The graph below shows the probability that a player who averages half a touchdown per game will score 0, 1, 2, 3, or 4 touchdowns in any given game:
How did I generate that? Using something called the Poisson Distribution, which Doug Drinen wrote about nine years ago. Let’s reproduce the relevant text here: [continue reading…]
Reminder: The 2015 Game Scripts page is now updated for week 9 results.
Last week, I wrote about the Rams heavy commitment to the running game. Well, in week 9, the Buffalo Bills had just 15 pass plays — 12 pass attempts and 3 sacks — while passing on just 29.4% of all plays. Both of those numbers set new team lows for the 2015 season.
Below are the Game Scripts data for each team in week 9. [continue reading…]
With six teams on bye this week, that left 26 teams playing in week nine. Not a single one of the main quarterbacks for any of those teams averaged fewer than 4.00 Adjusted Net Yards per Attempt. That’s incredible: overall, quarterbacks this week averaged an insane 7.12 ANY/A. Take a look: the table below shows the passing stats from all 30 players who threw a pass in week 9. I have calculated the Adjusted Net Yards per Attempt for each player as well, along with their VALUE (ANY/A minus league average multiplied by number of dropbacks) provided relative to league average, with one catch: league average is 7.12. As a result, all of the quarterback grades feel a little depressed. [continue reading…]
Today’s guest post/contest comes from Adam Harstad, a co-writer of mine at Footballguys.com. You can follow Adam on twitter at @AdamHarstad.
This guy can pick up first downs
Regular readers of Football Perspective are well acquainted with the sneaky-greatness of DeAndre Hopkins, who led the NFL in percentage of his team’s receiving yards in 2014 despite not even leading his own team in targets. [1]Hopkins had 127 targets in 16 games, or 7.9 per game. Then-teammate Andre Johnson had 146 targets in 15 games, or 9.7 per game. And, indeed, by “percentage of team receiving yards”, Hopkins is having another terrific season; his 37.0% share is slightly above the league-leading 35.0% he posted last year, (though it trails the 38.6% share he carried through his team’s first 14 games in 2014).
But Hopkins is having an even better season by a far less esoteric statistic: receiving first downs. As best as I have been able to determine, the all-time record for receiving first downs in a season is 92, set by Marvin Harrison in 2002 and tied by Calvin Johnson in 2012. [2]Obviously play-by-play data is virtually impossible to come by for older seasons. Thanks to frequent guest contributor Bryan Frye, I have complete first-down data going back to 1992; however (a) the … Continue reading Through eight games this year, Hopkins has converted for a new set of downs a remarkable 54 times, putting him on pace for 108, a ridiculous 17.4% more than the previous NFL record. (For context, if a quarterback wanted to break Peyton Manning’s single-season passing yardage record by 17.4%, he would need to throw for 6430 yards.) [continue reading…]
References
| ↑1 | Hopkins had 127 targets in 16 games, or 7.9 per game. Then-teammate Andre Johnson had 146 targets in 15 games, or 9.7 per game. |
|---|---|
| ↑2 | Obviously play-by-play data is virtually impossible to come by for older seasons. Thanks to frequent guest contributor Bryan Frye, I have complete first-down data going back to 1992; however (a) the best first-down conversion rate by a receiver with 80 catches over that span was 85%, (Michael Irvin’s 75 first downs on 88 catches in 1993), (b) only 2.9% of 80-catch receiver since 1992 even managed to top an 80% first-down rate, and (c) there were only 12 seasons prior to 1992 that even had more than 92 total receptions. Assuming an 85% conversion rate, a receiver would have needed 109 receptions to beat 92 first downs. Assuming an 80% conversion rate, he would have needed 116 receptions. Art Monk had 106 receptions in 1984, but given his sub-13 yard per reception average, I find it impossible to believe he converted on 88% of them. So with all due respect to Jerry Rice’s 1990 season and Charley Hennigan’s 1964, I feel pretty confident calling 92 receiving first downs the all-time NFL record. |
Let’s start with a piece of good news: I have created a 2015 Game Scripts page, where you can access the Game Scripts data from every game this season. That will likely be updated on Tuesday or Wednesday of each week.
