≡ Menu

In 2006, playing for the Detroit Lions, quarterback Jon Kitna was responsible for every pass attempt by the team. He wound up throwing for 4,208 yards, and ran for another 156 yards. The Lions, being a terrible 3-13 team, finished 32nd in rushing attempts (because they were always losing) and 2nd in pass attempts (because they were always losing).

So you won’t be surprised to see that the Lions threw for a lot of yards (7th most in the NFL) and ran for not many yards (last). And since Kitna took every passing attempt, well, Kitna was responsible for most of the Lions total yards. In fact, the 4,364 yards he totaled wound up representing 81.7% of the Lions 5,337 team yards from scrimmage. That is the most in a single season by any player in NFL history… until, maybe, now.

Here’s a good tip: if you see a stat that says Superstar X has the most Y in history, and it doesn’t tell you who currently has the most Y in history, there’s a good chance it’s a player who isn’t very good. (Also, and this is a more rare rule: if a stat says since X date, it usually means another player had a better season before X date. That’s not the case here; the “in the Super Bowl era” modifier was not necessary.)

Wilson has all 2,543 passing yards thrown by the Seahawks this year, and he also leads the team in rushing. Wilson has 290 rushing yards, so he’s accounted for 2,833 yards for Seattle this year, or 82.1% of the team’s 3,443 yards through ten weeks.

Below are the current list of single-season leaders in percentage of team yards: [click to continue…]


Adjusted Completion Percentage, Part 2

Last summer, I discussed that while completion percentage is a bad statistic, there’s one simple way to improve the metric: include sacks in the denominator.

If a quarterback takes a sack, that is *worse* than an incomplete pass, but it is *better* for the quarterback’s completion percentage. That is Just Plain Wrong.

As it turns out, this really impacts Peyton Manning and, to a lesser extent, Drew Brees. In 2003, Manning led the NFL in both completion percentage (67.0%) and adjusted completion percentage (64.9%). Technically, Manning didn’t win any other completion percentage crowns, although PFR gives him a tie in 2012. 1 However, he won the adjusted completion percentage crown a whopping five more times in his career: 2004, 2006, 2008, 2012, and 2013.

Brees has three completion percentage crowns (a fourth may come this season), but two more adjusted completion percentage titles. From 2003 to 2016, Manning won six AC% titles, Brees won give, and the rest of the league (Cousins, Brady, Palmer) won just three. In fact, from the six-year period covering 2008 to 2013, Manning and Brees won all of the adjusted completion percentage crowns.

The full list of leaders in each year since the merger are presented below, along with where that quarterback ranked in raw completion percentage (using a minimum of 224 passing plays per 16 team games for both metrics): [click to continue…]

  1. Technically, he lost it to Matt Ryan that year once you go out to two decimal places, 68.62% to 68.61% (although if you include sacks in the denominator but keep the minimum at 224 passing plays, Alex Smith was the completion percentage champion in 2012). []

Week 10 Game Scripts: Saints Stick To the Ground Game

The Broncos, Saints, Packers, and Steelers were your run-happy teams of week 10.

Denver made the quarterback switch to Brock Osweiler, which is a good reason why they decided to be run-happy. Osweiler had 33 pass attempts, while Broncos running backs had 27 carries… in a game Denver trailed 27-9 at halftime. Consider that New England had a +14.5 Game Script… and finished with a higher pass ratio than Denver!

New Orleans called a running play on 24 straight plays, and passed on fewer than 35% of their plays, the first time that’s happened for the Saints since 2001. Even with a +16.6 Game Script, that’s still incredibly run-heavy.

Green Bay was in a tight game throughout with the Bears, but Jamaal Williams had 20 carries and Green Bay rushed 37 times against the Bears, compared to just 28 pass plays for Brett Hundley (who also had only two rushing attempts).

Finally, Pittsburgh trailed most of the day against the Colts, but Le’Veon Bell still had 26 carries. The Steelers had a nearly 50/50 pass/run ratio, remarkable for a team playing with a -4.7 Game Script.

The full week 10 Game Scripts below: [click to continue…]


Let’s get to the week 9 Game scripts! Yes, these are a week late: my apologies, as well, other topics wound up being covered last week.

The biggest stories of week 9 were the blowout wins by Los Angeles, Philadelphia, and New Orleans. The Rams and Saints followed that up with another pair of blowout wins in week 10, while the Eagles were on bye. But before turning to week 10, let’s review some of the biggest outliers from week nine.

In week 9, the Jets and Panthers were very run-heavy. Lest you forget, the Jets beat the Bills on Thursday night in week 9, and while quarterback Josh McCown did have 5 carries, the running backs combined for 36 carries, while McCown had just 21 attempts. The Jets blew out Buffalo, but consider that the Lions had a similar Game Script and passes on 50% of plays.

