Let’s start with the SRS ratings for every team in the NFL. The SRS ratings are generated based off of the points scored, points allowed, home field, and opponent for each game. In its simplest form, the SRS is just an SOS-adjusted version of points differential, although the devil is in the details. After running hundreds of iterations to get the ratings to converge (and awarding 3 points to the home team), below are the ratings through week 8: [click to continue…]
Against Indianapolis in week 8, Ben Roethlisberger was close to perfect. He completed 40 of 49 passes for 522 yards. He threw six touchdowns, and didn’t throw an interception or take a sack. That’s a magnificent performance: in fact, among players with an 80% completion percentage in a game, he set a record for completions. It goes without saying that 500+ yard games are rare, and 6+ TD games are rare, and the combination of both are really rare.
But was it the best passing game ever? Not so fast. Let’s start by calculating his Adjusted Net Yards per Attempt, which gives a 20-yard bonus for touchdown passes, a 45-yard penalty for interceptions, and deducts sack yardage from the numerator (and adds sacks to the denominator). Roethlisberger averaged 13.10 ANY/A, a sparkling number. That’s an outstanding number that needs no qualifier, but it’s even more impressive when you consider the opponent. Entering the day, the Colts were allowing just 5.52 ANY/A to opposing passers.
Therefore, the Steelers star averaged 7.58 more ANY/A against the Colts than the average passer in 2014. Over the course of 49 dropbacks, this means Roethlisberger produced a whopping 372 Adjusted Net Yards above average, with average being defined as what all other passers did against Indianapolis.
That number may not mean much in the abstract. But if the Colts defense continues to allow just 5.52 ANY/A to all other passers year, that would give Roethlisberger the 7th best passing game since 1960. [click to continue…]
In early September, Adam Steele, a longtime reader and commenter known by the username “Red” introduced us to his concept of Marginal Yards after the Catch. Today is Part II to that post. Adam lives in Superior, Colorado and enjoys digging beneath quarterback narratives to discover the truth; hey, who can blame him?
Introducing Marginal Air Yards
There are three components of Y/A: Completion %, Air Yards/Completion, and YAC/Completion. In my last post I looked at YAC, so today, let’s look at the other two components. By multiplying completion percentage and air yards per completion, we get air yards per attempt, which we can then modify to create Marginal Air Yards (mAir):
mAir = (Air Yards/Attempt – LgAvg Air Yards/Attempt)*Attempts
Here are the yearly Air Yard rates since 1992, with the table sorted by Air Yards per Attempt:: [click to continue…]
With just under five minutes left in last Sunday’s game against the Giants and his team trailing 27-20, Mike Smith went for it on 4th and 1 from his own 29 yard line. As was the case on repeated 4th down attempts the last time his team visited MetLife Stadium to face the Giants, the decision to be aggressive did not work out well. Matt Ryan was sacked for a nine-yard loss that effectively ended the game. If his previous behavior is any guide, Smith may learn the wrong lesson from that outcome and choose not to go for it again when the next similar opportunity arises. Smith illustrates better than any other coach the potential for fourth down failure to lead to future fourth down timidity.
Before those two failed Ryan fourth down sneaks against the Giants in that 2011 playoff game, Smith actually was one of the more enlightened coaches on fourth down strategy. From 2008-2011, Smith was the third-most aggressive coach of the last twenty years, at least according to Football Outsiders’ Aggressiveness Index. Dating Smith’s turning point is a little tough. He got burned going for it in Week 10 of the 2011 regular season, when he tried a sneak on 4th and inches from his own 29 in overtime against the Saints. He punted in a couple of situations where he usually went for it late in the 2011 season, but then was aggressive closer in against the Giants. By the 2012 regular season, Smith hadn’t just abandoned his prior tendency for aggressive strategy. He entirely reversed it. In 2012, he was the least aggressive coach in football, only going for it once in 91 qualifying fourth-down tries. He was similarly passive in 2013. His fourth down decision last Sunday was surprising given that trend.
To see Smith’s evolution on fourth down strategy, consider his decisions on 4th and 3 or less when between the opponent’s 10- and 40-yard lines. To consider only situations where there was a real choice while keeping things as simple as possible, I look only at first-half decisions along with third-quarter decisions where the margin was ten points or less. [click to continue…]
Allow me to present to you Atlanta running back Antone Smith’s 2014 play-by-play log in its entirety:
|Week 1 vs. NO|
|2||11:40||0 - 13||2nd-and-10||own 20||rushed for 2 yards|
|2||09:16||0 - 13||1st-and-10||opp 31||rushed for 5 yards|
|3||00:33||17 - 20||2nd-and-9||own 46||caught pass for 54 yards TOUCHDOWN|
|Week 2 vs. CIN|
|2||14:49||41701||1st-and-10||own 28||caught pass for 4 yards|
|2||01:14||41708||3rd-and-4||own 38||target of incomplete pass|
|4||10:23||41722||1st-and-10||opp 35||caught pass for 15 yards (first down)|
|4||00:54||41936||1st-and-10||own 41||target of incomplete pass|
|Week 3 vs. TB|
|1||04:21||36708||1st-and-9||opp 9||rushed for 4 yards|
|2||09:00||28 - 0||1st-and-10||opp 11||rushed for 10 yards (first down)|
|3||02:36||49 - 0||1st-and-10||opp 36||rushed for -2 yards|
|3||01:59||49 - 0||2nd-and-12||opp 38||rushed for 38 yards TOUCHDOWN|
|Week 4 vs. MIN|
|1||05:23||0 - 7||3rd-and-2||opp 29||rushed for 2 yards (first down)|
|1||04:47||0 - 7||1st-and-10||opp 27||rushed for 3 yards|
|2||14:55||41834||2nd-and-10||own 31||rushed for 9 yards|
|3||01:40||21 - 27||1st-and-10||opp 48||rushed for 48 yards TOUCHDOWN|
|Week 5 vs. NYG|
|1||03:42||0 - 7||1st-and-10||opp 23||rushed for 2 yards|
|2||14:59||41827||3rd-and-4||opp 4||caught pass for 1 yards|
|2||12:33||41919||1st-and-10||own 25||caught pass for 8 yards|
|3||05:51||13 - 10||3rd-and-4||own 26||caught pass for 74 yards TOUCHDOWN|
That’s four long touchdowns on 17 offensive touches. On his four scoring plays, Smith has gained an incredible 214 yards. That’s the most in the NFL so far, with Steve Smith (162 yards) and Jordy Nelson (160) rounding out the top three. Perhaps even more incredible is that Smith has gained 214 yards on scoring plays despite gaining only 63 yards on non-scoring plays. Here’s a chart I tweeted a couple of days ago, showing yards gained on TDs on the X-axis and yards gained on all other plays on the Y-axis: [click to continue…]
In the third quarter on Monday night, I texted my Patriots fan buddy Matt, “Is it possible that we suck? Maybe the run is finally over.” Bill Barnwell mused on this, and Aaron Schatz also wrote about it. It was hard not to think that, given the way the Patriots were manhandled by a mediocre team playing without several key players. It looked every bit as bad as the 41-14 score and maybe worse.