The Rams are suddenly interesting to watch on offense. St. Louis traded into the top ten to draft Todd Gurley and Tavon Austin, and both players showed off their talents in week eight. Gurley had a 71-yard run, the highlight of a 20-carry, 133-yard day, while Austin ran for a touchdown and caught a 66-yard touchdown. Together, the duo combined for 265 yards from scrimmage as part of a run-heavy day for St. Louis. The team rushed 41 times compared to just 23 dropbacks, giving St. Louis the most run-happy identity of the week.
In losing efforts, two other teams stood out as run-happy. One, surprisingly, was Green Bay: The Packers were blown out, the sort of environment that usually leads to a 40-pass day. Instead, Aaron Rodgers had just 25 pass attempts, while the Packers finished with 21 carries. That ratio was supported by the efficiency metrics, though, as Green Bay shockingly averaged just 2.0 Net Yards per Attempt (compared to 4.3 yards per carry on the ground.)
The table below shows the week 8 Game Scripts data: [continue reading…]
Barnidge even scored against the Broncos… twice
For seven years, Gary Barnidge was one of the hundreds of nondescript players in the NFL. He played in 92 games for two teams, but logged just 25 starts. He caught 44 passes for 603 yards, an average of less than 100 yards a season, with just three touchdowns. He averaged 6.6 yards per game during this seven-season stretch in his 20s.
But Barnidge turned 30 on September 22nd, and then transformed into one of the most dominant tight ends in the NFL. Barnidge has played in six games since his 30th birthday, and has caught 36 passes for 512 yards and 6 touchdowns.
Let’s say Barnidge finishes the season with 1,000 yards in 16 games. That would mean his career average in receiving yards per game will have jumped from 6.6 to 14.8, a pretty remarkable increase for a player in his eighth season. In fact, Barnidge would become just the third player to have his career receiving yards per game double at any point after their fifth season in a year where they gained at least 500 receiving yards.
The last player to do so is Jim Jensen, a utility football player for the Dolphins who saw time at quarterback, running back, wide receiver, and tight end in his career. In his 20s, he totaled just 414 receiving yards in seven seasons, but he broke out in 1988 with 58 catches for 652 yards and five touchdowns. [continue reading…]
Writing is a tough business. That is especially true for sports writers, and sad news emerged yesterday: after four years, Grantland decided to suspend publication, effective immediately. There are a lot of great writers at Grantland, and the site’s lead football writer, Bill Barnwell, was one of the biggest supporters and promoters of this site. Bill was always willing and eager to link to Football Perspective from his larger platform, and that helped grow the readership of this site tremendously. I’ll always be grateful to him for the respect and admiration he showed for FP.
On a larger level, it’s frustrating and disappointing to see so many good writers unemployed: I wish all of them the best in their future endeavors. Another casualty is the Grantland NFL Podcast, hosted by Barnwell and Robert Mays. You can follow then on twitter at @BillBarnwell and @RobertMays, respectively, and I recommend that you do so you can follow them wherever they wind up. Their podcast made the NFL season simply more fun to follow, and hopefully we get to hear it again one day.
In one of their last podcasts, Mays noted that Jarvis Landry ran the 40-yard dash in 4.77 seconds, an incredibly slow time for one of the game’s most explosive young players. [1]Landry did tweak his hamstring at the combine during this run, which obviously may have impacted things. Mays made the comment that 5’11 wide receivers who run the 40 in 4.77 seconds aren’t supposed to do the things that Landry has done, and well, I agree.
In March 2014, I looked at the 40-yard dash times of all players since 1999, courtesy of the good folks at NFLSavant.com. I then took that database, and measured it against all wide receivers since 1999 who have averaged at least 50 career receiving yards per game (Landry is at 53 yards per game).
