise that all interceptions were completely random. What if every quarterback’s pass attempts each had a 2.9% chance of being intercepted, regardless of who he is or who the opponent is. What would the distribution look like then? How different would this purely random distribution be from the actual distribution we observe in the real NFL?

The Weekly League: Notes and Ideas for Week Eleven

This week's edition of The Weekly League features:

1. Previews of the Green Bay-Minnesota, Oakland-Pittsburgh, Indianapolis-New England, and Denver-San Diego games.

2. Responses to three readers on matters sundry.

and

3. Tons of whimsy.

The Four Factors represent each team's raw performance in four important categories (pass and rush efficiency, pass and rush efficiency against) relative to league average (where 100 is league average and anything above is good).

Generic Win Probability (GWP) is the probability a team would beat the league-average team at a neutral site. It can be found for all teams here. Game Probability (PROB) is each respective team's chance of winning this particular contest. Those numbers (along with methodology) can be found here.

Finally, a glossary of all unfamiliar terms can be found here.

Notes to Three Readers
To the Reader Who Asked for Fewer NFC East Games
Your wish is my command. Or, at least, this wish is my command. Other wishes, probably not so much.

To the Reader Who Asked Why the Unlucky Teams on the GWP Table Are Green
I'm drawn towards -- and, I'm guessing, many readers here are drawn towards -- teams that are better than public perception might suggest. The green -- as opposed to the red -- reflects bad luck as a "virtue."

To the Reader Who Asked About My Impressive Jawline, Whether It's Natural
Yes. It is. Cento per cento.

Green Bay at Minnesota | Sunday, November 21 | 1:00pm ET
Four Factors


Notes
• This is what you might call "a clash of NFC North foes."
• So that's one thing.
• But another thing is: have you ever frigging seen Clay Matthews play?
• He's 24th among linebackers in +WPA and sixth in EPA/G.
• But he's more like first or second overall in the category of "being scary."

Roundup 11/20/10

How does David Garrard have the second highest passer rating in the league?

After reading this, I don't think Bill Belichick really understands statistics the way many people thinks he does. He knows that most stats are junk trivia, but still seems unsure which ones are worthwhile. He appears to grasp that the bottom line goal is net point differential.

How to use PFR's 'Play Index' to search for player and team milestones.

How good are the 2010 Jets?

How unlucky are the 2010 Lions? (My 2 cents: Lions are not very good, but they're improving. I've got them 28th in terms of efficiency. They're lucky in terms of points scored, considering their ability to move the ball, but they're unlucky in terms of converting points into wins.)

What are statistics, anyway? As Tango put it: "If a blogger does it it's crap; If a coach does it it's great."

Game Probabilities: Week 11

Weekly game probabilities are available now at the nytimes.com Fifth Down.

Washington Post: The Silver Lining

Today's post at the Washington Post's Redskins Insider is a desperate attempt by an amateur football analyst to make sense of Monday night's game.

-Beat downs like that aren't are rare as most think.
-Vick's performance was historic any way you cut it.
-Meaningless silver lining: RB Keiland Williams had a good game.
-McNabb's three interceptions are not what they seem.
-Tennessee is good but hurt.

Efficiency Rankings - Week 11

One of the more perplexing things about the rankings lately is that the 7-2 Patriots have a net negative balance of offensive and defensive YPA. NE averages 6.5 net YPA but gives up 6.8 net YPA, while their other stats are very average. How are they winning and beating quality opponents (0.60 Opponent GWP)?

Perhaps part of the answer is consistency. Their offensive Success Rate, adjusted for opponent is 2nd in the league. They might only be getting 6.5 YPA, but if they got 6 or 7 yards on every attempt, they'd be unbeatable.

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

Offensive rank (ORANK) is offensive generic win probability, which is based on each team's offensive efficiency stats only. In other words, it's the team's GWP assuming it had a league-average defense. DRANK is is a team's generic win probability rank assuming it had a league-average offense.