I remember the last time I wondered if the Pats were done. In a 34-14 loss to the Browns in 2010, the Patriots looked pretty impotent. In that game, as in the Chiefs one, the Pats had just under 300 yards of offense. Peyton Hillis ran over the Patriots. Of course, that wasn’t the end. Maybe this time is different, though. If anything the Chiefs game was even worse, so it’s possible this time really is the end.1
Will the Patriots offense be good later this year? To provide a little insight into this, I went back and looked at performance trends for quarterbacks who have had long careers. The first table looks at quarterbacks since 1969 who have the biggest single-season drops in adjusted net yards per attempt (ANY/A) from the previous five year trend. I look just at quarterbacks with at least 100 attempts in a season and I weight by the number of attempts when calculating the average ANY/A over the previous five years.
- And those Pats were 6-1 at the time of the loss to the Browns. [↩]
I am getting some well-deserved crap from people about just how bad my predictions have been so far. The Arizona Cardinals have already somehow outperformed the number of wins I gave them. The Jacksonville Jaguars, my pick to win the AFC South at 8-8, at one point in the game against the Colts had been outscored 112-13 over a stretch of about nine quarters. And my pick to win the NFC North at 14-2 could be 0-3 if Marty Mornhinweg let his head coach call the timeouts.1
But I did win my first Stone-Cold Mega-Lock of the Week with my very comfortable tease of the Bengals and Falcons. So things are looking up and I’m taking that as license to check out some historical betting data for anything that might seem appealing after three weeks.
Last year’s Carolina Panthers are the inspiration for the analysis here. After three weeks, they were a 1-2 team with a big positive point differential. The Panthers last year lost 12-7 to Seattle and 24-23 to Buffalo before annihilating the Giants 38-0. Despite VOA liking the Panthers even after just three games, the betting market came around later in at least one way. The Panthers were at 3/1 to make the playoffs last year after three weeks, even though Football Outsiders had their playoff odds at over 50% at that time.
Is it possible that teams like the 2013 Panthers have historically been undervalued? It seems likely that Carolina was a little undervalued last year after three weeks. By looking at point spread data, we can see if teams that have likely been better than their records have been good bets in the early part of the season. Specifically, I’m going to look at whether betting on early-season underachievers (teams with deceptively poor records) or against overachievers has been profitable now and in the past.
Data and Methods
Feel free to skip this part, but here’s the background for those interested. I have put together Pro Football Reference’s point spread data for all games from 1979 to 2012. This sample is good enough for the tests of long-term and recent betting strategies that I want to do.
I’m going to look at betting outcomes in games 4-8 for teams that are either losing teams (winning percentage below 0.5) with strong Pythagorean records or winning teams with weak Pythagorean records. I will keep things simple and define Pythagorean wins here as:
Pythagorean Wins = (Previous Points Scored ^2.53)/(Previous Points Scored^2.53 + Previous Points Allowed^2.53)
In a continuing effort to avoid unnecessary complications, I’m just going to split the data up over time, looking separately at results before and after 2000.
Betting On and Against Pythagorean Outliers
Below is how you would have done over time if you bet on or against two kinds of teams:
- Overachievers: Teams with winning records with bad point differentials for their records
- Underachievers: Teams with losing records with good point differentials for their records
An overachiever is more specifically a team with a winning record that has a Pythagorean winning percentage at least 25 percentage points worse than their actual winning percentage. An underachiever has a Pythagorean winning percentage at least 25 percentage points better than actual.
|1979-1999||174-142-11 (55.1%)||141-146-5 (50.9%)|
|2000-2012||109-99-4 (52.4%)||108-100-8 (52.0%)|
The results show that, before 2000, you would have won most of the time betting on overachieving teams, teams that were not as good as their records would suggest. I was surprised by that and it even made me wonder if I made a coding mistake. I certainly expected that any tendency away from an even split would have been in favor of betting against teams with good records and relatively poor point differentials. Note that the even split occurred in the past for the underachievers, the teams with good point differentials and poor records.
More recently, the data come pretty close to an even split for betting both on the overachievers and the underachievers. Betting on the overachievers and the underachievers has been successful about 52% of the time since 1999.
So the overall message is that there is little value now or in the past in identifying Pythagorean outliers and either riding the teams with deceptively poor records or fading the teams with misleadingly good ones. In fact, the only pattern from the past suggested it was a good idea to ride the teams with misleadingly good records. I tried to check this out a bit to just see if it was just betting on teams with good records that was profitable, but betting on all teams with winning percentages over .750 has gone almost exactly dead even over time. It would be great to hear any thoughts you might have in the comments for this pattern. I feel like I’m missing something.
Overall, the message here is the one that we get most of the time if we try to find patterns that might lead to a consistently profitable and simple betting strategy. It just ain’t there. That doesn’t make this a bad post, though: as Chase once noted, an answer of “not useful” is often just as meaningful as any other answers.
The Stone-Cold (I Think There May Be a 60% Chance This Bet Will Win) Mega-Lock of the Week
So I am now 33% on my Stone-Cold Mega Locks of the Week. If I get the next two, I will be at 60%. If I get the next two after that, I’ll be at 71%. I kind of think I should be able to claim extra points already, Chris Berman-style, for my tease last week, since the Falcons and Bengals won by a combined 89-21 score that wasn’t that close. But I will instead put my faith in the always reliable larger sample size that will bear out these predictions living up to their title.2
Two-team teaser: Pittsburgh down to -1.5 and Indianapolis down to -1.5
This week, I like another two-team teaser of two home teams, this time down to 1.5 points. I particularly like the Steelers down to 1.5 points. I do not understand how they could be the same offense for quarters 3-8 of this season as they were for the other high-efficiency ones. Still, I like the Steelers (#10 in DVOA) at home against the Buccaneers (#32).