There are 51 wide receivers who entered the NFL since 1999, have averaged at least 50 receiving yards per game, and ran the 40-yard dash at the combine. [2]Michael Crabtree, Josh Gordon, Wes Welker, Roddy White, Allen Hurns, John Brown, Willie Snead, Mike Evans, and Brandon Marshall all have also averaged 50 yards per game, but either didn’t run … Continue reading In the graph below, I’ve plotted the height (on the X-Axis) and 40-yard dash time (on the Y-Axis, in reverse order, so the fastest and biggest receivers should be on the top right) of each of those wide receivers. As you can see, Landry is indeed an outlier, as Mays suggested: [continue reading…]
References
| ↑1 | Landry did tweak his hamstring at the combine during this run, which obviously may have impacted things. |
|---|---|
| ↑2 | Michael Crabtree, Josh Gordon, Wes Welker, Roddy White, Allen Hurns, John Brown, Willie Snead, Mike Evans, and Brandon Marshall all have also averaged 50 yards per game, but either didn’t run the 40 at the combine or that data wasn’t collected by NFL Savant. |
The New England Patriots on Sunday provided one of the most incredible pass/run ratios in recent history. Last year, teams were very pass-happy against the Jets, as a result of New York having a great run defense and a terrible pass defense. The Jets pass defense is significantly better this year, but that didn’t stop the Patriots from pretending the run option didn’t exist.
New England finished the day with 54 pass attempts, 3 sacks, and just 9 carries, representing an incredible 86.4% pass ratio. If you consider that Tom Brady had two scrambles and a third “carry” that went down for zero yards but was a sack on a pass play where Brady managed to get back to the line of scrimmage, and the Patriots really meant to pass on 60 plays, while calling runs just six times. A fourth run was a Brady sneak, leaving just five rushing attempts for the rest of the team that totaled exactly one yard. Brady was effective but not stellar in the passing game, but it was pretty clear that passing was the best option for the Patriots offense on nearly every play.
Two other big notes from week 7: Washington fell behind 24-0 against Tampa Bay, but won in the final minute of the game. That gave Washington a remarkable victory in a game where the team posted a -9.3 Game Script, topping the Bears game against the Chiefs for largest comeback as measured by Game Script. And, on the far other end of the spectrum, Miami produced an unreal 25.9 Game Script, the top score of the season. There have been three Game Scripts this year of over 20 points, and two of them have come against the Texans. The third was the previous high of the season, Arizona’s 24.3 Game Script against the 49ers. [continue reading…]
In my Washington Post article this week, I noted that Ryan Fitzpatrick was doing well as Jets quarterback in part because he was often playing with the lead. Fitzpatrick threw a whopping 58 times against the Eagles, and all but two of those plays came with the Jets trailing. In his other four games, Fitzpatrick threw just 18 passes while trailing. And, this year, Fitzpatrick has a 6.5 ANY/A average while throwing passes with the lead, and 4.8 ANY/A while throwing passes while trailing. [1]Frankly, even that understates the split. Fitzpatrick was terrible in the Eagles game, which, admittedly, may have more to do with the Eagles defense than the Game Script. But Fitzpatrick averaged … Continue reading
But I thought it would be fun to see how every quarterback has fared this year while leading and then while trailing, with a minimum of 30 pass attempts in each situation. That’s what graphed below, and the two guys who really stand out are Cam Newton and Andy Dalton. The Bengals quarterback has been outstanding this year in both situations, while the Panthers quarterback has been significantly more impressive this year while trailing. In the graph below, the X-Axis shows ANY/A while leading; for Newton, that’s a pedestrian 5.5 ANY/A. The Y-Axis shows ANY/A while trailing, which is an incredible 9.2. [continue reading…]
References
| ↑1 | Frankly, even that understates the split. Fitzpatrick was terrible in the Eagles game, which, admittedly, may have more to do with the Eagles defense than the Game Script. But Fitzpatrick averaged 2.98 ANY/A that day. In 13 passes against the Browns — all of which came in the 2nd quarter with the Jets trailing by 3 or 7 points — he averaged 7.1 ANY/A. And in 5 passes against Washington, Fitzpatrick averaged 20.2 ANY/A, which was largely the result of yards after the catch gained by his receivers. |
|---|
The Jets have long been one of the most run-happy teams in the NFL, particularly after accounting for Game Script. That rung true again in week 6, when New York was in control for most of the game against Washington. The Jets had a solid +6.1 Game Script, but most teams would still pass at least half of the time with that sort of game flow. But not the Jets, who passed on just 38.8% of pass plays and recorded a league-high 41 runs in week 6.
