≡ Menu

2016 AV-Adjusted Team Age: Defense

Being young isn’t by itself a virtue: the Browns ranked in the bottom 5 in points allowed, yards allowed, net yards per attempt allowed, net yards per rush allowed, turnovers forced, and first downs allowed. But Cleveland was, by far, the youngest defense in the NFL last season.

Yesterday, we looked at the age-adjusted offenses from 2016. Today we do the same for defenses, and the Browns were the youngest group in the league last year, with an average age of just 25.2 years. [click to continue…]


2016 AV-Adjusted Team Age: Offense

After each of of the last five years, I’ve presented the AV-adjusted age of each roster in the NFL. Measuring team age in the NFL is tricky. You don’t want to calculate the average age of a 53-man roster and call that the “team age” because the age of a team’s starters is much more relevant than the age of a team’s reserves. The average age of a team’s starting lineup isn’t perfect, either. The age of the quarterback and key offensive and defensive players should count for more than the age of a less relevant starter. Ideally, you want to calculate a team’s average age by placing greater weight on the team’s most relevant players.

My solution has been to use the Approximate Value numbers from Pro-Football-Reference.com, and to calculate age using each player’s precise age as of September 1 of the year in question.  Today, we will look at offenses; tomorrow, we will crunch these same numbers for team defenses. The table below shows the average AV-adjusted age of each offense, along with its total number of points of AV. Last year, the Rams, Jaguars, and Titans were the three youngest offenses. Each of those three are still in the top five this year, joined by the Bucs at #1 and the Seahawks at #4. [click to continue…]


The Colts were 0.2 points per game better than average last year, as measured by the Simple Rating System (which takes the points scored and allowed in each game, and adjusts for opponent strength and home field advantage).

The Vikings were 0.9. points per game better than average in 2016, and hosted the Colts in week 15.  Given those facts, we would expect Minnesota to have won by 3.7 points.  Instead, Indianapolis upset the Vikings, 34-6, beating the expected line by 31.7 points.  That was the least-conforming game of 2016 (you can view the least-conforming games of 2015 here).

The table below shows all 512 regular season games from 2016, and how it differed from expectations. Here’s how to read the first line. The second-least conforming game was came in week 3, and we can use it to help guide us through the table below. The Eagles hosted the Steelers, and Philadelphia had an SRS rating of +3.7, while Pittsburgh had an SRS of +4.7. As a result, we would expect the Eagles to lose by 2 points. Instead, they won 34-3, exceeding expectations by 29 points.
[click to continue…]

{ 1 comment }

2016 Pythagenpat Records

If you’re not familiar with how to calculate Pythagenpat records, you can read this post.

But the short version is, this is a slight upgrade on using Pythagorean records, which I assume most of you are familiar with. The formula to calculate a team’s Pythagorean winning percentage is always some variation of:

(Points Scored^2) / (Points Scored ^2 + Points Allowed^2)

Here, instead of using 2 as the exponent, we use a dynamic exponent that changes based on how much scoring occurs in each team’s games. Here are the 2016 Pythagenpat records: [click to continue…]


Three years ago, I looked at the single-season leaders in percentage of team rushing yards. Then and now, the top two seasons belonged to Edgerrin James: he had 94% and 92% of the Colts rushing yards in his first two seasons in the league. There were only three other seasons where a running back had at least 90% of his team’s rushing yards: Emmitt Smith in 1991, Barry Sanders in 1994, and … Travis Henry in 2002. In that post, I calculated for each team the percentage of his team rushing yards gained by that team’s top rusher. Then I calculated the league average percentage gained by each team’s top rusher, and plotted how that varied over time. This was intended to measure how running back back committee centric the league was in each year.

For a less rigorous method to measure RBBC-ness, you can see this post, which looked at games with more than 15 carries.