Carolina beat Atlanta in a close game where the Panthers trailed for most of the first half. Still, behind Cam Newton and his 9 carries, Carolina wound up passing just 25 times while running 38 times! That’s really run-heavy.

The full Game Scripts data below: [click to continue…]


The Jets had an ugly 15-10 loss to the Bucs today, and quarterback Josh McCown was as responsible for it as anyone. Prior to some garbage yard throws, he had passed for just 157 net yards on 41 dropbacks with an interception, and the Jets first 11 drives (before a meaningless touchdown) ended with 7 punts, 2 turnovers, 1 FG attempt, and 1 turnover on downs.

But in the final seconds of the game, McCown managed to throw his 14th touchdown pass of the season. That set a new single-season career high for McCown, which is notable: that’s the oldest age any player set their single-season career high in passing touchdowns.

As I wrote earlier, McCown has turned into one of the great late bloomers in quarterback history. Of McCown’s 70 career starts, half of them have come with him at 34.4 years of age or older, giving him the fifth oldest median age of start in league history. But now he has another record all to his own.

Warren Moon set a career high with 33 touchdown passes at age 34 in 1990; 5 years later, Moon tied that mark at the age of 39. But he didn’t set a new career high at age 39, so the tie goes to McCown.

Similarly, Craig Morton originally set a career high in passing touchdowns in 1969 at the age of 26 with 21 scoring strikes; at age 38, in 1981, he again threw 21 touchdown passes.

Five player — Y.A. Tittle, Roger Staubach, John Elway, Steve Young, and Peyton Manning — set a new career high in touchdown passes at the age of 37. Those are the men McCown pushed aside it he record books today.

There are 301 quarterbacks in NFL history who threw for at least 10 touchdown passes in one season and are at least 35 years old in 2017. The graph below shows for each age, the number of QBs who set their career high at that age (and quarterbacks who tie that number later in their career get a 0.5 for each year; so age 26 and age 38 each get 0.5 for Morton). [click to continue…]

{ 1 comment }

Today at Slate:

While we’ve known for a long time that going for it on fourth-and-short is the wise move, NFL coaches have typically eschewed this aggressive approach. Are coaches getting any smarter? I took a shot at answering that question in August, analyzing fourth-and-1 decisions that fell within the following three constraints:

  • The decision must have come in the first three quarters before end-of-game factors encourage or discourage aggressive play.
  • The offense had to be between its own 40-yard line and its opponent’s 40-yard line, so kicking a field goal wasn’t an option, but the team wasn’t so close to its own end zone as to make fourth down conservativism a defensible move.
  • The game needed to be competitive, defined as within 10 points, to ensure the scoreboard wasn’t the primary factor dictating those decisions.

From 1994–2004, teams went for it on these fourth-and-1 situations 28 percent of the time. From 2005–2014, that number ratcheted up, with teams going for it 35 percent of the time. And in 2015 and 2016, offenses stayed on the field for these fourth downs more than 40 percent.

That trend is still holding halfway through the 2017 season.

You can read the full article here.


Carolina is #1… in percentage of rushing yards not by running backs

The Carolina Panthers have rushed for 982 yards this year, an average of 109.1 per game.  That ranks 15th in the NFL, and just a hair above the league average rate of 108.1 rushing yards/game.  But the Panthers don’t have anything resembling a traditional ground game: of those 982 yards, starting running back Jonathan Stewart has just 350 of them, while quarterback Cam Newton has 341 rushing yards, the most of any quarterback in the NFL in 2017.

In addition, wide receivers Curtis Samuel, Damiere Byrd, and Russell Shepard have combined for 87 yards; that’s the third-most rushing yards in the league for any team behind the Rams (Tavon Austin) and Raiders (Cordarrelle Patterson) among non-QB/non-RBs. In fact, Panthers running backs are averaging just 61.6 rushing yards per game, the fewest in the NFL.

This is hardly shocking, of course: Newton has been an incredible rushing threat since he arrived in the NFL in 2011. But it’s still interesting to see the numbers and understand that the Panthers are an above average team in total rushing, but dead last in rushing by running backs. The table below shows each team’s rushing yards in 2017 through nine weeks, both by running backs only and overall. Here’s how to read the table below. The Jacksonville Jaguars rank 1st in running back rushing yards, with 145.1 per game. The Jaguars also rank 1st in total rushing yards per game, at 166.5. For Jacksonville, 87.2% of their rushing yards have come from running backs. The Panthers rank last in both running back rushing yards per game and percentage of rushing yards by their running backs. [click to continue…]


Kirk Cousins is Spreading It Around

In the summer, I wrote an article describing the increased emphasis on spreading the ball around in team passing games. Through nine weeks of the 2017 season, which teams have the most and least concentrated passing games?