I’m a little less sure about the other side of the tease, where I have Indianapolis (#21) over Tennessee (#25). In fact, I mainly just wanted to get the Pittsburgh end of the tease. I may be getting that queasy-knees feeling come Sunday. It’s hard to feel that way about Andrew Luck, but I didn’t imagine I’d ever be going into the water tethered to a Ryan Grigson-led team.
Season record: 1-2
Against Washington in week 3, the Eagles fell behind 17-7 before coming from behind and again emerging victorious. As a result — and after trailing the Jags 17-0 and the Colts 20-6 — Philadelphia became the first team since at least 1940 to start a season 3-0 despite trailing by at least 10 points in each game.
In fact, only three teams had ever overcome a deficit of a touchdown or greater in each of their first three games: the 2000 Rams, the 2000 Jets, and the 1960 Giants. Those teams finished the season 10-6, 9-7, and 6-4-2 respectively, which means they went just 16-17-2 the rest of the season after starting 9-0.
In general, teams that have started 3-0 despite constantly falling behind have not been as successful over the rest of the season as other 3-0 teams. In fact, if you add up the worst margin for each 3-0 team in each game, 25 teams have trailed by an “aggregate” of 21+ points in those three games. On average, those teams won just 53.5% of the remainder of their games. [click to continue…]
Just above these words, it says “posted by Chase.” And it was literally posted by Chase, but the words below the line belong to Bryan Frye, a longtime reader and commenter who has agreed to write this guest post for us. And I thank him for it. Bryan lives in Yorktown, Virginia, and operates his own great site at nflsgreatest.co.nf, where he focuses on NFL stats and history.
In August, I introduced a concept on my site to better adjust the NFL’s passer rating for the league passing environment. I love Pro Football Reference’s use of the Advanced Passing Index for passer rating (Rate+), but it still bothered me that the internal math of the NFL’s formula remained the same.
The NFL’s official passer rating formula is based on four variables: completion percentage, yards per attempt, touchdown percentage, and interception rate. Each of those variables are then used to determine four different variables, as seen below:
A = (Cmp% – .3) * 5
B = (Y/A – 3) * .25
C = TD% * 20
D = 2.375 – Int% * 25
Passer rating is then calculated as follows, provided that each variable is capped at 2.375 and has a floor of zero:
(A + B + C + D)/(0.06)
For each component, a score of 1 represents the ideal average passer. Because the formula is based on a league average completion rate of 50%, modern passers significantly exceed that; pre-modern passers rarely reached it. Similarly, the NFL’s model is based on a 5.5% interception rate and a 5% touchdown rate. Thanks to a Greg Cook injury (and Bill Walsh’s genius reaction to it), those numbers have also changed significantly. Last year, the league interception and touchdown rates were 2.8% and 4.4%, respectively. [click to continue…]
Since 1940, there have been 616 times where a team rushed for at least 125 yards and completed at least 75% of its passes. On Sunday, when Washington pulled off that feat against the Texans, they became the first team to fail to score double digit points in the process.
In the second half, both RG3 and Niles Paul lost fumbles inside the Houston 10-yard line; that obviously contributed to the team failing to score more than 6 points. But Griffin’s 78.4% completion percentage was also pretty misleading. Griffin’s average throw went just 5.8 yards in the air, and his average completion covered just 3.9 yards before including his receiver’s yards gained after the catch. Both of those averages put ranked 30th among 32 qualifying passers. But while short throws can be part of an effective offense, on Sunday, that wasn’t the case for Washington. Consider:
- A 4th and 10 completion to Roy Helu for 6 yards
- A 3rd and 16 completion (on the Washington 15) to Helu for 9 yards
- A 3rd and 13 completion to DeSean Jackson for 0 yards
- A 2nd and 25 completion to Jackson for 0 yards
- A 2nd and 19 completion to Pierre Garcon for 3 yards
- A 2nd and 14 completion to Logal Paulsen for -3 yards
- A 2nd and 8 completion to Garcon for 3 yards
- A 2nd and 1 completion to Jackson for 0 yards
- Four 1st and 10 completions to Jordan Reed, Paulsen, Paul, and Darrel Young for 4, 3, 2, and 1 yard(s), respectively.
Sure, Griffin completed 29 of his 37 passes, but 12 of his completions did little or nothing to help his offense. He also was sacked three times. As a result, just 17 of his 40 dropbacks — or 42.5% — were successful completions.
To be fair, this isn’t as much a knock of Griffin as the Washington offense as a whole, or perhaps just a counter to those who like to rely on completion percentage or its brother, passer rating. If Griffin’s targets could have gained more yards after the catch, things would have looked a lot different. And against the frightening pass rush of J.J. Watt and company,1 short passes make some sense. But looking at Griffin’s completion percentage and concluding he had a good game is kind of silly. Again, more a knock on the misuse of statistics than the player.
Football Outsiders considers a completion that fails to gain a first down on 3rd or 4th down, a completion that fails to gain at least 60% of the distance needed on 2nd down, or a completion that fails to gain at least 45% of the needed yards on 1st down to all be failed completions. Those cut-offs seem reasonable enough to use for theses purposes. Looking at the numbers, Griffin led the NFL in failed completions in week one.
Here’s how to read the table below. In week 1, Griffin completed 29 of 37 passes, producing a completion percentage of 78.4%. However, 12 of his completions were failed completions, as identified above. That means 41.4% of his completions were failed completions. He also took 3 sacks; as a result, just 42.5% of his dropbacks were successful completions. The difference between his raw completion percentage and his SCmp/DB average was 35.9%. [click to continue…]
Is week 1 a window into a team’s soul? Or is week 1 best left ignored by analysts, since results are skewed by teams that are still shaking off the rust from the summer? As it turns out, week 1 isn’t just like any other week: it’s more like any other week than, uh, any other week. What do I mean by that?