Both methods show RBBC being heavy in the ’70s, and the stud RB era peaking about 10 years ago.  But if you want to measure rushing concentration, a better method is probably to use the formula described yesterday. So for each team, I calculated the percentage of team rushing yards gained by every player on the team, squared that result, and then summed those numbers for each player on the team. You can read yesterday’s post for more info on the methodology, but here were the results for 2016: [click to continue…]


Tyreek Hill was noticeably absent from yesterday’s list of yards from scrimmage leaders. The main reason for that? Hill was a part-time player for the first seven weeks, failing to take the field on even half of Kansas City’s offensive snaps in even a single game.  By the end of the year, he was a more regular part of the offense, although he never participated in 70% or more of the Chiefs offensive plays in any regular season game (and in the playoff loss to Pittsburgh, he was present for 69% of Kansas City’s offensive snaps). The graph below shows the percentage of offensive snaps he was on the field each week of the 2016 regular season:

[click to continue…]


Doug Drinen wrote this article 11 years ago, and it serves as a good reminder to always look at offensive numbers in the context of a player’s team. Yesterday, I looked at tackle leaders as a percentage of team tackles.  Today we will do the same thing with yards from scrimmage.

Arizona running back David Johnson led the NFL in yards from scrimmage last year with 2,118 yards. The Cardinals as a team gained 6,157 yards of offense (before deducting for sack yards lost), which means Johnson gained 34.4% of his team’s total output. That also led the league. However, Steelers RB Le’Veon Bell missed three games due to suspension and sat out a meaningless week 17 game.  Bell averaged 157 yards per game last year, the third-most in NFL history. He was responsible for 30.7% of the Steelers total yards from scrimmage last year, but on a pro-rated basis (i.e., multiplying that by 16/12), that jumps to an insane (although not historically extraordinary) 40.9%.

That’s the column the table is sorted by below. Here’s how to read Bell’s line. He gained 1,884 yards for Pittsburgh, while the Steelers as a team had 6,137 total yards. Bell therefore was responsible for 30.7% of Pittsburgh’s yards, but he only played in 12 games. On a pro-rated basis, he ranks first at 40.9%. The table below shows the top 75 leaders in this metric, minimum 6 games played: [click to continue…]


Positive Air Yards per Attempt: 2017 Update

Adam Steele is back for another guest post. You can view all of Adam’s posts here. As always, we thank him for contributing.

Positive Yards Per Attempt: 2017 Update

If I could only share one thing from my time doing football analytics, it would be the following principle: Positive plays carry more weight than negative plays in determining the winner of a football game. I’ve already written a couple of articles on this subject and hope to further the cause with this update.


For those of you who don’t feel like reading the previous two posts, I’ll give you the basic gist. Since passing has a far greater impact on winning than running, I’ve focused my research on quarterbacks, but the principle applies to the entire offense (defense, not so sure). Despite everyone constantly harping on turnover avoidance, a potent passing offense is usually able to overcome giveaways. Conversely, avoiding turnovers is normally not enough to overcome a weak passing game. Furthermore, turnovers are highly random and situation dependent, so it follows that turnovers are a very poor method of gauging quarterback performance. Even though sacks are largely the quarterback’s fault, they are also very context dependent and only contribute a small amount in determining game outcomes. More importantly, the majority of signal callers trade sacks for interceptions or vice versa, so it’s no really fair to include one but not the other. [click to continue…]


The Jay Schroeder Index

Yesterday, I looked at the quarterbacks who were the biggest checkdown artists: i.e., which players had the best completion percentages and lowest yards per completion averages. I measured this by calculating how many standard deviations above/below average each quarterback was in those two categories in each year.

Today, the reverse. And the big winner is rookie Terry Bradshaw. We all know Bradshaw stunk as a rookie. He had a whopping 11.0% interception rate, which was horrible even for 1970. In fact, he has the second most attempts in history by a player with an 11% or worse interception rate. And since Bradshaw also ranked dead last in completion percentage, he ranked 2nd to last in ANY/A that year.

Of course, you might wonder: how could someone with the worst completion percentage and by far the worst interception rate not rank last (by a mile) in ANY/A? Well, it’s because Bradshaw ranked 2nd in the NFL in yards per completion as a rookie. He was your ultimate boom/bust passer, finishing 2.75 standard deviations below average in completion percentage and 2.18 standard deviations above average in yards per completion.

The top of the list features a bunch of interesting names, but I’m calling this the Jay Schroeder Index for a reason.  Schroeder only had 8 seasons where he threw at least 200 passes, but he makes the top 200 in 6 of those 8 seasons!  Schroeder made the list in ’86, ’87, and ’88 (despite moving from the Redskins to the Raiders this year), and then in ’90, ’91, and ’92.  He only missed the list in 1989 during this run, and that’s because he threw just 194 passes.  But in 1989, of the 34 quarterbacks with at least 150 pass attempts, Schroeder had the lowest completion percentage (46.9%) and by far the highest yards per completion average (17.0, the best of his career).  In other words, Schroeder had a top-200 season in 6 out of 7 straight years, with the lone exception being perhaps his most Schroeder-esque season! Of course, Schroeder’s love of the deep ball isn’t new to readers of this site.