Four of those runs were from Ryan Fitzpatrick (an 18-yard touchdown, a 15-yard scramble, and two kneels), but Chris Ivory and Zac Stacy rushed 33 times for 192 yards (less impressive: Bilal Powell with four carries for -2 yards). The Jets are always going to run the ball, and when the Game Script goes their way, they will put up some truly impressive rushing totals. There have been just four games this year where a team has rushed more than 40 times, and two of those came in New York’s last four games. With a date in Foxboro this weekend, tracking the Jets pass/run ratio will be very interesting if the Game Script doesn’t go the Jets way. In New York’s only loss, the Jets had 59 passes and just 16 runs.
Without Ben Roethlisberger — and with Le’Veon Bell — the Steelers have become quite the run-happy team. Pittsburgh has now run more than its passed over the last three weeks, including a pass ratio of just under 40% despite posting a negative Game Script against the Cardinals (don’t let the 12-point final margin of victory fool you). The Steelers have completed just 43 passes over the team’s last three games.
Below are the week 6 Game Scripts data. As you can see, the biggest comeback of the week belongs to the Carolina Panthers. [continue reading…]
Last week, Michigan topped the SRS. Following the Gift Six, the Wolverines fall to the fifth spot after one of the craziest games in recent history. Jumping into the top spot is Baylor, after the Bears scored 56+ points for the sixth time in six games this year.
Baylor wide receiver Corey Coleman already has 16 touchdowns this year. 16! In six games! Okay, the Bears have only played two games of note — against Texas Tech two weeks ago and against West Virginia on Saturday — but the Bears also have the track record to show that they’re a top five team. Are they truly the best team in college football? We won’t find out more until a date with Oklahoma in four weeks, and the showdown with TCU two weeks later still looms as a de facto playoff game.
Without further ado, below are the SRS ratings through seven weeks. As always thanks to Dr. Peter R. Wolfe for providing the weekly game logs. [continue reading…]
With just 21 catches for 204 yards and two touchdowns through five games, Jimmy Graham is hardly making a big impact in Seattle. Consider that over his last four years in New Orleans, he averaged 5.6 receptions, 69.7 yards and 0.73 touchdowns per game, while he is at 4.2, 40.8, and 0.4 in those metrics, respectively, so far with the Seahawks.
So what’s wrong? Well, let’s start by focusing just on receiving yards. The drop from 69.7 to 40.8 is quite significant, but is there one main factor driving it? We can break receiving yards down into several components. For example, we can parse out four different metrics from simple receiving yards:
Receiving Yards = Team Pass Attempts * (Targets/Team Pass Attempt) * (Receptions/Target) * (Yards/Reception)
In the first four weeks, the biggest “comeback” as measured by Game Script belonged to the Atlanta Falcons, who defeated the Cowboys despite posting a Game Script of -5.4. But Chicago won on Sunday despite a Game Script of -8.9! This game seemed like a Chiefs win from before kickoff — when Kansas City was a 9-point favorite — until the very end.
The Chiefs led 14-3 early in the second quarter, 17-3 at halftime, and 17-6 entering the 4th quarter. With five minutes left, Kansas City led by that same score. With 3:11 to go, Jay Cutler found Marquess Wilson for a 22-yard touchdown on 3rd-and-6, giving the Bears new life. After the failed two point conversion, the Chiefs went 3-and-out, and Cutler took over on his own 33 with 2:04 to go. He drove the Bears down the field again, and found Matt Forte for the game-winning touchdown with 23 seconds remaining. The Bears drive chart reads: Punt, Fumble, Field Goal, Punt, Punt, Punt, End of Half, Field Goal, punt, Downs, followed by an 88-yard touchdown drive and a 67-yard touchdown drive. File this in your memory bank the next time a coach decides to take the conservative approach because his defense had been shutting down the opponent all day.
And while not on the same level, the Browns (-4.9), Bengals (-3.8), Steelers (-3.4), Bills (-2.7), and Falcons (-2.1) all won with negative Game Scripts. That always make the numbers a bit more interesting to look at. So let’s do just that: below are the week 5 Game Scripts data: [continue reading…]
For most NFL fans, the book on Andy Dalton has been written in permanent ink. But this week at the Washington Post, I write why 2015 may in fact be his breakout season.
So, through three weeks, it’s easy to dismiss the great numbers that Dalton has produced as the product of a small sample size. On 94 passing drop backs, he’s thrown for 866 yards and 8 touchdowns with just two sacks and one interception. That translates to a 10.32 Adjusted Net Yards per Attempt average, the best in football through three weeks. But is there any reason that Dalton, who has had hot streaks before, can maintain this level of play?