GWP is based on a logistic regression model applied to current team stats. The model includes offensive and defensive passing and running efficiency, offensive turnover rates, defensive interception rates, and team penalty rates. If you're scratching your head wondering why a team is ranked where it is, just scroll down to the second table to see the stats of all 32 teams.

Best EPA Performances of the Past 10 Years

Michael Vick blew up the Redskins Monday night with 333 passing yards, 4 passing TDs, 80 rushing yards, 2 rushing TDs, and no turnovers. It was an almost flawless individual performance from beginning to end, worth 29.4 Expected Points Added (EPA). How does it compare to the top individual performances since 2000?

The decade's top performance belongs to Tom Brady, 2007 edition. His performance in the Patriots' 59-12 win at Buffalo was worth 33.3 points. Here are the rest of the top performances:

Predictivity

Success Rate (SR) is a simple measure of whether or not a play improves an offense's expected net point potential. It essentially ignores the magnitude of a play's result, and instead focuses only on whether a play was simply a good outcome or bad outcome.

Although team SR statistics ignore important information in terms of explaining past wins, it may be able to predict future outcomes better than other measures. A team's SR on run plays is particularly informative, because it is not sensitive to the low-frequency but high-impact events that are largely subject to randomness, such as long broken runs or turnovers.

Compared to simple running efficiency, run SR correlates better with winning (0.39 compared to 0.15). This is telling and helpful, but it only accounts for past outcomes. How well a statistic predicts future outcomes is not just about the parlor game of picking winners. Stats that predict future outcomes measure the signal of how good a team really is, underneath all the noise of randomness.

In other words, there is no 'right now.' There is no is. There is the known past, clouded by randomness, and there is the unknown future, clouded by uncertainty. Now is merely the ephemeral intersection between the past and future. When trying to measure team strength or player ability, the focus should be on how well a team or player is likely to play in the future.

The Weekly League: Notes for Week Ten

This week's edition of The Weekly League features:

1. Previews of the Minnesota-Chicago, New England-Pittsburgh, and Philadelphia-Washington games.

2. The GWP win/luck table that makes all the ladies scream.

and

3. Hella chagrin.

The Four Factors you see for each game represent each team's raw performance thus far in four important categories (pass and rush efficiency, pass and rush efficiency against) relative to league average (where 100 is league average and anything above is good).

Along with the Four Factors, you'll see two other numbers: Generic Win Probability (GWP) and Game Probability (PROB). The GWP is the probability a team would beat the league average team at a neutral site. It can be found for all teams here. The PROB is each respective team's chance of winning this particular contest. Your host, Brian Burke, provides PROBs to the New York Times each week, and those numbers (along with methodology) can be found here.

The following games have been chosen as they'll be available to the greatest portion of the network-watching audience, per the NFL maps at the506.com.

Finally, a glossary of all unfamiliar terms can be found here.

Minnesota at Chicago | Sunday, November 14 | 1:00pm ET
Four Factors


Notes
• As noted by our host Brian Burke, Jared Allen -- despite having what appears to be an underwhelming year by his standards -- is still second among defensive ends with a 0.96 +WPA this season.
• Allen is, among other things, second in QB hits, with 15.
• That thing about Allen is bad, on account of -- through Week Eight, at least -- Chicago was sporting the worst offensive line (as measure by WPA) in the league.
• Probably a whole bunch of that is from the game where the New York Giants sacked Jay Cutler something like 43 times in the first half.
• Yep. Just checked it out. It was 43 times exactly.

Roundup 11/13/10

Should Antwaan Randle El be used in trick plays more often?

How lucky or unlucky have today's crop of QBs been? The Rivers Index tells us.

"Getting the most out of a professional athlete does not involve filling his head with useless facts and statistics and probabilities, and filling him with fear of what may happen if he forgets them." That's the perspective of Nate Jackson, a TE who was cut from the Browns in August. That seems right on the money to me. The picture he paints of the Browns' training camp reminds me of Plebe Summer. Don't miss the interview Jackson links to in his article.

John Candido has posted up to date play-by-play data for everyone over at the Community site.

The Lions are improving.