One way to measures this is to calculate the percentage of team targets had by every player on each team, square that result, and sum those squared results to get a team grade. Let’s use the Steelers as an example. Pittsburgh has 273 team targets this year, and star receiver Antonio Brown has seen 94, or 34%, of those targets. The square of 34% is 11.9%; perform those calculations for every Steelers who has a target this year, and the sum of those squares is 19.6%.

[click to continue…]

{ 1 comment }

Jared Goff was the worst quarterback in the NFL last season, and had one of the worst rookie quarterback seasons in modern history. This year, Goff is averaging 8.04 ANY/A, the 2nd-best in the NFL, and is on pace to set the NFL record for the largest year-over-year increase in Adjusted Net Yards per Attempt.

Dak Prescott had one of the best rookie quarterback seasons of all time in 2016, taking Dallas from the basement to the division crown. Prescott averaged 7.86 ANY/A last year, and isn’t far behind this year: he’s at 7.13 ANY/A, 7th-best in the NFL.

Carson Wentz had a very up-and-down rookie season, ultimately finishing with poor numbers and showing a lack of big play ability. This year, he’s having an MVP-caliber season: he ranks 5th in ANY/A at 7.55, and leads the NFL with 23 touchdowns as his Eagles have a 8-1 record. [click to continue…]


Largest Decreases in Team Scoring

Yesterday, I looked at the largest increases in team scoring from one year to the next. Today, the opposite: which teams have seen the largest decreases in scoring?

In the post-merger era, that “honor” would belong to the 1974 Falcons. In 1973, the Falcons averaged 22.7 points per game, 7th-best in the NFL. The team was led by fullback Dave Hampton and quarterback Bob Lee, and while both returned the next season, the results were disastrous. Atlanta averaged just 7.9 points per game, the lowest in the NFL. Along with the 1977 (not ’76) Bucs, the ’74 Falcons are one of just two teams since 1950 to average fewer than 8 points per game.

In more modern times, the 2015 Cowboys (after losing Tony Romo), 2011 Colts (after losing Peyton Manning), and 2010 Vikings (in year two under Brett Favre) are the biggest decliners. The top 100 biggest declines below: [click to continue…]


Largest Increases in Team Scoring

Last year, the Los Angeles Rams scored 224 points, or just 14.0 points per game. That ranked last in the NFL, 40 points behind the 31st-ranked Cleveland Browns. But as noted earlier this week, the 2017 Rams have scored 212 through seven games, a 30.3 points per game average that ranks 2nd in the NFL.

If that holds, the Rams increase of 16.3 points per game would rank as the third largest ever, and the biggest increase since 1950. The table below shows the 100 biggest per-game increases in scoring in pro football history: [click to continue…]


Let’s get to the week 8 Game scripts! The Raiders and Redskins stood out as pass-happy this week, as Derek Carr and Kirk Cousins kept passing in losing efforts. They were the two most pass-happy teams without accounting for Game Scripts in week 8, at 78% (Oakland) and 74% (Washington), respectively. But even after accounting for Game Script, both team’s pass-heavy nature stands out.

The Washington-Dallas game was competitive most of the way, but Cousins had 43 dropbacks while the team had just 15 rushing attempts.  Meanwhile, the Raiders called 49 pass plays against just 15 rushing plays, although that may have been due to Marshawn Lynch being suspended and both backup running backs underwhelming. This was just the second time in the last five years the Raiders passed on 77% of more of their plays. 

The full Game Scripts data from week 8, below: [click to continue…]


Jameis Winston‘s average pass completion has traveled 8.24 yards in the air, the longest distance in the league.

On Carson Wentz‘s average completion, the ball traveled 8.11 air yards.

And Marcus Mariota? His average completion picked up 7.61 air yards before being caught.

Those are the three most vertical passers in the NFL this season by that metric.

Last year, Winston and Mariota ranked 2nd and 3rd in this category: Winston’s average gain was 7.89 yards before being caught, Mariota’s 7.86. Cam Newton led all passers at 8.14 air yards on completed passes. But Wentz? He ranked 27th out of 30 qualifying passers, at 5.34 yards.

In 2015 — the rookie years for Winston and Mariota — Winston ranked 2nd behind Carson Palmer with an 8.13 average; Mariota was in the top 10 at 7.24.

Wentz is having a remarkable season: he ranks 4th in yards per pass attempt, and ranks 1st in yards per completion. Winston ranks only 4th in yards per completion, while Mariota is down at 11th in yards per completion. That’s because those two — and especially Mariota — aren’t getting much yards after the catch from their receivers. Mariota and Winston are both getitng just 4.2 yards of YAC per completed pass, ranking them both in the bottom six of that metric. Wentz ranks 19th with 4.8 YAC per completion.