You can read the full article here.
Brees may not be throwing for awhile
Believe it or not, Marques Colston and Brees have connected for 72 touchdowns, tied with Philip Rivers and Antonio Gates for the 5th most in NFL history by any receiver/quarterback combination. I’ve written about that streak before, but here’s something else unique to consider: Colston has never caught a touchdown pass from anyone other than Brees. [continue reading…]
It’s a little late, but good news: I have the week 1 Game Scripts!
Regular readers know all about Game Scripts, but you can learn more about them here. Essentially, Game Scripts is the term I’ve used to represent the average margin of lead or deficit over the course of every second of a game.
In week 1, three won in week 1 despite having a negative Game Script: the Dolphins trailed by 1 point, on average, throughout the game against Washington, Dallas had a -2.7 Game Script against the Giants, and the Chargers came back from a 21-3 deficit to win, which produced a -4.8 Game Script.
Below are the results of every game from week 1. [continue reading…]
We spend a lot of time focusing on pass offenses, but not necessarily as much looking at the other side of the ball. After running my Rearview ANY/A numbers, it struck me just how bad Washington’s pass defense was last year. And if it feels like that team has struggled against the pass for awhile, well, that’s because it has. Over the last five years, there’s been no worse pass defense in the NFL.
As regular readers know, ANY/A stands for Adjusted Net Yards per Attempt, which is Yards per Attempt with a 45-yard penalty for interceptions, a 20-yard bonus for passing touchdowns, and includes sack data. Relative ANY/A simply subtracts the league average ANY/A from each team’s individual ANY/A. Last year, Washington’s pass defense allowed 7.88 ANY/A while the league average was 6.14; as a result, the team’s pass defense had a RANY/A of -1.7. In fact, the team’s defense has had a negative RANY/A in each of the last five years. [continue reading…]
On Monday, I looked at the SOS-adjusted Adjusted Net Yards per Attempt ratings of every quarterback and defense in the NFL in 2014. And just like last year, I want to follow that post by looking at the best and worst games of the year, from the perspectives of both the quarterbacks and the defenses.
Let’s start with the top 100 passing games from 2014. The top spot belongs to Ben Roethlisberger, for his scorched-earth performance against Indianapolis. The Steelers star threw for 522 yards and 6 touchdowns on just 49 pass attempts with no sacks or interceptions. For the game, that means Roethlisberger averaged 13.10 ANY/A. The league-average last season was 6.13 ANY/A, which means Roethlisberger was 6.97 ANY/A above average. Now since the game came against a Colts team that was 0.28 ANY/A worse than average last year, we have to reduce that by the same number. That puts Roethlisberger at 6.70 ANY/A above expectation; multiply that by his 49 dropbacks, and he produced 328 adjusted net yards of value above average after adjusting for strength of schedule. That was easily the top game of 2014. [continue reading…]
Every year at Footballguys.com, I publish an article called Rearview QB, which adjusts the fantasy football statistics for quarterbacks (and defenses) for strength of schedule. I’ve also done the same thing for years (including last season) using ANY/A instead of fantasy points, which helps us fully understand the best and worst real life performances each year. Today I deliver the results from 2014.
Let’s start with the basics. Adjusted Net Yards per Attempt is defined as (Passing Yards + 20 * Passing Touchdowns – 45 * Interceptions – Sack Yards Lost) divided by (Pass Attempts plus Sacks). ANY/A is my favorite explanatory passing statistic — it is very good at telling you the amount of value provided (or not provided) by a passer in a given game, season, or career.
Let’s start with some basic information. The league average ANY/A for quarterbacks in 2014 was 6.13, the highest in NFL history. Aaron Rodgers led the way with a 8.65 ANY/A average, the highest rate in the league among the 39 quarterbacks who started at least five games. Since the Packers quarterback had 520 pass attempts and was dropped for 28 sacks, that means he was producing 2.52 ANY/A (i.e., his Relative ANY/A) over league average on 548 dropbacks. That means Rodgers is credited with 1,383 Adjusted Net Yards above average, a metric labeled “VALUE” in the table below. That was the most in the NFL last year: [continue reading…]
Bryan Frye, owner and operator of the great site http://www.thegridfe.com, is back for another guest post. You can also view all of Bryan’s guest posts at Football Perspective at this link, and follow him on twitter @LaverneusDingle.