Ryan and Flacco

Matt Ryan and Joe Flacco are squaring off tonight. Both QBs were taken in the 1st round in 2008 and they're often compared. Their performance is amazingly similar this season in many different measures, but each player has taken a very different path to get where they are.

Both QBs improved over the course of their rookie seasons, and you can even see it clearly in their stats.

While Flacco rode his defense to success his rookie year, Ryan's play in 2008 was the reason for the Falcons' success. Since then, however, Flacco has improved, making the standard jump from his first year to his second.

Ryan is said to have regressed slightly in his second year, but his apparent decline was due primarily to situational factors. His EPA in 2008 and 2009 was nearly equal, but his WPA was cut in half.

Game Probabilities: Week 10

Weekly game probabilities are available now at the nytimes.com Fifth Down.

Mid-Season WPA All Stars: Defense

It's just past the halfway point in the season, and we recently took a look at the top offensive players in terms of Win Probability Added (WPA). Now it’s time to look at the top defenders. Instead of WPA, we’ll use +WPA, explained here.

Let’s start with DEs. Kyle Vanden Bosch leads all ends with a +WPA of 1.37. He’s credited with 37 total tackles, giving him a tackle factor of 1.42—meaning he has 42% more of his team tackles than you’d expect from his position. Vanden Bosch racked up 5 sacks, 1 pass defended, 10 QB hits, 9 tackles for losses, and a forced fumble. His biggest play was at the end of the now infamous Redskins game in which he caused Rex Grossman’s fumble returned by teammate Ndomukong Suh for a touchdown.

Jared Allen is the runner-up so far DE. The party line is that he’s having an off-year with “only” 4 sacks, but that’s not all a DE does. Allen has 15 QB hits, often forcing hurried passes or even interceptions. He also has 26 total tackles, 3 passes defended, 2 tackles for losses and an interception. Justin Tuck also deserves a mention with his insanely high +EPA. Tuck’s best plays have, unfortunately for him, come in low leverage situations. Sorry Justin, no Advanced NFL Stats sweatshirt for you.

Mid-Season WPA All Stars: Offense

It’s just past the halfway point in the season, and it’s a good opportunity to take a look at which players are making the biggest impacts. I’ll look at offensive players in terms of their total contribution to their team’s fortunes using Win Probability Added (WPA). Although not perfect, there is no other metric better suited for judging MVP performance .

Starting with QBs, none other than Drew Brees leads the list with +2.67 WPA . Although the party line is that Brees is having a down year so far, primarily due to his 12 interceptions, WPA disagrees. Brees has led multiple comeback drives, even in games that were ultimately lost. Some of his biggest plays came in the overtime loss to the Falcons. Brees led the Saints into potentially game-winning FG position, but lost the game thanks to a missed kick. Brees isn’t simply lucky either. He’s 3rd in Success Rate (SR) and 3rd in EPA. Brees’ interception rate will almost certainly regress to a more normal level, and I'd bet we'll see the Saints offense to return to form in the second half of the season.

Joe Flacco is the runner up with +2.43 WPA, but he actually ties Brees with +0.30 WPA per game. The Ravens passing game has been the best component of Baltimore’s team so far. WPA is capturing things that other stats can’t, like deep passes that draw pass interference calls or good decisions to throw balls away or even take a sack. Flacco’s other numbers are rather average. His WPA is so high because of his comeback wins against the Jets, Browns, Bills, and Steelers.

Washington Post: Who Is Making an Impact?

Today's post at the Washington Post's Redskins Insider takes a look at which Redskins players on offense have made the biggest impact through the first half of the 2010 season.

Efficiency Rankings - Week 10

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

Offensive rank (ORANK) is offensive generic win probability, which is based on each team's offensive efficiency stats only. In other words, it's the team's GWP assuming it had a league-average defense. DRANK is is a team's generic win probability rank assuming it had a league-average offense.

GWP is based on a logistic regression model applied to current team stats. The model includes offensive and defensive passing and running efficiency, offensive turnover rates, defensive interception rates, and team penalty rates. If you're scratching your head wondering why a team is ranked where it is, just scroll down to the second table to see the stats of all 32 teams.