The table below shows the top 200 seasons based on the Schroeder Index, using the same formula as yesterday: [click to continue…]


The most efficient runner in NFL history? That depends.

Jamaal Charles is now a Denver Bronco, making him the second superstar running back in two weeks to join a new team at the tail end of his career. In his prime, Charles was a very good receiver and a player that could be the centerpiece of an offense. However, he will likely be remembered for a singular skill: rushing efficiency.

Charles has a career YPC average of 5.45, easily the best in history among running backs in the NFL. That number is at least a little misleading. While rushing efficiency has not soared the way passing efficiency has, we are currently in a high-YPC environment. Two years ago, I calculated era-adjusted yards per carry: at the time, Charles was at 5.49, while the league average was 4.21. For reference, the league average during the careers of Jim Brown, Gale Sayers, and Barry Sanders was 4.08, 3.95, and 3.93, respectively.

I am not a big fan of yards per carry as a statistic, but hey, it’s still interesting trivia. It’s a little silly and mostly an academic exercise, but let’s pretend that we replaced every Charles rush attempt with a league average rush attempt. How much worse off would Kansas City have been? Well, a whole lot. Let’s use his 2010 season as an example. He had 230 carries for 1,467 yards, producing an incredible 6.38 YPC average. The league average that season was 4.21, meaning he was 2.17 YPC above-average. Given his 230 carries, we would have expected him to rush for just 968 yards, meaning he produced 499 rushing yards above average. And for his career? Charles is at +1657. [click to continue…]


Peterson with a rare cameo by a good quarterback.

After a ten-year career with the Vikings, Adrian Peterson is now headed to New Orleans where he will get to play with Drew Brees.  It will be the second time Peterson has played with a Hall of Fame quarterback, after Brett Favre’s stint with the team beginning in 2009.

In ’09, the Vikings had a Relative ANY/A of +2.05, easily the best passing game the franchise has produced in the last decade.  In fact, the only other time in the last ten years that Minnesota had an above-average ANY/A was last year, when Peterson rushed for just 72 yards in three games.

Most of his time in Minnesota, though, the team’s passing attack has been below-average — or outright bad.  For example, in 2012, Peterson rushed for 2,097 yards.  That represented 17.9% of his career total, and it came when the Vikings had a Relative ANY/A of -0.94.  Overall Peterson has a weighted average RANY/A — i.e., the Vikings RANY/A in each season of Peterson’s career, weighted by the number of rushing yards Peterson had — for his career of -0.52.  Take a look. [click to continue…]


Manning didn’t have much help during his career

Yesterday, I looked at quarterbacks from 2016 who started at least 8 games and threw at least 150 passes. For those passers, I calculated how many standard deviations above average they were in Relative ANY/A (i.e., how much better they were, statistically, than average) and in winning percentage. I sorted the list by the difference between the two, to find the quarterbacks whose stats and winning percentages diverged by the largest amounts.

What about historically? I performed the same study going back to 1970. And the season that stands out the most is Archie Manning’s 1980 season. That year the Saints were the worst team in the league: New Orleans went 1-15, and every other team won at least 4 games.1 Manning started every game for the team because he actually had a strong season, at least statistically: he ranked 9th out of 30 qualifying passers in ANY/A, and had a Relative ANY/A of +0.53. That, of course, is pretty unusual given his team’s 1-15 record.

That stands out as the biggest example of a divergence of stats being more impressive than team record. The best 100 seasons (although by default, the table only lists the top 20) are below: [click to continue…]

  1. The Saints’ troubles continued into the draft; New Orleans selected George Rogers first overall, when two of the top four, and three of the top eight players went on to be Hall of Famers. []

Kessler in a losing effort.

In 2016, Browns rookie quarterback Cody Kessler had an uneven year. He went 0-8, but he ranked 24th in ANY/A out of the 31 quarterbacks who started at least 8 games and threw at least 150 passes. His stats weren’t great, but they weren’t 0-8 bad, either. In PFR’s Adjusted Net Yards per Attempt Index, which attempts to adjust for era, Kessler ranked 15th out of the 43 rookie passers to meet the 8 start/150 attempt threshold. It was a pretty good rookie season that came with an 0-8 record.