Let’s begin with a hypothesis. The best teams in the league are [more/less] likely to win in week 1 than they are normally. This is because the best teams are [at their best/rusty] in week 1. How would we go about proving this to be true?
One method would be to take a weighted average winning percentage of teams in week one, with the weight being on the team’s actual season-ending winning percentage. For example, the Patriots went 16-0 in 2007, which means New England was responsible for 6.25% of all wins in the NFL that season. That year, the Colts went 13-3, so Indianapolis was responsible for 5.1% of all wins that year. If we want to know whether good teams play [better/worse] in week 1, we care a lot more about how teams like the ’07 Patriots and Colts fared than the average team.
By using weighted average winning percentages, we place more weight on the results of the best teams, which is exactly what we want to do. So when the ’07 Patriots and ’07 Colts won in week one, rather than being responsible for 6.25% of the league, they are now are responsible for over 11% of the NFL’s weighted week 1 winning percentage. Of course, you can probably figure out pretty quickly that by using this methodology, we are ensuring that the “average” winning percentage over the course of the season will be quite a bit over .500, since the best teams will win more often than not. And that’s exactly what we see: the average weighted winning percentage across all weeks, using this methodology, was 0.574. As it turns out, that’s exactly what the average is in week 1, too. [click to continue…]
Just above these words, it says “posted by Chase.” And it was literally posted by Chase, but the words below the line belong to Adam Steele, a longtime reader and commenter known by the username “Red”. And I thank him for it. Adam lives in Superior, Colorado and enjoys digging beneath quarterback narratives to discover the truth; hey, who can blame him? One other house-keeping note: I normally provide guest posters with a chance to review my edits prior to posting. But due to time constraints (hey, projecting every quarterback in the NFL wasn’t going to write itself!), I wasn’t able to engage in the usual back and forth discussion with Adam that I’ve done with other guest posters. As a result, I’m apologizing in advance if Adam thinks my edits have changed the intent of his words. But in any event, sit back and get ready to read a very fun post on yards after the catch. When I envisioned guest submissions coming along, stuff like this is exactly what I had in mind.
Introducing Marginal YAC
A quarterback throws a two yard dump off pass to his running back, who proceeds to juke a couple defenders and run 78 yards into the endzone. Naturally, the quarterback deserves credit for an 80 yard pass. Wait, what? Sounds illogical, but that’s the way the NFL has been keeping records since 1932, when it first began recording individual player yardage totals. The inclusion of YAC — yards after the catch — in a quarterback’s passing yards total can really distort efficiency stats, which in turn may distort the way he is perceived.
In response, I created a metric called Marginal YAC (mYAC), which measures how much YAC a quarterback has benefited from compared to an average passer. Its calculation is very straightforward:
mYAC = (YAC/completion – LgAvg YAC/completion) * Completions
I have quarterback YAC data going back to 1992 for every quarterback season with at least 100 pass attempts.1 That gives us a healthy sample of 965 seasons to analyze, and includes the full careers of every contemporary quarterback. But first, let’s get a sense of what’s average here. The table below shows the league-wide YAC rates since 1992: [click to continue…]
- This data comes courtesy of sportingcharts.com. It’s obviously unofficial, but there doesn’t seem to be any noticeable biases from one team to another. Some unofficial stats, such as passes defensed or quarterback pressures, can vary wildly depending on the scorekeeper, but Sporting Charts’ YAC stats seem pretty fair, from what I can tell. Here is a link to the 2013 data. Chase note: I have not had the chance to compare these numbers to what is on NFLGSIS, but that’s a good idea. [↩]
Below are my 2014 projected quarterback rankings. Let me be very clear at the top of this post as to exactly what these rankings mean: they represent my projections of the order in which these quarterbacks will finish in my preferred measure of quarterback play. Everyone has their own measuring sticks when it comes to quarterbacks; for me, it’s Adjusted Net Yards provided above league-average. As a reminder, here is how we calculate that metric.
First, we start with Adjusted Net Yards per Attempt, which is calculated as follows:
(Passing Yards + 20 * PassTDs – 45 * INTs – Sack Yards Lost) / (Pass Attempts + Sacks)
Then, we take each quarterback’s ANY/A average, and subtract from that number the league average ANY/A metric, which should be around 5.9 ANY/A. Then, we multiply that difference by the quarterback’s number of dropbacks.
Last year, Peyton Manning led the league in this category, with 2,037 Adjusted Net Yards of value provided above average. The benefit to this approach to ranking passers is that the results are easy to test. At the end of the season, we can calculate the actual results, and then look back and laugh at this post.
So, ranking 1-32, here is how I project the top quarterback for each team to finish in 2014.1) Peyton Manning, Denver Broncos
There’s a reason Manning is the heavy favorite to repeat as NFL MVP. The Broncos lost Eric Decker and Knowshon Moreno, and Wes Welker’s concussion concerns only worsened this preseason. No matter: Manning remains the gold standard. Denver added Emmanuel Sanders in the offseason, and he caught five passes for 128 yards with two touchdowns against Houston in the preseason. Manning has led the NFL in sack rate in three of his last four seasons, and the return of Ryan Clady should make Manning even more difficult to sack in 2014. No need to over think this one: Manning is the clear favorite to again provide the most value of any quarterback in the league.
2) Aaron Rodgers, Green Bay Packers
3) Drew Brees, New Orleans Saints
Choosing between Brees and Rodgers is tough, but the return of a healthy Randall Cobb and the departure of Darren Sproles is enough to tip the scales towards Rodgers for me. Green Bay tends to forget about the little things — Corey Linsley, a fourth round pick, will be the team’s starting center — but Rodgers has a way of curing all ills. Brees turns 36 in January, which is yet another reason to break ties in favor of Rodgers. Since ’09, Rodgers is the league-leader in ANY/A, while over that period, Brees has thrown the most touchdowns and gained the most yards. If Manning isn’t the king in 2014, it’s a good bet that either Rodgers or Brees took the crown. [click to continue…]
In 1995, Football Outsiders graded the Eagles special teams as the worst in the NFL. The next two years, Philadelphia ranked 20th and 26th, respectively. In 1998, after hiring a new special teams coordinator, the team still finished just 25th. But, over the next eight years, the Eagles’ special teams flipped dramatically, ranking as the second-best in football during that period. In fact, from 2000-2004, Philadelphia ranked in the top five in the Football Outsiders’ special teams ratings each season.