On Monday, I looked at which quarterbacks since 1960 helped or hurt their teams the most by taking or avoiding sacks. On Tuesday, I looked at the quarterbacks who gained the most or fewest yards through the air per attempt or dropback, and on Wednesday, we looked at completions relative to league average. Yesterday, the metric of the day was touchdown pass rate.
As promised, this article, “Dr. Safelove or: How I Learned to Start Worrying and Fear the Bomb,” centers on interceptions. The methodology here is no different from before: Figure out each player’s rate stats relative to the average of the rest of the league minus that player (LMP) that year and multiply it by his attempts to find the marginal total.
The caveat for this article is a big one: it is mathematically impossible for modern players to rank highly on a per play basis. In 1945, Sammy Baugh threw interceptions at a rate 7.4% lower than his peers. Because the league average today is so low (about 2.5%), a current quarterback would have to throw negative interceptions to match a -7.4% relative pick rate. Even if a quarterback threw 700 passes without an interception, the best he could possibly do is about -2.5%. [continue reading…]
Bryan Frye, owner and operator of the great site http://www.thegridfe.com, is back for another guest post. You can also view all of Bryan’s guest posts at Football Perspective at this link, and follow him on twitter @LaverneusDingle.
On Monday, I looked at which quarterbacks since 1960 helped or hurt their teams the most by taking or avoiding sacks. On Tuesday, I looked at the quarterbacks who gained the most or fewest yards through the air per attempt or dropback. And yesterday, we looked at completions relative to league average.
Today’s article, “Mile High Club or: Scoring through the Air,” is an examination of how often quarterbacks threw touchdowns. [1]Note that I didn’t say “how well quarterbacks threw touchdowns.” A screen with 80 YAC, a bomb to a wide open speedster, and a missile into tight coverage all count for six on the stat sheets.
I have used the same methodology as before (similar to Chase’s model for Relative Adjusted Net Yards), and I have maintained the same minimum attempt cutoffs. That means we’ll only look at seasons with 224 or more attempts and careers with 1,000 or more attempts. Like before, I didn’t prorate for shorter seasons. [2]Feel free to copy the table and make your own spreadsheet if you’d like. Or don’t. I’m not going to tell you how to live your life. [continue reading…]
References
| ↑1 | Note that I didn’t say “how well quarterbacks threw touchdowns.” A screen with 80 YAC, a bomb to a wide open speedster, and a missile into tight coverage all count for six on the stat sheets. |
|---|---|
| ↑2 | Feel free to copy the table and make your own spreadsheet if you’d like. Or don’t. I’m not going to tell you how to live your life. |
Bryan Frye, owner and operator of the great site http://www.thegridfe.com, is back for another guest post. You can also view all of Bryan’s guest posts at Football Perspective at this link, and follow him on twitter @LaverneusDingle.
On Monday, I looked at which quarterbacks since 1960 helped or hurt their teams the most by taking or avoiding sacks. Yesterday, I looked at the quarterbacks who gained the most or fewest yards through the air per attempt or dropback. As you may have guessed, I’m keeping the theme going today. This article, “Sharpshooters or: Quarterbacks who were Good at Completing Passes,” is an examination of how passers stacked up statistically against their peers in the not-super-important category of completion rate. [continue reading…]
Bryan Frye, owner and operator of the great site http://www.thegridfe.com, is back for another guest post. You can also view all of Bryan’s guest posts at Football Perspective at this link, and follow him on twitter @LaverneusDingle.
Yesterday, I looked at which quarterbacks since 1960 helped or hurt their teams the most by taking or avoiding sacks. In this post, “Frequent Flyers or: Quarterbacks who Gained Yards through the Air,” I’ll do something similar but with passing yards instead of sacks.
Because we have the necessary passing stats dating back to 1932, I can take this study back nearly three decades further than the previous one. However, I will use Chase’s estimated sack statistics to examine net yards for all post-1960 quarterbacks.