And then there was Brock Osweiler.  The Texans quarterback — now on the Browns — was dead last with a pitiful 4.34 ANY/A average last season.  But for the second year in a row, Osweiler produced a winning record despite poor play; Houston went 8-6 with Osweiler under center.

I calculated the winning percentage and Relative ANY/A (i.e., ANY/A adjusted for era) for each passer since 1970 to meet the 8 start/150 attempt threshold.  I then calculated the standard deviations above/below average each passer was in each category.  Here are the results for 2016, and here’s how to read the Kessler line: he started 8 games for the Browns and had a 0.000 winning percentage.  His Relative ANY/A was -0.34, so just a hair below league average.  He was 2.53 standard deviations below average in winning percentage, but only 0.28 standard deviations below average in RANY/A.  As a result, he was 2.24 standard deviations better in RANY/A than he was in winning percentage; that was the highest number on the list.  Passers at the top had much better stats than wins; passers at the bottom (highlighted by Osweiler) had better wins than stats. [click to continue…]


Last year, Tyrod Taylor led all quarterbacks with 580 rushing yards. Colin Kaepernick, in 12 games, ranked 2nd with 468 rushing yards, and no other quarterback had even 400 rushing yards. But Aaron Rodgers, Blake Bortles, Cam Newton, Marcus Mariota, and Andrew Luck all had at least 300 rushing yards, so 7 out of 32 teams had a quarterback with at least that many yards.

How does that compare historically? Two years ago, in one of my favorite posts/methodologies, I looked at how to measure quarterback rushing yards. Here’s what I did.

1) Calculate the percentage of league-wide passing yards by each player in each season. For example, Tyrod Taylor was responsible for 2.3% of all passing yards in 2016.

2) Calculate the weighted average league-wide rushing yards for each season. So we take the result in step 1 and multiply that by each player’s number of rushing yards. For Taylor, this means multiplying 2.3% by 580 for a result of 13.4 rushing yards. Perform this calculation for each player in each season and sum the results to obtain a league-wide total. For 2016, this total was 150.9 rushing yards (obviously Taylor was the biggest contributor among quarterbacks).

3) For non-16 game seasons, pro-rate to 16 games.

Perform this calculation for each season since 1950, and you get the following results: [click to continue…]


Joe Montana had what many consider to be the best performance in Super Bowl history. In Super Bowl XXIV against the Broncos, Montana completed 22 of 29 passes for 297 yards and 5 touchdowns, with 1 sack for 0 yards. Jerry Rice was the biggest beneficiary, catching 7 passes for 148 yards and 3 touchdowns, in a 55-10 blowout of the Broncos.

Do the math, and Montana averaged 13.23 Adjusted Net Yards per attempt that day. Making it even more impressive is that he was facing a Broncos defense that allowed just 3.89 ANY/A to opposing passers during the regular season. That means Montana averaged 9.35 additional ANY/A relative to the average Broncos opponent. Over 30 dropbacks, that’s 280 Adjusted Net Yards of Value that Montana added. That’s the most in Super Bowl history, just ahead of what Doug Williams did two years earlier against the Broncos.

In that game, Williams was 18/29 for 340 yards with 4 TDs and 1 INT, and one sack for 10 yards. That’s an ANY/A of 12.17, but it came against a slightly tougher defense: the Broncos allowed 3.77 ANY/A that season. So Williams was 8.40 ANY/A better than “expected” against Denver, over 30 dropbacks; that means he produced 252 ANY of value in the Super Bowl.

Below are those numbers for each of the 128 passers in Super Bowl history. For Super Bowls prior to 1981, I had to use estimated sack data rather than actual, with the formula for estimated sacks being simply (Team Sacks) * (QB Pass Attempts/Team Pass Attempts). [click to continue…]


Atlanta had a really, really good offense this year. My favorite statistic: the Falcons had 59 drives end in a punt or a turnover, and 58 end in a touchdown.  Atlanta averaged 3.03 points per drive this year, and yet, the offense has been even better in the playoffs.