The Randomness of Win-Loss Records

Bill Parcells famously said "You are what your record says you are." Although that's undeniably true with regard to how the NFL selects playoff teams, and I wholeheartedly believe a leader needs to think that way, Parcells is only 58% correct. That's not a joke. It's 58%, and here's how something like that can be measured.

One staple of statistical analysis in sports is estimating how much of a given process is the result of skill and how much is the result of randomness or 'luck'. By luck, I’m not referring to leprechauns, fate, or anything superstitious. Real randomness is far more boring. Imagine flipping a perfectly fair coin 10 times. It would actually be uncommon  for the coin to come out 5 heads and 5 tails. (In fact, it would only happen 24% of the time). But if you flipped the coin an infinite number of times, the rate of heads would be certain to approach 50%. The difference between what we actually observe over the short-run and what we would observe over an infinite number of trials is known as sample error. No matter how many times you actually flip the coin, it’s only a sample of the infinitely possible times the coin could be flipped.

As a prime example, the NFL's short 16-game regular season schedule produces a great deal of sample error. To figure out how much randomness is involved in any one season, we can calculate the variance in team winning percentage that we would expect from a random binomial process, like coin flips. Then we can calculate the variance from the team records we actually observe. The difference is the variance due to true team ability.

Ndamukong Suh's XP

David writes in to ask, "It seems to me that Detroit would have been better off going for a two-point conversion after their kicker got injured, rather than having Ndamukong Suh attempt the PAT. What kind of confidence should Detroit have had in Suh's leg to make kicking make more sense than going for two?"

With Lions' place kicker Jason Hanson out due to injury, rookie defensive tackle Ndamukong Suh was called upon to attempt an extra point. Suh was chosen because he had won a place kicking competition in training camp to be the back-up kicker. Suh hit the right upright and missed the extra point, which turned out make a crucial difference in the game.

The Lions had scored 6 points to take a 3-point lead, 13-10, with 9:04 left in the 3rd quarter. A successful extra point would make the lead 4 points, and there can be a world of difference between a 3- and 4-point lead.

Roundup 11/6/10

Can wheat production estimates, candy bar weights, and attendance at Wimbledon improve your fantasy football team? Stein's Paradox says you can, but is it truly useful?

Wrong. It's actually 96.8%.

Neil Paine at PFR estimates how many wins Philip Rivers should have based on his insanely high Adjusted Yards Per Attempt. Neil also looks at how the Patriots are winning this year despite pedestrian yardage stats.

I have to say, I'm really impressed by the NFL Network's top 100 players. I thought I'd disagree far more than I do. I learned a lot from the series. Together with their series America's Game, NFL Network is doing some great work. But that should be no surprise, as both series are products of the indispensable NFL Films. One interesting thing is the comparison the expert panel's ranking with the fans' ranking.

The only issue I'll raise is that the #1 and #4 players had each other. Half of Jerry Rice's 10 all-pro seasons were partially thanks to Joe Montana, and all of Joe Montana's all-pro seasons were with Jerry Rice as his top receiver. And of course, they both benefited from Bill Walsh's visionary passing offense. Not that they don't belong at the top of the list, but no two other players at the very top of the list are as directly connected as Rice and Montana.

The Weekly League: Notes for Week Nine

This week's edition of The Weekly League features:

1. Game previews for Indianapolis-Philadelphia, Dallas-Green Bay, and Pittsburgh-Cincinnati.
2. An untinentionally glowing review of Ben Roethlisberger.

and

3. Equal parts vim and vigor.

The Four Factors you see for each game represent each team's raw performance thus far in four important categories (pass and rush efficiency, pass and rush efficiency against) relative to league average (where 100 is league average and anything above is good).

Along with the Four Factors, you'll see two other numbers: Generic Win Probability (GWP) and Game Probability (PROB). The GWP is the probability a team would beat the league average team at a neutral site. It can be found for all teams here. The PROB is each respective team's chance of winning this particular contest. Your host, Brian Burke, provides PROBs to the New York Times each week, and those numbers (along with methodology) can be found here.