When the Ravens hired the coordinator of those special teams, John Harbaugh, as their head coach in 2008, Baltimore turned one of the more surprising coaching hires in recent history into one of the best. Based on where the team was when it hired him, Harbaugh’s first three years were about the best since 1990 of any coach not named Harbaugh, at least according to DVOA. The Ravens made the playoffs in Harbaugh’s first five seasons, winning the Super Bowl in the last of those. Harbaugh’s success even caused Chase to wonder whether it would change the way teams hired head coaches.
Since Harbaugh was so successful as a coordinator, does that mean he was a good bet to be a successful head coach? At first glance, you might think just about every coordinator who gets promoted or poached to become a head coach was very successful in his previous job. As it turns out, that’s not always the case. Once we correct for expectations, a little more than one in four hired head coaches actually underperformed in their previous jobs, at least according to DVOA.
Consider one man who performed particularly poorly as a coordinator: Eric Mangini. The 2005 New England defense had a DVOA that was 15.2 points lower than we would have predicted based on the Patriots’ performance in the preceding seasons. He was not so much of a (Man)genius to have a good defense in 2005, and that may have given some hint that he was not the greatest bet to succeed as a head coach, either.1
This leads to an obvious question: on average, have teams done better when they have hired head coaches who were actually good in their previous jobs (either as coordinators or head coaches)? Let’s take this to the data. [click to continue…]
The AFC East was a very stable division over the past two years. The Patriots won 12 games in 2012 and 12 more in 2013. The Bills, with six wins in 2013, also repeated their 2012 win total. Miami won 7 games in 2012, and then 8 last year. And the Jets followed up a 6-10 season in 2012 with an 8-8 season last year. That’s about as stable as a division can get. The four teams saw their win totals move by an aggregate of just three wins, making the 2012-2013 AFC East the most stable division since realignment.
On the other end of the spectrum: the NFC South. The Falcons dropped from 13 wins in 2012 to just four last year. The Panthers jumped from 7 wins in 2012 to 12 last year, and it didn’t even take Bill Parcells to do it. New Orleans also won seven games in 2012, but jumped to 12 wins in 2013. The team that saw the least movement in the NFC South last year was Tampa Bay, but the Bucs still fell from 7 wins to 4 wins, matching the total movement by all AFC East teams. As a group, NFC South teams had a change of 21 wins from 2012 to 2013, the most of any division since realignment.
That’s hardly new for the NFC South, or for that matter, the AFC East. Since realignment, the NFC South has easily been the league’s most unstable division: the Falcons, Saints, Bucs, and Panthers have seen their win totals fluctuate by an average total of 18.8 wins per year, beginning with the 2002-2003 seasons. The AFC East has been incredibly stable: no team has ever finished with more wins than New England, while the Bills have finished last or tied for last eight times since realignment. As a result, the average movement among AFC East teams — in the aggregate — has been just 6.3 wins.
Change in Wins/Yr
Change # Wins/Tm Yr
Kansas City ranked 4th in Adjusted Net Yards per Attempt in 2002, 1st in 2003, 3rd in 2004, and 2nd in 2005. In terms of Adjusted Yards per Carry, the Chiefs were 2nd in 2002, 3rd in 2003, 1st in 2004, and 3rd in 2005. That’s an incredible streak of not just dominance, but balanced dominance. And Kansas City missed the playoffs in three of those four years! (pours one out for Jason Lisk).
On Monday, we looked at some great defenses that missed the playoffs. Today, a look at some of the best offenses to stay home for the winter. And in the last 15 years, the 2002 Chiefs, 2004 Chiefs, and 2005 Chiefs are the only teams to rank in the top five in both ANY/A and AYPC and miss the playoffs.
What other teams since the merger met those criteria? [click to continue…]
Just above these words, it says “posted by Chase.” And it was literally posted by Chase, but the words below the line belong to Bryan Frye, a longtime reader and commenter who has agreed to write this guest post for us. And I thank him for it. Bryan lives in Yorktown, Virginia, and operates his own great site at nflsgreatest.co.nf, where he focuses on NFL stats and history.
In February, Chase used a regressed version of Football Outsiders’ DVOA metric to derive 2014 expected wins. If you are reading this site, you probably have some familiarity with Football Outsiders and DVOA, FO’s main efficiency statistic. Given the granularity of DVOA, it is no surprise that Year N DVOA correlates more strongly with Year N + 1 wins (correlation coefficient of .39) than Year N wins does (correlation coefficient of .32).
By now, even casual NFL fans probably have at least heard of Pythagorean wins, and regular readers of this site are certainly familiar with the concept. Typically, an analyst uses Pythagorean records to see which teams overachieved and underachieved, which can help us predict next year’s sleepers and paper tigers. Well, I wondered what would happen if we combined the two formulae to make a “DVOA-adjusted Pythagorean Expectation” (or something cooler sounding; you be the judge).
Going back to 1989, the earliest year for DVOA, I used the offensive, defensive, and special teams components of DVOA to adjust the normal input for Pythagorean wins (points). Because DVOA is measured as a percentage, I adjusted the league average points per team game accordingly (I split special teams DVOA between offense and defense). Let’s use Seattle, which led the league in DVOA in 2013, as an example.
In 2013, the league average points per game was 23.4. Last year, Seattle had an offensive DVOA of 9.4% and a defensive DVOA of -25.9% (in Football Outsiders’ world, a negative DVOA is better for defenses). The Seahawks also had a special teams DVOA of 4.7%. So to calculate Seattle’s DVOA-adjusted points per game average, we would use the following formula:
23.4 + [23.4 * (9.4% + 4.7%/2)] = 26.15 DVOA-adjusted PPG scored
And to calculate the team’s DVOA-adjusted PPG allowed average, we would perform the following calculation: [click to continue…]
Over at Five Thirty Eight, I look at whether the Broncos pass offense, or the Seahawks pass defense, is more immune from regression to the mean.
As a general rule, elite offenses are further from league average than great defenses, so offensive regression isn’t as likely as defensive regression. It helps, too, that research has shown offenses to be more consistent from year to year than defenses. All else being equal, we would expect the Broncos to be the more likely team to repeat last year’s brilliant performance.But all else isn’t equal. Denver produced 2013’s record-breaking numbers while playing defenses from the AFC South and the NFC East; those will be replaced this year by the AFC East and the NFC West, divisions that present much more formidable challenges. That’s a significant change.