The graph below shows Air Yards per completed passes for each quarterback in the 2017 season on the X-Axis, and Yards After the Catch per completed passes on the Y-Axis. Mariota, Wentz, and Winston are all to the far right of the graph, of course: [click to continue…]


Adam Gase was hired as the Dolphins head coach last year. His tenure with the team has been both successful and underwhelming, which is pretty hard to do. The Dolphins are 14-9 under Gase, tied with the Packers for the 8th-best record in the NFL. It feels hard to imagine, but Miami has a better record than Philadephia or Denver since 2016, and has as many wins as the Falcons.

On the other hand, Miami has a -77 points differential, which is the 7th-worst in the league. That’s a very stark difference: most teams have records that are proportional to their points differential, but not Miami. Tennessee (11th in record, 20th in points differential) and Houston (14th, 23rd) are the next two biggest outliers in that direction, with winning percentage ranks that are 9 slots better than their points differential ranks; Miami is at +17.5, by being tied for 8th in record and 26th in points differential. The Saints (t-19th; 8th), Jaguars (29th; 19th), and Chargers (t-27th; 18th) are the biggest underachievers by this method.

The graph below shows each team’s winning percentage (on the X-Axis) and points differential (on the Y-Axis) since 2016. Miami is a pretty large outlier: [click to continue…]


Last week, the Bears and Steelers were two of the most run-heavy teams in the league. That repeated itself in week 7, as Chicago and Pittsburgh finished with the two lowest pass ratios of the week.

The Bears ran just 37 plays on Sunday against the Panthers, the fewest by any NFL team in a game since 2010. So while the headlines may have been that Chicago threw just 7 passes (plus four sacks), to be fair to the Bears, that represented a 30% pass ratio, higher than what the team did last week.  Chicago had two long return touchdowns, which limited the offense to just 9 drives, six of which were three-and-outs.  But the Bears are clearly looking to throw as infrequently as possible, making them the most run-happy team in the NFL.

In Pittsburgh, Le’Veon Bell had over 30 carries for the second straight week. The Steelers had a Game Script of +7.3, but consider that they had a lower run rate than the Cowboys or Jaguars, who both had nearly 20-point Game Scripts! Pittsburgh finished the day with 25 runs and 43 pass attempts. The Steelers defense limited the Bengals to just 18 yards on 6 second-half drives (two of which ended on interceptions); with that dominant a performance, expect Pittsburgh to continue to rely on the ground game.
[click to continue…]


Drew Brees Finally Has A Defense Again

Did you know: The Saints currently lead the NFC in the Simple Rating System? And for a change, it’s not *all* because of Drew Brees and the passing game. The graph below shows how the Saints pass offense and pass defense have fared in ANY/A in each year since 2006. As a rule of thumb, you want the gold line (offense) to be a lot higher than the black line (defense); but despite Brees, that hasn’t always been the case.

As an aside, how remarkable is it that the 2015 Saints with such a historically bad defense still went 7-9?

Anyway, you can see that the pass defense is faring very well by Saints standards, similar to what New Orleans had in 2009 (13-3), 2010 (11-5) and 2013 (11-5).  New Orleans is on a four-game winning streak, and all four games have been won in large part due to the defense: [click to continue…]


You remember the 2008 Bengals, don’t you? Remarkably, that Cincinnati team was led by two quarterbacks still in the league: Carson Palmer started the first 4 games (all losses) before elbow issues caused him to shut things down for the rest of the year. From there, a young Ryan Fitzpatrick took over, leading the team to a 4-7-1 record the rest of the way.

Cincinnati had some talented weapons at wide receiver. T.J. Houshmandzadeh had led the NFL in receptions in 2007, and in 2008 he still had another 92 receptions. Chad Johnson had just come off his fifth straight Pro Bowl season, but the ’08 year was the beginning of the end for the man once known as Ochocinco. Chris Henry, who had been a big play receiver the past few years in Cincinnati, was reduced to a possession player in this offense in ’08.

The offensive coordinator was longtime coach Bob Bratkowski, who manned that role in Cincinnati from 2001 to 2010. But the 2008 season was very different. The Bengals averaged just 8.83 yards per completion, the single lowest output in NFL history. On a team with two longtime NFL quarterbacks and two Pro Bowl wide receivers, Cincinnati somehow couldn’t manage to push the ball down the field with any sort of consistency. The longest reception of the year was a 79-yard completion to… running back Cedric Benson! Johnson and Henry combined for 72 receptions, but none of them went for more than 26 yards.