There was no stopping Matt Ryan and the Falcons against Green Bay, as the group scored 44 points on 9 drives in the NFC Championship Game. In the division round, the Falcons scored 36 points on 9 or 10 drives against Seattle, depending on whether you want to treat the Falcons final drive of the game as a real drive.  In two NFC playoff games, Atlanta’s offense has scored 10 touchdowns, seen 5 drives end on punts, 3 end on field goals, with zero turnovers and one drive end with the clock running out.

Scoring 80 points on 18 or 19 drives translates to an average of 4.21 or 4.44 points per drive. Take an average of those two numbers, and the offense is still averaging a whopping 4.32 points per drive. How remarkable is that? Well, it’s the best average for any of the 102 Super Bowl teams in their pre-Super Bowl playoff games.

The NFL has not historically recorded drive stats, so I previously wrote how one can estimate the number of offensive drives a team has in a game or season.  I used that formula to measure the best playoff offenses entering the Super Bowl; unsurprisingly, the 1990 Bills were the previous hottest offense.

Against Miami in the division round, Buffalo had between 10 and 12 drives, depending on how you treat the final drives of the half (the Bills received the ball with 14 seconds left on their own 32, and took a knee) and the game (Buffalo received the ball with just over one minute to go, and ran three times for a first down to run out the clock). Those other ten drives ended as follows, in order: Touchdown, Field Goal, Field Goal, Touchdown, Touchdown, Interception, Field Goal, Touchdown, Touchdown, Punt. That’s 44 points on 10 real drives.

The next week, in the AFC Championship Game against the Raiders, the Bills had 11 or 12 drives, as the final drive of the game featured Buffalo taking a pair of knees to close out a 51-3 victory. The first 11 drives went: TD, TD, Interception, TD, missed FG, TD, TD, Punt, TD, FG, Punt.  That’s 44 points (Buffalo also scored on a pick six, and one extra point was missed) on 11 drives. [click to continue…]


Background reading:

Part I

Part II

Part III

I’m going to assume you have read the first three parts of this series; today, I want to go through how to adjust passer rating by era while keeping the weights of 5, .25, 20, and 25 on the four variables. As a reminder, here are the formulas used for the four variables in passer rating, once you ignore the upper and lower limits:

A = (Cmp% – .30) * 5
B = (Y/A – 3.0) * .25
C = TD% * 20
D = 2.375 – Int% * 25

Adjusted Completion Percentage

For completion percentage, we can do a simple era adjustment because the multiplier is not directly tied to league average. Instead, league average is intended to be 20% higher than the floor, which is 0.30 in the original formula. So we need to rewrite completion percentage as simply

A = (Cmp% – (League_Avg_Cmp% – 0.20) ) * 5

So in an environment where the league average completion percentage was 50%, you would insert 0.3 in the blue parenthetical; in 2016, tho, you would insert 43.0%. [click to continue…]


The regular season is now over, and it ended with a whimper. Of the 16 games in week 17, 6 featured margins of 17+ points, and a 7th had a Game Script of +15.2. There were two big comebacks, but they came in meaningless games: The Colts overcame a 17-0 deficit to win 24-20, with the final margin coming on a touchdown pass in the final seconds.

The Steelers, with their stars rested, trailed most of the day against the Browns. This game wasn’t meaningless to San Francisco, though: Cleveland led 14-0 early and 14-7 entering the 4th quarter; had the Browns won, the 49ers would have had the number one pick. Instead, Pittsburgh won in overtime, cementing a 1-15 year for the Browns.

Below are the week 17 Game Scripts. [click to continue…]


Note: the 2016 Game Scripts page is now updated through week 16.

The big win of the week, not surprisingly, came from the Patriots over the Jets. New England was a 17-point favorite over New York, and won by 38 with a Game Script of +20.6. That’s the third best Game Script of the season.

In the world of misleading final scores, the Cowboys beat the Lions by 21 points, but with a Game Script of only +6.6. The Lions actually led in this game, 21-14, late in the 2nd quarter, and trailed by only 7 with 20 minutes left. Dallas then scored two quick touchdowns, and neither team scored in the final ten minutes.

And the comeback of the week belongs to San Francisco. With 5:14 left, the 49ers trailed by 14 points, facing 3rd-and-10 from the Rams 13-yard line. Colin Kaepernick scrambled for a 13-yard touchdown, the 49ers defense forced a three-and-out, and the 49ers put together their second straight touchdown drive of 73+ yards. Then, the 49ers went for two — which is silly, given that the 49ers didn’t go for 2 after the first touchdown — and Kaepernick scrambled for the conversion. That’s how San Francisco won, 22-21, with a -5.5 Game Script.