According to Football Outsiders, Denver played the third-easiest slate of opposing defenses in 2013. Based purely on adjusted net yards per attempt, the average defense Manning faced last year was 0.44 ANY/A below average, and that’s after adjusting those defenses’ ratings for the fact that they played Manning. Only Alex Smith and Robert Griffin III faced more cupcakes. Last year, Manning didn’t Omaha against a single defense that ranked in the top eight in strength-of-schedule ANY/A; this year, he’s set to face six opponents that ranked in the top eight in that metric in 2013.
You can read the full article here.
Neil Paine showed some interesting evidence relating to this idea on Friday. Looking at team performance since 2009 for teams with new quarterbacks, Neil showed that preseason passing efficiency helps predict regular season passing efficiency. It’s important to note that part of this result may have been pretty predictable even before we watched those preseason games. The 2012 Redskins replaced Rex Grossman and John Beck with the #2 pick in the draft who would have been #1 in an average year. So we would expect a big improvement to come just by way of moving from Grossman to a healthy RGIII. [click to continue…]
In the Super Bowl era, there has been just one team that was both the youngest in the league and one of the five best teams in football: the 2012 Seattle Seahawks. As friend of Football Perspective Neil Paine recently pointed out, being young and great has historically been a good predictor of teams that have become dynasties. Consider the table below. It captures every team since 1966 that ranked amongst the five youngest teams by Approximate Value (AV)-weighted age and had at least 12 Pythagenpat wins, adjusting everything to a 16-game schedule.1
There are seven unique teams on this list, not counting the two repeaters. When trying to predict what’s going to happen with the Seahawks, there are two different ways to look at this list. The first looks good for their dynasty potential. The first two teams on the list, the ’72 Steelers and the ’92 Cowboys went on to win multiple Super Bowls. The closest comparison in terms of age also looks pretty good. Teams used to be younger, so the best comparison probably isn’t the ’72 Steelers, who were even younger by age but were only the fifth-youngest team in 1972, but the ’92-’93 Cowboys. They are the only other team on this list to be so young and so good.
Of course, even the Cowboys had a pretty short run. Their stay at the top was nothing like the ’70s Steelers or ’80s Niners, who were also quite young.2 Free agency helped to minimize their time on top. The ’90s Cowboys were the first great team in the free agency era. Players gained full freedom of movement only in the year after their first Super Bowl. Plan B free agency allowed limited movement starting in 1989.
Free agency and the salary cap help to explain the path of the other four teams on the list. They point towards a more cautious prediction about the Seahawks’ dynasty hopes. Between them, the ’99 Rams, ’01 Bears, ’06 Chargers, and ’07 Colts won one Super Bowl and played in two others. Within three years of their great-and-young season, only the Chargers were significantly better than league-average.
These more recent examples may do a better job of predicting the Seahawks future success. Before the beginning of full free agency in 1993, good-and-relatively-young teams appear to have generally followed a clear and sustained upwards trajectory over the long term. Since then, however, success has generally been less sustainable. The table below looks at teams’ strengths over time according to PFR’s Simple Rating System.3 Here I’ve made the cutoff any team that was in the five youngest teams in a given year and also had a SRS rating of at least 6. The table shows the trend in strength for the previous season and the following three seasons.
SRS Wins (t+3)
One surprising pattern in these data is just how infrequently young teams won in the past. From 1966-1992, only five teams were among the five youngest and still had an SRS of at least 6. Since 1993, it’s happened 16 times. In the past, teams had more of an opportunity to gradually build strength. So it looks like there was a greater share of young teams building for something and old teams trying to stay on top. Since 1993, the standard deviation of team ages is about 20% smaller than it was before that. In the last ten years, the standard deviation is about 30% smaller than it was before 1993. The ages of rosters are more compressed than they used to be.
The other thing to take away from these tables is the dropoff in years 2 and 3 since full free agency. For the pre-1993 teams, the good-and-young teams held much of their value. After starting at an average SRS of 8.96, they were still at 7.04 two years later and then 5.3 three years later. Since 1993, teams have deteriorated more quickly. From an average of 9.09, the more recent high quality young teams fell to 4.95 two years later and all the way to 2.09 three years later.
Since there are only five teams in the pre-1993 group, we want to be careful with interpreting too much into the earlier data. It’s possible that the ’72 Steelers and ’81 Niners are anomalies. At the same time, the success three years later is skewed downwards by the ’75 Colts, who would have been much stronger in ’78 if they had a healthy Bert Jones.
With the bigger set of more recent teams, the clear takeaway is that in the current era, even very good and young teams are just slightly better than average than three years later. The Seahawks may buck this trend, but they probably won’t. With Russell Wilson to sign and long-term cap hits for players like Richard Sherman and Earl Thomas, they’re more likely to have a brief run than a long one.
Another alternative may be available, though. If Wilson makes the leap into the Brady-Manning class (he may) and Pete Carroll turns out to be a truly elite coach (also possible), they may be able to fashion a New England-kind of dynasty. That sort of dynasty is not really built on youth. Consider the aging patterns of the last five teams of the decade.
The ‘60s Packers, ‘70s Steelers, ’80s Niners, ‘90s Cowboys all showed the same pattern of being relatively young and then progressively aging during their runs. On the other hand, the Patriots show an entirely different pattern. They’re the only dynasty to actually not age as their run progressed. They started old and stayed old through their Super Bowl years. While the Seahawks are starting off younger than those Patriots teams, excellence at QB and coach still offers them their best hope of building a dynasty in the current NFL. The benefits of being young and good are much more fleeting than they used to be.