The math is simple: for each player, subtract his individual raw totals from those of every other quarterback in the league to find the league minus player (LMP) Y/A or NY/A. Next, subtract the LMP rates from the individual player rates to find each player’s marginal rate stats. Last, multiply each quarterback’s marginal Y/A by his attempts (or marginal NY/A by his dropbacks) to find marginal yards (or marginal net yards). [1]It took as much time to explain as it did to set up in Excel.
Enough explanation – Let’s look at some stats. The first table displays the 1,563 qualifying QB seasons, sorted by marginal yards. Read it thus: In 2001, Kurt Warner threw 546 passes for 4,830 yards, giving him 8.85 Y/A. The average of the rest of the league was 6.69, so Warner had a marginal Y/A of 2.15. This means his 2001 season is worth 1,176 yards above expectation. [continue reading…]
References
| ↑1 | It took as much time to explain as it did to set up in Excel. |
|---|
Bryan Frye, owner and operator of the great site http://www.thegridfe.com, is back for another guest post. You can also view all of Bryan’s guest posts at Football Perspective at this link, and follow him on twitter @LaverneusDingle.
Last Wednesday, Chase unveiled his estimated sack numbers for 1960-1968. [1]I can neither confirm nor deny that he did this, at least in part, to give Joe Namath some love. I already had this post planned, but I wanted to wait for the estimated stats before running the numbers, as doing so would allow me to go back to 1960 instead of 1969.
This article, “Upright Citizens (Quarterbacks who Avoided Sacks)” is a brief examination of those quarterbacks who saved their teams valuable field position by avoiding sacks. By extension, it is also an examination of those quarterbacks who did the opposite. When Chase presented his 1960-1968 data, he included everyone who threw a pass during that timeframe. Because I am only concerned with quarterbacks, I have removed all non-quarterback plays and recalculated the metrics. [continue reading…]
References
| ↑1 | I can neither confirm nor deny that he did this, at least in part, to give Joe Namath some love. |
|---|
You remember the November 20th game between the Bears and Lions in 1960, right? If you look at the boxscore on PFR, you will see that Detroit quarterback Jim Ninowski was 10 for 26 for 121 yards with 0 touchdown passes and 2 interceptions. You’ll also see that the Lions as a team went 10 for 26 for 121 yards with 0 touchdown passes, 2 interceptions, and 12 sacks for 107 yards. But the PFR boxscore does not indicate how many sacks Ninowski took that game, because the individual game log data wasn’t kept on that metric.
But, you know, I’m a pretty smart guy. I have a feeling that Ninowski was probably sacked 12 times in that game for 107 yards. I could be wrong, of course — maybe a backup came in and took two dropbacks, and was sacked on both of them — but it seems like making a good faith effort here is better than ignoring it completely. [continue reading…]
Last weekend, we looked at the league-average ratios between receiving yards and touchdowns, and which players scored far more touchdowns than we would expect. Today, we do the same but for rushing yards.
For whatever reason, Jerome Bettis’ 2005 has become etched in the memories of many folks. That year, he rushed for 368 yards and 9 touchdowns. Back in ’05, the NFL average was 133.6 rushing yards per rushing touchdown. So we would expect Bettis, with 368 rushing yards, to rush for 2.8 touchdowns. That means Bettis actually rushed for 6.2 more touchdowns than we would “expect” given his rushing total. [continue reading…]
On Tuesday, I looked at which receivers produced the most Adjusted Catch Yards over the baseline of worst starter. Yesterday, I used that data to help identify which receivers produced their numbers in the most pass-happy offenses. Today, instead of measuring wide receivers by how often their teams passed, I want to measure them by how well they passed.
Some teams are very efficient at passing because they have great wide receivers: to be clear, today’s post doesn’t prove anything about which way the causation arrow runs. But I do think it’s worth quantifying the reality that receivers produce their numbers in very disparte environments. Let’s use Joey Galloway as an example. Galloway, longtime readers will recall, was a favorite of an early iteration of Doug Drinen’s attempts at ranking wide receivers. For similar reasons, Galloway comes out “very good” in this system, if good means producing numbers while playing for bad passing offenses (a proxy, one could argue, for playing with bad quarterbacks).