Below are the week 16 Game Scripts results: [click to continue…]


Are Teams Not Throwing Enough Interceptions?

In before the first “Well the Jets sure are” comment…..

On Saturday night, the Texans/Bengals game opened with 12 straight punts. Here’s the drive chart, in reverse chronological order, after twelve drives:

[click to continue…]


A number of teams produced blowout wins in week 13, with two of those games coming on national television.  The Colts destroyed the Jets on Monday Night Football, 41-10, with a Game Script of +19.9 in a game that was never close. The Ravens murdered the Dolphins, 38-6, producing a Game Script of +18.1. And the Seahawks had a Game Script of +16.8 in a Sunday Night massacre against Carolina.

Only two teams won with negative Game Scripts: Oakland trailed Buffalo for much of the game, and was down 24-9 halfway through the third quarter. The Raiders ultimately won by two touchdowns, despite a Game Script of -1.1. The biggest comeback belonged to Tampa Bay, who won with a -2.2 Game Script in San Diego. The Chargers led for most of the game, including with a 21-17 lead entering the fourth quarter. The Bucs scored 11 points in the final frame to pull out the victory.

Below are the week 13 game scripts: [click to continue…]


The 2016 Bills Are The 1973 Bills, Reincarnate

In 1988, the Dolphins went 6-10. That was the only sub-.500 season the Don Shula/Dan Marino Dolphins ever had.

You probably won’t be shocked to learn that those Dolphins finished dead last in rushing yards and first in passing yards. After all, Miami finished first in pass attempts, and the team ranked 2nd in NY/A; meanwhile, the Dolphins ranked last in rushing attempts, and 23rd in yards per carry. The presence of Marino, a bad running game centered around Lorenzo Hampton and Troy Stradford, and a 6-10 record all paved the way for the 1st/last split.

In 2005, the Cardinals went 5-11, and also ranked 1st in passing yards and last in rushing yards. That team’s running game was terrible: Marcel Shipp and J.J. Arrington were the backs, and the team ranked last in yards per carry *and* rushing attempts (and rushing TDs), as Arizona finished 190 rushing yards behind every other team. But with Kurt Warner at the helm, a bad record, and that running game, Arizona finished 50 passes (including sacks) more than than any other team, and 327 more yards.

That doesn’t sound so weird, does it? But since 1970, those are the only teams to rank 1st in passing yards and last in rushing yards (seven others raked 1st/2nd and last/2nd to last). And only three teams have done the reverse, finishing first in rushing yards and last in passing yards.

The first, unsurprisingly, was the O.J. Simpson-led Buffalo Bills in 1973 during his historic campaign. Buffalo went 9-5 and finished first in YPC and 2nd in attempts, as Simpson had a 332/2003/6.0 stat line, while Jim Braxton (108/494/4.6) and Larry Watkins (98/414/4.2) produced solid numbers in support. Joe Ferguson was not very good at quarterback: Buffalo ranked last in pass attempts, 3rd-to-last in NY/A, and therefore last in passing yards (and TDs).

The presence of the 2003 Ravens in this group is not going to surprise any folks, either. The 10-6 Ravens had a great defense and a fantastic running game led by Jamal Lewis, who rushed for over 2,000 yards. Baltimore finished 1st in rushing attempts and 3rd in yards per carry, with Lewis doing most of the heavy lifting there. With Kyle Boller and Anthony Wright, the passing attack was pretty bad: it ranked 27th in NY/A and 32nd in attempts, so the last-place ranking in passing yards makes sense.

The third team is one most of you could probably guess: it’s the 2006 Michael Vick-led Atlanta Falcons. Atlanta finished 1st in rushing attempts *and* 1st in yards per carry, joining the famous 1978 Patriots as the only teams since the merger to pull off that feat. The Falcons rushed for 2,939 yards, the most by any team since 1984. The passing game led by Vick was not very good: Atlanta ranked 29th in NY/A, and since it ranked 32nd in attempts, it ranked last in passing yards.

So why bring up those teams today? The 2016 Bills rank 2nd to last in pass attempts (by 1, to Miami) and 2nd in rushing attempts (Dallas), in a very 1973 Bills-like fashion. Buffalo easily leads the league in yards per carry (5.3), although right now the Cowboys (thanks to quantity) are only half a yard per game behind the Bills. LeSean McCoy is averaging 5.2 yards per carry, Mike Gillislee is at 5.8, and Tyrod Taylor is at 6.4; that group is powering an insanely efficient running game.