- My AV-weighted age calculations are very similar to Chase’s, but not always exactly the same. For example, I have Seattle third in 2013, while he has them second. We both had Seattle at 26 years, but I have Cleveland also at 26, instead of 26.1. [↩]
- They were the third-youngest team in 1981, their first championship year. [↩]
- I thank Bryan Frye for sharing his SRS dataset. [↩]
Yesterday, we looked at which quarterbacks were the best at yards per completion after adjusting for league average. Today, we’ll do the same thing for wide receivers and yards per completion.A small tweak is necessary to the formula. You can skip down to the results section if you don’t care about the math, but I suppose most of my readers want to know what goes in the sausage. We can’t just use league-wide yards per completion rates, since that average includes receptions by non-wide receivers. One way around this is to calculate the league average YPC for wide receivers only; that’s easy to do for 2013, but less easy to do for the earlier years of NFL history when the distinction among the positions was not so clear. So, after playing around with a few different methods, I’ve decided to instead use 120% of the league average YPC rate, and give wide receivers credit for their yards over expectation using that inflated number.
For example, in 1983, James Lofton caught 58 passes for 1,300 yards for the Packers, a 22.4 YPC average. That year, the average reception went for 12.63 yards; 120% of that average is 15.2, which means we would give Lofton credit only for his yards over the product of 15.2 and 58, or 879. Since Lofton actually had 1,300 yards, he gets credit for 421 yards over expectation.
The next year, Lofton caught 62 passes for 1,361 yards (22.0). Since the average reception went for 12.66 yards, Lofton gets credit for his yards over (120% * 12.66 * 62), or 942. Lofton therefore is credited with 419 yards over expectation, nearly identical to his performance in the prior year. In fact, those were the 10th and 11th best season in NFL history by this method. [click to continue…]
In 2013, the average completion went for 11.63 yards. That’s a pretty low number historically, although it’s actually a bit higher than some of the recent NFL seasons. Take a look at how Yards per Completion has generally been declining throughout NFL history:
If you want to discuss the quarterbacks who excelled in this metric, controlling for era is crucial. One simple way to measure the best passers when it comes to YPC is to measure how they fare in this metric relative to league average, and multiply that difference by the player’s number of attempts. For example, Nick Foles averaged 14.2 YPC last year, which was 2.6 YPC above average. Over the course of his 317 pass attempts, we could say he provided 529 yards above the average completion. That was the highest in the NFL last year, while Matt Ryan produced the lowest average. [click to continue…]
Last week, I looked at the top receivers and the quarterbacks who threw it to them. Today, we flip that question around and look at which receivers the top quarterbacks threw to. I used the exact same methodology from the previous post, so please read that for the fine details.
For Peyton Manning, 20% of his career passing yards came via Marvin Harrison, and another 16% came from Reggie Wayne. Both of those numbers will decline the longer Manning plays, of course, but for now, those players dominate his list (Dallas Clark is third at seven percent). That’s a pretty stark departure from other quarterbacks such as say, I dunno, Tom Brady. For the Patriots signal caller, Wes Welker is his top man (13%), followed by Deion Branch (9%), Troy Brown (7%), Rob Gronkowski (7%), and then Randy Moss (5%).
The table below lists the top 7 receivers for each of the 200 quarterbacks with the most passing yards since 1960. The list is sorted by the quarterback’s career passing yards, and I have removed the percentage sign from the table to enable proper sorting. For example, here’s how to read Brett Favre’s line. He’s the career leader in passing yards, and played from 1992 to 2010. His top receiver was Donald Driver (9%), followed by Antonio Freeman (9%), Robert Brooks (6%), Sterling Sharpe (5%), Bill Schroeder (5%), Ahman Green (4%), and William Henderson. [click to continue…]
A couple of weeks ago, I posted a list of the best rushing teams in 2013 using Adjusted Yards per Carry. That metric, you may recall, is calculated as follows:
Rushing Yards + 20*RushingTDs + 9*RushingFirstDowns
We can use the same formula to grade every team across history. To account for era and quantity (having more above-average rush attempts is better), I calculated each team’s AdjYPC average, subtracted the league-average AdjYPC average, and multiplied that difference by the team’s number of rush attempts.
The top team by this method isn’t even an NFL team: it’s the 1948 San Francisco 49ers. You may recall that the 49ers and Browns staged two epic battles that season, and may have been the best two teams in pro football. That season, San Francisco averaged 6.1 yards per carry and rushed for 35 touchdowns on 600 carries; along with 152 first downs, and the 49ers averaged 9.5 Adjusted yards per Carry. That’s the highest average ever, just narrowly topping the production of the franchise’s Million Dollar Backfield six years later. Joe Perry was on both teams because Joe Perry was the man.
The table below shows the traditional rushing data for the top 200 rushing teams of all time; the VALUE column represents the number of Adjusted Rushing Yards produced above average (i.e., relative to the league average AdjYPC). I’ve also listed the three most prominent rushers on each team. [click to continue…]
Mendenhall ended the season with 687 yards on 217 yards, a 3.2 yards per carry average. Ellington finished his rookie year with 118 carries for 652 yards, producing 5.5 yards per rush. One way to measure the magnitude of the difference in the effectiveness of these two players — and boy was there a large difference — is to simply look at the delta in the players’ yards per carry averages. In this case, that’s 2.36 yards per carry.
Where does that rank historically? Some teams — I’m looking at the Lions in the early Barry Sanders years — gave only a handful of carries to their backup running backs. So one thing we can do is to take the difference in the yards per carry between the team’s top two running backs and multiply that number by the number of carries by the running back with the lower number of carries. In each instance, I’ve defined the running back with the most carries as the team’s RB1, and the running back with the second most carries as the RB2. In Arizona’s case, that would mean multiplying -2.36 (Mendenhall’s average, since he was the RB1, minus Ellington’s average) by 118, the number of carries Ellington recorded. That produces a value of -278. [click to continue…]
One of my first posts at Football Perspective was one of my favorites: the top receivers and the men who threw it to them. I like referencing that post from time to time, so I decided to update the numbers through the 2013 season.
I looked at all regular season games since 19601, and calculated the percentage of passing yards produced from each quarterback. Then, I assigned that percentage to the number of receiving yards for each receiver. For example, in this Raiders game from 1995, Vince Evans threw for 75% of the Raiders passing yards, and Jeff Hostetler was responsible for the other 25%. Therefore, since Tim Brown gained 161 yards, 121 of those yards are assigned to the “Brown-Evans” pairing and 40 to the “Brown-Hostetler” pairing. Do this for every game since 1960, and you can then assign the percentage of career receiving yards each receiver gained from each quarterback.