So why am I bringing up the 2008 Bengals? Well, the 2017 Dolphins (through 5 games) and 2017 Ravens (through 6 games) are both averaging just 8.5 yards per completion. Yes, those gunslingers formerly known as Jay Cutler and Joe Flacco are running two of the most anemic passing attacks we have ever seen.

Here’s the breakdown on the Miami side: [click to continue…]


Through six weeks, the 49ers and Browns were both 0-6, while the Giants were 1-5. That’s bad, but it’s notable because those were the only three teams in the NFL with a record that was worse than 2-4. And on the flip side, only two teams — the 5-1 Chiefs and 5-1 Eagles — had a record that was better than 4-2. In other words, 27 of the 32 teams in the NFL were within two games of .500; or thought of differently, 84% of the NFL teams had a winning percentage between 0.333 and 0.667.

That… is… unusual. The graph below shows the percentage of NFL teams that had a record between 0.333 and 0.667 after six weeks in each year since 1970. As you can see, 2017 has set a new mark for parity: [click to continue…]


Chicago upset Baltimore in week 6, and rookie QB Mitch Trubisky picked up his first career victory in the process.  But if you want to award credit to the Bears for the win, the passing game would be a distant third.

The Bears picked up a first down on just 23.8% of all passing plays, and that includes a halfback pass that went for a touchdown; Trubisky gained 4 first downs on 20 dropbacks.  But the Bears defense was outstanding, limiting the Ravens passing offense to a 20.5% first down rateJoe Flacco had a 48.8 passer rating; for his career, he is 2-16 in the regular season when he has a passer rating below 55.0.

But for Chicago, it was the running game that carried the day.  The Bears passed on just 28% of dropbacks, and Tarik Cohen and Jordan Howard combined to run 50 times for 199 yards. They also caught two passes for 23 yards and threw one pass for a 21-yard touchdown.  Given how competitive the game was (Chicago had a Game Script of just +3.3), you could argue it was the most run-heavy game of the season.

The full week 6 Game Scripts data below: [click to continue…]


The Purple People Eaters

Last Monday, I provided some initial thoughts on the relative values of completion and passing first down percentage. The next day, I looked at how teams with disparate performances in those two metrics. And last Wednesday, I looked at the best passing offenses in NFL history in first down rate on passing plays.

Today? A look at the best pass defenses at preventing first downs.  This time, I am also going to era adjust these ratings. In 1969, the Vikings faced 459 pass plays (410 pass attempts, 49 sacks) but allowed only 88 passing first downs. That’s a remarkable rate of just 19.2%, the best since World War II.   It’s also the best rate on an era-adjusted basis.  The league average in the NFL in 1969 was 29.0%, which means this iteration of the Purple People Eaters was 9.8% better than league average, the highest differential ever. [click to continue…]


Brown is averaging more yards per game than his uniform number.

Yep, that’s pretty good.

Dating back to December 16, 2012, and including the playoffs, Steelers wide receiver Antonio Brown has 578 receptions for 7,755 yards in his last 77 games. Brown has 48 catches in six games this season for 700 yards, a 116.7 yards per game average.

Before the 2015 season, I wrote that Julio Jones had maintained a 100 receiving yards per game average over 57 straight games. I did not include the postseason when I wrote that post, but Jones still would have qualified had I done so: he had 5,703 receiving yards in his last 57 regular season games and 305 receiving yards in his 3 playoff games during that time. Through the end of last regular season, Jones was still keeping this pace up, at 7,417 receiving yards through his last 74 games.

And following the Super Bowl, Jones was at 7,751 yards — or just 4 yards behind Brown’s pace — through his last 77 games. Even through week 3 of this year, Jones had 8016 receiving yards in his last 80 games, but he has had two poor games since: as a result, he’s fall slightly under the 100 yard/game pace in his last 82 games.

But Jones still is at over 100 yards per game through his last 79 games (that’s because the first 3 games in his 82-game streak weren’t great). In his last 79 games, Jones has 7,932 receiving yards, a 100.4 yards per game average. [click to continue…]


The Jaguars used the fourth overall pick on LSU RB Leonard Fournette. The addition of Fournette was supposed to change the team’s identity, and it seems… to have worked? Fournette finished with 28 carries for 181 yards and 2 TDs, although the last of those carries was a 90-yard touchdown run when the game was over. Still, the Jaguars finished with 37 carries against just 16 pass plays: that’s a 30% pass rate, easily the lowest of the week. Jacksonville has a great defense and seems to have figured out its running game, and that may be the formula for success for this team.

Blake Bortles completed just 8 passes on his 16 dropbacks for only 5 first downs, but that was enough given that the Jaguars forced five interceptions and returned two of those for touchdowns. The Raiders and Bears were the other teams that showed up as run-happy this week, which probably isn’t surprising given that both teams were starting new quarterbacks in week five. For Oakland, EJ Manuel had 29 dropbacks but gained just 137 yards, while the running game had 25 carries for 108 yards and a touchdown (Manuel had two carries for 15 yards). For Chicago, Mitch Trubisky picked up just 108 yards on 26 dropbacks, while the running game had 115 yards on 29 attempts (with Trubisky picking up 22 yards on the ground on three carries).