The passing game, meanwhile, is nearly as bad as the running game is good. That’s mainly because of a drop in yards per completion (from 8th last year to 25th in 2016). Taylor averaged 7.10 ANY/A this year and 5.73 this year; Buffalo ranks in the bottom 5 of the league in NY/A, so given the 31st-place ranking in attempts, it’s not too shocking that the Bills are last in passing yards (though the 49ers are less than 50 yards ahead of them).

As a result, the Bills look a lot like the ’73 Bills, and those two teams could make up half of the franchises since 1970 to rank last in passing yards and first in rushing yards.


The Dolphins have now won five straight games for the first time since 2008, with Sunday’s win being the most remarkable: Miami won with a Game Script of -6.8, as the offense had a very slow start to the day:


In week 11, Miami was the only team to win with a noteworthy negative Game script: technically, the Raiders and Giants won with them, too. Below are the full results from week 11: [click to continue…]


Blair Walsh In Perspective: Game-By-Game EPA

Minnesota Vikings kicker Blair Walsh was released by the Vikings this week, and given his struggles this year, it’s hard to argue with Minnesota’s decision. Walsh will be infamously remembered for missing a chip shot in the playoffs against the Seahawks last year, and those demons have carried over to his 2016 performance.

How much so? I looked at every kick of Walsh’s career, beginning in his rookie season of 2012. For every made extra point in 2012, 2013, or 2014, I gave him +0.01 points, and +0.06 points for every made extra point in 2015 or 2016. Then, for every miss, he received -0.99 or -0.94 points, as applicable.

Extra points were easy; field goals were slightly harder. The graph below shows the average success rate on field goals in 3-year increments, from 2012 to week 10 of 2016:


I used those numbers to give Walsh points for each field goal attempt, too. For example, 48-50 yard kicks have been made 70% of the time over the last five years, so if Walsh attempted a 49-yard field goal, I gave him +0.9 points if he made it, and -2.1 points if he missed it.

Using that methodology, here is how Walsh has fared in every regular season game of his career:


As a rookie, Walsh was at +11.4 in this system, and would be even higher if I era-adjusted in sample (for convenience, I treated 2012 games the same as 2015 games, which probably is not appropriate). In 2016, he had -7.1 points by this system, including two miserable games in weeks 1 (missed extra point, missed 37-yarder, missed 56-yarder) and 9 (missed extra point, missed 46-yarder in a game the Vikings lost in overtime).


Week 10, 2016 Game Scripts: Break Up The Titans

The Titans were the big Game Scripts story of week ten, as Tennessee rolled out to a 21-0 lead against the Packers. The Titans have been remarkable over the last few weeks: since the start of week five, the team is averaging 33.7 points per game, the most in the NFL.

Tennessee has scored at least 25 points in six straight games for only the third time in franchise history, and only the second time since the merger. The Titans have crossed the 35-point mark in three straight games, a franchise first.  The team is 2nd in yards per carry and 6th in yards per attempt; this is an offense flying high right now. [click to continue…]

{ 1 comment }

Today at 538: the Steelers and Seahawks had some interesting two-point conversion decisions in week ten.

According to ESPN Stats & Information Group, there have been 1,045 two-point conversion attempts since 2001, with teams converting 501 of those tries. That’s a 47.9 percent conversion rate; given that a successful attempt yields 2 points, that means the expected value from an average 2-point try is 0.96 points.

Interestingly, that’s almost exactly what the expected value is from an extra point these days. Since the NFL moved extra-point kicks back to the 15-yard line last season, teams have a 94.4 percent success rate, which means that an extra point has an expected value of between 0.94 and 0.95 points.

This means that, all else being equal, the average team should be indifferent between going for two or kicking an extra point. Unless the game situation (i.e., late in the second half) or team composition (e.g., a bad kicker, or an offense or an opposing defense that is very good or very bad) changes the odds considerably, the decision to go for two or kick an extra point shouldn’t be controversial. In the long run, things will even out, because the expected value to the offense is essentially the same in both cases.

That’s the long run. In the short run, there will be ugly outcomes. And we saw two of those play out this weekend.

You can read the full article here.