The following games have been chosen as they'll be available to the greatest portion of the network-watching audience, per the NFL maps at the506.com.

Finally, a glossary of all unfamiliar terms can be found here.

Indianapolis at Philadelphia | Sunday, November 07 | 4:15pm ET
Four Factors


Notes
• Per the Interweb, DeSean Jackson returns this week after nearly getting killt by Dunta Robinson in Week Six.
• Per the Interweb, Michael Vick returns this week after getting sandwiched by some Washingtonians in Week Four.
• In Weeks One through Four -- i.e. the ones where Vick played -- DeSean Jackson was targeted 33 times, an average of just over eight per game. In Week Five, started by Kevin Kolb, Jackson was targeted only three times.
• Eight targets is kinda a lot to average. Consider: only 35 WRs were targeted 100+ times last season, and that only requires 6.25 targets per game.
• What this suggests -- but, of course, does not prove -- is that Vick looks for Jackson more than Kolb does.

Stories vs. Statistics

That's the title of an interesting essay by author of Innumeracy John Allen Paulos. Some highlights:

"...there is a tension between stories and statistics, and one under-appreciated contrast between them is simply the mindset with which we approach them. In listening to stories we tend to suspend disbelief in order to be entertained, whereas in evaluating statistics we generally have an opposite inclination to suspend belief in order not to be beguiled."

-and-

"Of course, the contrasts between stories and statistics don’t end here. Another example is the role of coincidences, which loom large in narratives, where they too frequently are invested with a significance that they don’t warrant probabilistically. The birthday paradox, small world links between people, psychics’ vaguely correct pronouncements, the sports pundit Paul the Octopus, and the various bible codes are all examples. In fact, if one considers any sufficiently large data set, such meaningless coincidences will naturally arise..."

Opponent-Adjusted Team Success Rate

Offensive and defensive Success Rate (SR) have been available on the advanced team stat pages all year. And recently I've been calculating opponent-adjusted team SR. Now I've been able to fully automate it, so it's always available following each wave of weekend games.

I like to look at SR for a couple reasons. First, it correlates well with winning. And second, it correlates well with itself, meaning it is relatively stable throughout the season. These are the two attributes we want in a stat for it to be predictive of future outcomes. For a couple weeks I've promised some hard numbers, but I haven't had time to wrap everything in a bow.

SR does not, however, tell the whole story. Teams that rely on high risk/reward plays, such as the Steelers this season, tend to be undervalued. SR is really a measure of consistency.

Weekly Game Probabilities - Week 9

Weekly game probabilities are available now at the nytimes.com Fifth Down. This week I also highlight the KC-OAK game and take a look at how coming off a bye week may give the edge to a couple teams this week.

Washington Post: McNabb and Grossman in the Clutch

This week's post at the Washington Post's Redskins Insider examined the epically bad decision to pull Donovan McNabb in favor of Rex Grossman in the final 2 minutes of last Sunday's game against the Lions. Was it a vindictive over-reaction by an arrogant head coach, or was it just a spiteful hasty judgment of an egotistical head coach?

Efficiency Rankings - Week 9

The team rankings below are in terms of generic win probability. The GWP is the probability a team would beat the league average team at a neutral site. Each team's opponent's average GWP is also listed, which can be considered to-date strength of schedule, and all ratings include adjustments for opponent strength.

Offensive rank (ORANK) is offensive generic win probability, which is based on each team's offensive efficiency stats only. In other words, it's the team's GWP assuming it had a league-average defense. DRANK is is a team's generic win probability rank assuming it had a league-average offense.

Saints Fake FG?

Chris asked the following:

The Saints had 4th and 3 on the Steelers 13 with no timeouts and 12 or 13 seconds left in the first half. The Saints lined up for a field goal and then changed it up to show that they were going to run a play. The Steelers called a timeout, and after the timeout, the Saints kicked the field goal.