Below are the week 5 Game Scripts data:
[click to continue…]


Against the Saints in week four, Jay Cutler completed 20 of 28 passes for 164 yards, which translates to a sparkling 71.4% completion percentage. But that was about as misleading as it gets. Cutler also was sacked four times — a 12.5% sack rate — and several of those completions were pretty meaningless. In fact, Cutler threw just 7 first downs against New Orleans.

When you hear the 71.4% completion rate, you think: pretty good. You’d be wrong. Completion percentage ignores sacks (it shouldn’t), and it treats a completed pass for a first down the same as a completed pass for zero yards. On 32 dropbacks, Cutler threw just 7 first downs — a 21.9% rate that is more meaningful than his completion percentage. Why is it more meaningful? Well, the Dolphins were shutout against the Saints.

Want another example? Against the Redskins in week 3, Derek Carr completed 19 of 31 passes, for a nominally effective 61% completion rate. But Carr was also sacked four times (which, again, should be in the denominator when looking at completion percentage) and picked up just three first downs. Three! So while he completed 61% of his passes, Carr threw for a first down on only 9% of his pass plays against Washington.

How about from this weekend? In his first start of the year, Titans quarterback Matt Cassel completed 66% of his passes and produced a passer rating of 85.5 against Miami. That’s pretty good, right? Well, it isn’t when you have drives that like this that increase your completion percentage and passer rating:

Cassel was sacked six times on the day and threw for just 9 first downs. So while he was 21/32 on the stat sheet, he was also 9/38 at throwing for first downs, a very poor 24% rate. The Titans had 14 drives, and one of them was a 4-play drive for -3 yards that resulted in a field goal because it started at the Dolphins 24; the other 13 drives produced one touchdown, two fumbles, and ten punts. Tennessee lost, 16-10, despite Cassel completing 66% of his passes: or, maybe they lost because Cassel completed 66% of his passes playing that style.

A high completion percentage shouldn’t be any offense’s goal; instead, it feels like more and more quarterbacks (and offensive coordinators) are treating it like the ends and not the means.

The graph below shows completion percentage in the NFL (excluding the AFL) from 1950 through five weeks of 2017. That line is in blue and plotted against the Left Y-Axis; as you can see, it’s been increasing steadily over the last seven decades.  Plotted in orange and against the Right Y-Axis is the percentage of pass plays that have gone for first downs.  That’s also increasing, although it’s been a little bit bumpier. [click to continue…]


Week 4 Game Scripts: Buffalo Runs To Victory

The Buffalo Bills and their run-heavy ways are at it again. The Bills upset the Falcons in week 4, but it wasn’t because their offense stole the show. Buffalo scored only 23 points, and that included a defensive touchdown, a 55-yard field goal, and a 56-yard field goal. The Bills had just 15 first downs, and four of the team’s 8 drives ended with three or fewer yards. But Buffalo is going to stick with the ground game, even when it isn’t working. And that makes them a unique team in 2017.

The highlight for the offense came on a 19-play, 82-yard drive that took over 11 minutes off the clock. Here’s the full summary: bonus points if you can figure out whether or not this counts as run-heavy:

Full Play-By-Play Table
Quarter Time Down ToGo Location Detail BUF ATL EPB EPA Win%
3 8:20 1 10 BUF 12 LeSean McCoy left tackle for 8 yards (tackle by Keanu Neal) 14 10 -0.350 0.290 44.4
3 7:49 2 2 BUF 20 LeSean McCoy for 2 yards. Penalty on Eric Wood: Offensive Holding, 10 yards 14 10 0.290 -0.780 49.1
3 7:31 2 10 BUF 12 Mike Tolbert right guard for 7 yards (tackle by Dontari Poe) 14 10 -0.780 -0.600 48.1
3 6:51 3 3 BUF 19 Tyrod Taylor right end for 5 yards (tackle by Deion Jones) 14 10 -0.600 0.540 42.1
3 6:17 1 10 BUF 24 LeSean McCoy right end for 7 yards (tackle by Keanu Neal) 14 10 0.540 0.940 39.7
3 5:29 2 3 BUF 31 LeSean McCoy right tackle for 7 yards (tackle by Deion Jones) 14 10 0.940 1.470 36.4
3 4:51 1 10 BUF 38 Mike Tolbert right guard for 3 yards (tackle by Brooks Reed) 14 10 1.470 1.330 36.5
3 4:14 2 7 BUF 41 Mike Tolbert left guard for 6 yards (tackle by De’Vondre Campbell) 14 10 1.330 1.420 35.4
3 3:35 3 1 BUF 47 LeSean McCoy right tackle for 4 yards (tackle by Deion Jones) 14 10 1.420 2.320 30.4
3 3:02 1 10 ATL 49 Tyrod Taylor pass complete deep left to Zay Jones for 18 yards 14 10 2.320 3.510 24.5
3 2:19 1 10 ATL 31 LeSean McCoy right tackle for 4 yards (tackle by Keanu Neal) 14 10 3.510 3.510 23.6
3 1:39 2 6 ATL 27 LeSean McCoy right tackle for 7 yards (tackle by Sharrod Neasman) 14 10 3.510 4.240 19.8
3 1:01 1 10 ATL 20 Mike Tolbert left end for 8 yards (tackle by Damontae Kazee). 14 10 4.240 5.040 16.1
3 0:09 2 2 ATL 12 Patrick DiMarco left guard for -2 yards (tackle by Grady Jarrett) 14 10 5.040 3.890 19.3
4th Quarter
4 15:00 3 4 ATL 14 Penalty on John Miller: False Start, 5 yards (no play) 14 10 3.890 3.100 22.5
4 15:00 3 9 ATL 19 Tyrod Taylor pass complete short middle to Andre Holmes for 10 yards 14 10 3.100 5.140 14.5
4 14:20 1 9 ATL 9 LeSean McCoy left end for 6 yards (tackle by Brooks Reed and Keanu Neal) 14 10 5.140 5.530 12.4
4 13:37 2 3 ATL 3 LeSean McCoy right tackle for 1 yard (tackle by Brooks Reed and Grady Jarrett) 14 10 5.530 4.950 13.3
4 12:59 3 2 ATL 2 Tyrod Taylor pass complete short left to Jordan Matthews for 1 yard 14 10 4.950 3.550 17.7
4 12:09 4 1 ATL 1 Penalty on Tyrod Taylor: Delay of Game, 5 yards (no play) 14 10 3.550 2.990 18.8
4 12:03 4 6 ATL 6 Steven Hauschka 24 yard field goal good 17 10 2.990 3.000 18.6

The drive opened with NINE straight running plays, and 15 of the 18 plays were rushes! A pass on 1st-and-10 following those nine runs was the only non-third down pass of the drive. That’s just crazy. The Bills ended the day passing on just 39% of plays despite this being a back-and-forth game throughout.

Below are the week 4 Game Scripts: [click to continue…]


They’re not very good.

You’ve undoubtedly heard that only one team in NFL history has started a season 0-4 and made the playoffs. That team was the 1992 San Diego Chargers, who shocked the world by going 6-10 in 1990, 4-12 in 1991, and then making the playoffs in 1992 after an 0-4 start.

But as Jason Lisk has pointed out before, many stories about 0-X teams missing the playoffs ignore the fact that 0-X teams usually are not very good. There are four 0-4 teams in the NFL right now: the two worst teams in the NFL by record last year, the Browns and 49ers, and two pretty good teams from last year, the Giants and Chargers. If New York or Los Angeles was to make the playoffs this year, it would be pretty remarkable. But probably not as remarkable as you might think.

Prior to 2017, there were 113 teams that began a season 0-4 in the 16 game era, which might make you think there’s only about a 1% chance of making the playoffs from this far behind. Indeed, of that group, the ’92 Chargers won 11 games, no team won 10 games, and the 2004 Bills were the only team to win 9 games.

But what would the Giants or Chargers need to do to make the playoffs? Probably win 10 games: i.e., finish 10-2 in their final 12 games. How rare is that? Prior to 2017, there have been 1,116 team seasons during the 16-game era. And 80 of those teams went 10-2 or better in their final 12 games (including the ’03 and ’07 Patriots, and ’04 Steelers, who went 12-0). In other words, about 7% of NFL teams finish 10-2 in their last 12 games. Which is a lot higher than one percent.

So the real question when discussing the Chargers and Giants isn’t how likely is an 0-4 team to make the playoffs, but how likely as the Chargers and Giants to play like a top 7% team the rest of the season.


Over their last 16 games — the final 12 games of 2016 and the first 4 games of 2017 — the Cleveland Browns are 1-15. Over the Browns last 16 games before that, Cleveland is 2-14: Cleveland began the 2016 season 0-4, and was 2-10 over the final 12 games of 2015. But the first game in that streak was the 5th game of the 2015 season, a rare Browns victory. Which means if the Browns lose to the Jets on Sunday, Cleveland will be 1-15 in their last 16 games, and 1-15 in the 16 games prior to that.