There was just one 4th quarter comeback in week nine, and it came in the Lions/Vikings game.  Trailing 13-9 with 4:14 left, Minnesota embarked on a 19-play, 79-yard drive for a touchdown to take a 16-13 lead.  That would have been the only 4th quarter comeback of the week, but Matthew Stafford completed two passes for 35 yards to put Detroit in position for a field goal to tie the game.  Matt Prater connected from 58 yards, and the Lions won in overtime.

But the Lions led for most of that game, and finished with a Game Script of +2.3.  Only one other winning team in week nine had a fourth quarter score to take the lead: Miami, who returned a kickoff for a touchdown against the Jets. But Miami led 14-13 at halftime, and 20-13 at the end of the third quarter; the Dolphins finished with a Game Script of +2.3.

So there were no teams that won games in week nine with a negative Game Script.  Below are the full results: [click to continue…]


538: Are The Eagles Better Than Their Record?

Today at 538: the Eagles appear to be much better than your average .500 team.

The Philadelphia Eagles are one of the more confusing teams in the NFL. At 4-4, it’s easy to assume that the Eagles are an average team, yet Philly has outscored opponents by 57 points this season, the third-best differential behind the 7-1 Cowboys and 7-1 Patriots. Furthermore, Football Outsiders has the Eagles first in the NFL in defense-adjusted value over average, a metric that measures team performance on a play-by-play basis. So what’s the deal — are the Eagles secretly one of the best teams in the league, or have they somehow gamed the system?

The obvious reason the Eagles are 4-4 despite putting up impressive numbers in the two stats mentioned above is that they clustered a lot of very strong play into just four games. In the team’s four wins, the Eagles have outscored opponents by a total of 76 points, an average of 19 points per victory. That makes Philadelphia one of four teams with an average margin of victory of at least 19 points in wins, joined by the Steelers (19.3 in four wins), Cardinals (23.3 in three wins) and 49ers (28.0 in one win).

You can read the full article here.

{ 1 comment }

Three victorious teams stood out in week eight as pass-heavy:

  • New England blew out the Bills, and led 41-17 until the final minute of the game. But despite a Game Script of +13.0, that didn’t stop the Patriots from throwing on over 60% of all plays. Tom Brady has deservedly received a lot of press this week, but the ratio against Buffalo was also a sign of an emerging problem: the running game hasn’t been very good. LeGarrette Blount had 18 carries for 43 yards, and New England’s running game has been inconsistent all year. Of course, that could just lead to more Brady throws, which may not be such a bad thing.
  • For Oakland, Derek Carr had 59 pass attempts (and just two sacks) in a monster game against the Bucs. The Raiders running backs had some success, but this was a competitive game throughout. That’s a sign, tho, that the Raiders want to put the ball in the hands of Carr and Michael Crabtree (16 targets) and Amari Cooper (15 targets).
  • The Chiefs rolled against the Colts, but even though Alex Smith went down, Kansas City stayed pass happy under Nick Foles. Kansas City passed more than you would expect from a team with a +8.0 Game Script, but that also may be a sign that the Colts pass defense is so bad that teams will pass on it regardless of situation.

Below are the Week 8 Game Scripts numbers. [click to continue…]


2016 ANY/A Update

Matt Ryan is having a career year, in a not disimilar way from what Carson Palmer did last season. Thanks to a superstar receiver and an offensive coaching staff that is drawing rave reviews, Ryan is having the sort of once-in-a-career year expected from a top-3 pick.  In fact, Ryan is even ahead of Palmer’s pace from last year:

Rk Age Year Lg Tm G W L Cmp Att Cmp% Yds TD Int Rate Sk Yds ANY/A
1 Matt Ryan 31 2016 NFL ATL 6 4 2 143 210 68.10 2075 15 3 117.9 15 98 9.52
2 Carson Palmer 35 2015 NFL CRD 6 4 2 125 193 64.77 1737 14 5 106.9 8 42 8.71

The Falcons ranked 17th in ANY/A last year, and 1st this year; Atlanta’s offensive ANY/A has jumped by 3.34 ANY/A, the biggest leap in the league. You might think the Jets — 14th in ANY/A last year, 32nd this year — would have the biggest decline, but New York’s dip is only the second worst. That’s because Palmer, who had a very lofty perch from which to fall, has been far below-average this season: [click to continue…]

Previous Posts