For example, 32% of Brown’s yards came from Rich Gannon, 26% from Hostetler, 12% from Jeff George, and 9% from Jay Schroeder. That breakdown isn’t too unique: in fact, of the six receivers with the most receiving yards since 1960, all six (including Brown) gained between 29% and 37% of their career receiving yards from their top quarterback.
The table below lists the top 7 quarterbacks for each receiver, although I only included quarterbacks who were responsible for at least five percentage of the receiver’s yards. It includes the 200 players with the most receiving yards since 1960. [click to continue…]
There is no doubt that in modern times, passing is king. But until pretty recently, we were at the peak level in NFL history with respect to individual rushing performance. On the team level, rushing production ebbed and flows, with high points in the late ’40s, mid-’50s, and mid-’70s, but on the individual level, the 2006 season may have been the high point.
That year, Pittsburgh’s Willie Parker rushed 337 times for 1,494 yards and scored 13 touchdowns, but Parker ranked just sixth in rushing yards. He also caught 31 passes for 222 yards, but Parker ranked only 7th in yards from scrimmage. That season, the average leading rusher on the 32 teams gained 1,124 rushing yards. Again, that was average. Last year, the average leading rusher gained 912 yards. Consider that in 1991, after Emmitt Smith, Barry Sanders, and Thurman Thomas, the fourth-leading rusher was New York’s Rodney Hampton, and he gained 1,059 yards. In other words, 2006 and its surrounding seasons — even if it might not feel like it — really was a different era of football for running back statistics.
The graph below shows the average rushing yards gained by the leading rushing for each team in every season since 1932. All team seasons of fewer than 16 games were pro-rated to 16 games1; the NFL line is in blue, while the AFL/AAFC line is in red. [click to continue…]
- But this does not pro-rate for injury. [↩]
Yesterday, I looked at the best AV-weighted winning percentages of offensive players. Today, we examine the same numbers but for defensive players and kickers since 1960. Again, players who entered the league prior to 1960 are included, but for purposes of this study, only their 1960+ seasons count (assuming they produced at least 50 points of AV). That’s a pretty important bit of detail to mention when it comes to the top player on the list. The player with the best AV-adjusted winning percentage since 1960 is Packers linebacker Bill Forester, who entered the NFL in 1953 but only gets credit for his 1960-1963 seasons in Green Bay (spoiler: those were pretty good ones). After him, of course, we have yet another Patriots lineman. Today it’s Vince Wilfork: [click to continue…]
By this methodology, Dan Koppen has the highest AV-weighted career winning percentage of any offensive player since 1960. The table below shows his AV and team’s winning percentage in each season of his career. Because Koppen’s best season came in 2007, when the Patriots went 16-0, Koppen’s career winning percentage gets a big boost from that season (18.7% of his career winning percentage comes from ’07 since 18.7% of his career AV comes from that year). On the other hand, Koppen played in just one total game for the 13-3 Patriots (2011) and the 13-3 Broncos (2013), so he gets almost no credit for those performances. Of course, he doesn’t need it, because his average season, after adjusting the weights based on his AV grades, was a 13-3 season.
% of Car AV
The table below shows the top 500 career AV-adjusted winning percentages among all offensive player since 1960 (minimum: 50 points of AV). As always, players who entered the NFL before 1960 are included but only their seasons beginning in 1960 count. The table below is fully sortable and searchable, so get to searching and leave your thoughts in the comments. [click to continue…]
Last week, Brian Burke provided some excellent data on the value of a first down. I began working on today’s post last offseason, but as you’ll see in a few minutes, I wasn’t quite comfortable with the results. But here’s what I did.
For all teams from 1989 to 2012, I recorded for each team’s average:
- Yards per carry;
- Touchdowns per carry;
- First downs per carry;
- Rush VOA from Football Outsiders (DVOA is FO’s main statistic, but it is adjusted for SOS; VOA is the unadjusted metric).
Then, I ran a series of regressions to help better understand the “proper” weights on the running game.1 First, I used yards per carry and touchdowns per carry as my input, and VOA as my output. The best-fit formula was:
-0.647 + 0.128 * YPC + 4.615 * TD/Carry
Understanding that formula isn’t important.2 What we care about is the correlation coefficient (0.65) and the relationship between the YPC variable and the TD/carry variable.
Here we run into our first problem: 4.615 is 36 times as large as 0.128. This would imply that a touchdown is 36 times as valuable as a regular yard (or 35, if you subtract the yard gained on the score). That seems very high, as 20 is the generally accepted standard conversion rate for a touchdown.
What if we introduce first downs per carry into the equation? Then we get this best-fit formula:
-0.761 + 0.081 * YPC + 2.954 * TD/Carry + 1.593 * FD/Carry
Here, the R^2 is 0.72, which is an indication that first downs matter (or, to put it another way, DVOA gives rewards for first downs, as it should). Unfortunately, the TD variable remains very high (it’s now nearly 37 times as large as the YPC variable), but we get the bit of insight I was looking for: a first down is worth 0.54 touchdowns.
The first down variable is about 20 times as large as the YPC variable, but that seems way off to me. Instead, if we think a TD is really worth 20 yards, this puts the value of a first down at about 10.8 yards. That’s not too far off from Brian’s 8.7 average, which I think makes sense to round to 9.0 once you remember that a few first downs happen on 4th downs, which was ignored in Brian’s analysis.
What helps bridge the gap between the two valuations? Burke’s method looks at the marginal value of a first down based on an Expected Points model. What I did with Football Outsiders’ numbers was to try to correlate first downs with rushing value. But that could lead to overstating the value of a first down, if first downs are correlated with other things (like say, short-yardage success in general). Teams that are good at rushing for first downs might be better at X, Y, and Z than other teams, and being good at X, Y, and Z could lead to a higher DVOA grade. As a result, I like Brian’s result better, and I think FO’s numbers serve as a good gut check. And while the number is a few yards higher, that’s arguably the correct result, and at least the number is higher in the right direction.
As for the idea that touchdowns are worth 36 yards and first downs are worth 19 yards? Well, I’m not so sure about that. My hunch though, is that Football Outsiders (wisely) cares a lot more about say, success rate, than just generic yards per carry. So perhaps this represents a devaluation of YPC as much as anything else.
But I’ll open it up to the crowd: what thoughts do you guys have for finding the value of a first down?