In the post-game press conference, Sean Payton confirmed that the Saints were actually prepared to run a play and weren't just trying to get the Steelers to take a timeout.

I guess my question is this: would this fake field goal be a) a boneheaded call, or b) the most boneheaded call ever? I guess that Payton might have had some kind of spectacular play called, but it would have to be pretty spectacular to overcome the horrible odds that were against it. Only a touchdown (from the 13 yard line) would have given them a much better expectation than they had before the play, and most other results would have given them a far worse expectation.

At the end of the half, it's a slightly easier analysis because you don't have to kick off and a FG is worth a full 3 points and a TD is worth a full 7. Also, you don't need to worry about the EP value for missing the FG or failing to convert. It's just zero.

FGs are good from the 13 91% of the time, so a FG attempt is worth 2.7 EP.

Therefore, to go for the TD, you'd need the EP for going for it, regardless of whether it's a fake FG or normal scrimmage play, to be at least as high as the EP for the FG attempt. So:

Valuing Offensive Line Performance

Valuing the play of an offensive line is naturally difficult. Unlike other positions, an offensive lineman's performance on any given play is marked by the absence of things, such as sacks and stuffs. It's possible to cobble together a kluge, combining a variety of things like sacks or runs that make it past a couple yards. While this would probably give us some idea of which lines have played better than others, it's the kind of inelegant stat I like to avoid whenever possible. I like simpler, more graceful stats with meaningful units I can wrap my head around. WPA is a stat like that, its unit simply being wins. So when I was thinking about offensive line performance, I started by thinking about wins and win probability.

The irony of an offensive line is that its function is completely defensive. When you think about it, the offensive line's job is to protect the ball carrier or passer from attacking tacklers. Whether it's a run or pass play, the defense is really the side on the attack, and it's the offensive line's job to defend against their onslaught. Blocking, by its nature, is defensive.

At the core, the offensive line's job is to prevent the defensive line, and in most cases the linebackers, from doing their jobs. The lower the success of the defensive front seven, the higher the success of the offensive line. In fact, there's not much more to it than that. It doesn't matter whether a block is a quick chip or a Michael-Oher-from-Blindside-drive-block through the back of the end zone, as long as it prevents the defender from making the play. Realistically, an offensive lineman can do no better than prevent his counterpart from making a play.

To measure offensive line performance, we start by measuring defensive play-making, which is precisely the purpose of +WPA. +WPA (and +EPA) measures defender play-making by summing the WPA of all the plays in which a defender is credited for a defensive victory. In other words, only plays that are setbacks for the offense are counted toward a defender's +WPA. The half of all plays that are offensive victories should not be charged against the defender who made the tackle. Consider the backside DE who tracks down a RB from behind 10 yards down-field. The WPA was positive for the offense, but at the same time the DE should not be penalized for his effort. (This concept is explained in depth here.)

Who's Throwing Deep?

There are a couple updates to the individual stat pages.

First, there's a new column for deep percentage for quarterbacks. 'Deep%' tells us the percentage of pass attempts that are deeper than 15 yds throw the air. Ben Roethlisburger and Charlie Batch are at the top of the list. Vince Young, Joe Flacco, Derek Anderson, and Bruce Gradkowski round out the list of most vertical passers. Deep% has always been on the WR page, but now it's there for QBs too. Depth data is only available for seasons since 2006.

Great Article on Haley's Aggressive Playcalling

I love it when reporters "get it." This article is one great example. I spoke with Kent Babb of the Kansas City Star about Todd Haley's unorthodox fourth down doctrine earlier this week. Haley, who already earned one Advanced NFL Stats Coach of the Week Award (and a cool sweatshirt) for his aggressiveness, went for it two more times this past Sunday. He was one for two on those, but the way most people would look at it, he "passed up points" by not attempting field goals in a game that ultimately went into overtime.

Kent wrote me Sunday to ask if going for it on 4th and 2 from the 19 was a good idea today. One of my points in response was that sometimes kickers miss 36-yard field goals. As I typed that, Ryan Succop missed a potentially game-winning 36-yard field goal into swirling Arrowhead winds.