How bad is a 2-29 stretch over 31 games? Well, the Browns are just the fourth franchise to pull off that feat.

The World War II Cardinals

The Chicago Cardinals lost 29 games in a row during World War II, and went 1-36 during one 37-game stretch. [click to continue…]


Making The Playoffs With A Backup QB

The Texans opened this season with Tom Savage as the team’s starting quarterback. This was intentional, and by that I mean Houston really wanted to do this (as opposed to situations like the 2016 Patriots or 2014 Panthers who opened the season with — despite what pedants might say — their starting quarterback on the sidelines due to suspension or injury). Even the 2016 Cowboys opened the season with Prescott as their starting quarterback, although that wasn’t exactly how they opened the preseason.

But if the Texans make the playoffs, it will be because of Deshaun Watson (at least, as opposed to because of Tom Savage). Absent injury, Watson will be the team’s starting quarterback for the majority (if not all) of Houston’s wins in 2016. That separates Houston from teams like the 2016 Dolphins or 2014 Cardinals, who may have made the playoffs with backup quarterbacks but still saw Tannehill and Palmer lead their team in wins.

So how rare would it be for a playoff team to make the playoffs while riding a true backup quarterback? Obviously the definition of a true backup is open to interpretation, but I am referring to situations where the quarterback who led the team in wins was not atop the team’s depth chart as the season began (excluding injuries/suspension — Watson didn’t begin the year on the bench for that reason). [click to continue…]


Field Goal Rates Throughout NFL History

Yesterday, I wrote about Nick Lowery and why I think he was the greatest field goal kicker in NFL history.  That post was pretty long — I probably should have broken it into two parts — but I’d welcome any more discussion on the topic here or there.

So today I’ll keep it short and sweet: a reminder on how necessary era adjustments are when discussing field goal kickers. The graph below shows the field goal success rate throughout history. From 1960 to 1964, the average success rate was 50 percent. Over the last five years, the average rate was 85 percent.

Even more remarkable is that kicks are being attempted from farther away now, too. In 1960, the average kick was from 30.9 yards away; the average successful kick was from 26.2 yards out, while the average miss was from 36.0 yards away. Well, in 2016, the average kick was from 37.7 yards away; the average successful kick was from 36.2 yards out — farther than the average miss in 1960! — while the average miss was from 46.2 yards away.

The graph below shows the average length of each field goal attempt, in blue, each field goal made, in orange, and each field goal miss, in gray. [click to continue…]


Lowery, Anderson, Andersen, and Stenerud In Four Charts

A couple of years ago, I wrote about the best field goal kickers in NFL history. That was a threepart series where I measured how accurate each field goal kicker has been after adjusting for era and distance. The result? Nick Lowery was, by a clear margin, the most valuable field goal kicker in NFL history. He made kicks at a rate nearly 10% higher than league average after adjusting for era and distance, an astonishing level of success considering his reputation hasn’t quite matched his production.

Today, I wanted to update that post and also provide a comparison of the four men generally considered in contention for the title of top field goal kicker in history: Jan Stenerud, the first pure placekicker to make the Hall of Fame, Morten Andersen, who became the second such Hall of Famer this year, Nick Lowery, my choice for the best kicker ever, and Gary Anderson, who had a long and distinguished career.

I used a simple methodology this time around to compare the four kickers: I catalogued all field goal attempts in NFL history into five yard ranges (i.e., 40-44, 45-49, 50-54, etc.). Then, I looked at the league average success rate that season and calculated the expected number of field goals an average kicker would be expected to make from that range. So if the league average rate on kicks from 40-44 yards was 75%, a kicker with 8 field goal attempts from that rage would be “expected” to make 6 of those attempts. Finally, I calculated how many field goals each kicker made above expectation, and then created the following four charts. So if a kicker made 7 out of 8, he would be at +1.0. I have coded particularly good outcomes in blue, and bad outcomes in red. Let’s get to it.

Jan Stenerud [click to continue…]


Week 3 Game Scripts (2017): Jaguars Ring the Ravens Bell

The Jaguars obliterated the Ravens in London in week 3: Jacksonville led 10-0 after the first quarter, 23-0 at halftime, and 37-0 entering the third quarter. The Jaguars led 44-0 with 3:30 minutes left, before the Ravens scored the final points of the game.

It was the best Game Script of the season, thanks to both an incredible defensive performance Baltimore’s first 11 drives ended with three turnovers, seven punts, and one turnover on downs, and averaged a total of just 10.5 yards per drive! The offense’s first ten drives resulted in five touchdowns, three field goals, and two punts, and averaged 40 yards per drive (which would jump to 48 yards/drive if you eliminated the drives that began in Ravens territory and resulted in a touchdown).

The full week 3 Game Scripts below: [click to continue…]

Previous Posts