Galloway produced 2,071 Adjusted Catch Yards above the baseline in his career, good for an unremarkable 84th place on Tuesday’s list. But let’s look at the 8 seasons that get Galloway there: [continue reading…]
Yesterday, I looked at which receivers produced the most Adjusted Catch Yards over the baseline of the worst starter. Today, I want to use that data to help identify which receivers put up their numbers in the most pass-happy offenses.
Let’s use Calvin Johnson as an example. He’s been with the Lions for each season of his career, and Detroit has been very pass-happy throughout his career. Last year, Detroit averaged averaged 40.56 dropbacks (pass attempts plus sacks) per game, while the league average was 37.29 dropbacks per game. So Detroit passed 108.8% as often as the average team.
In 2013, Detroit’s ratio to the league average was 108.2%, but it was 129.8% in 2012. To measure pass-happiness as it pertains to Johnson, we can’t just take Detroit’s average grade from ’07 to ’14; instead, we need to assign more weight to Johnson’s best years. Johnson gained 1,358 ACY over the baseline in 2012, which represents 29% of his career value of 4,721 ACY over the baseline. As a result, Detroit’s 129.8% ratio in 2012 needs to count for 29% of Johnson’s career pass-happy grade.
If we do this for each of the players in yesterday’s top 100, here are the results. [continue reading…]
Brown stuck the lanning.
Brown was dominant in 2014, and he led the NFL in more advanced systems, too. But today, I wanted to do something relatively simple. How do we compare Brown’s 2014 to say, three Packers greats from years past?
In 1992, Sterling Sharpe had 108 catches for 1,461 yards and 13 touchdowns. Those are pretty great numbers for 1992, although they don’t leap off the page the way Brown’s 2014 stat line does. If we go back farther, Billy Howton in 1956 had 55 receptions for 1,188 yards and 12 touchdowns. Like Brown, that was good enough to lead the NFL in two of the three major categories, and rank 2nd in the third. And 15 years earlier, Don Hutson caught 58 passes for 738 yards and 10 touchdowns. How do we compare that statline to Brown’s?
Here’s what I did.
1) Calculate each player’s Adjusted Catch Yards. For Brown, that’s 2,603. For Sharpe, Howton, and Hutson, it’s 2,261, 1,703, and 1,228, respectively.
2) Next, calculate the Adjusted Catch Yards for every other player in the NFL. Then, determine the baseline in each year, defined as the number of ACY by the Nth ranked player, where N equals the number of teams in the league. For Brown, that means using 1,398 Adjusted Catch Yards, the number produced by the 32nd-ranked player in ACY in 2014. For Sharpe, we use 1,078 ACY, the number gained by the 28th-ranked player in ’92. For Howton, it’s just 797, the number of ACY for the 12th-ranked player (keep in mind that ’56 was a very run-heavy year). And finally, for Huston, we use the 10th-ranked player from 1941, who gained only 413 Adjusted Catch Yards.
3) Next, we subtract the baseline from each player’s number of Adjusted Catch Yards. So Brown is credited with 1,205 ACY over the baseline, Sharpe gets 1,183 ACY over the baseline, Howton is 906 ACY over the baseline, and Hutson is 815 ACY over the baseline.
4) Finally, we must pro-rate for non-16 game seasons. For Brown and Sharpe, we don’t need to do anything, so Brown wins, 1,205 to 1,183. Howton played in a 12-game season, so we multiply his 906 by 16 and divide by 12, giving him 1,208 ACY, narrowly edging Brown. And in 1941, the NFL had an 11-game slate; multiply 815 by 16 and divide by 11, and Hutson is credited with 1,185 ACY.
As you can see, it wasn’t a coincidence I chose those three Packers seasons to compare to Brown. Those four seasons are the 19th-through-22nd best seasons of all time by this metric, and stand out as roughly equally dominant for their eras (both Sharpe and Hutson won the triple crown of receiving in their years).
This is not my preferred method of measuring wide receiver player, but it’s my favorite “simple” one. I put simple in quotes, of course, since there’s a lot of programming power behind generating these numbers. But at a high level, it’s simple: we combine the three main receiving stats into one, we adjust for era because the game has changed so much, and we pro-rate for years where the league didn’t play 16 games. Nothing more, nothing less. [continue reading…]
In 2013, NFL players combined for 129,177 receiving yards and 804 receiving touchdowns. That means, on average, a touchdown was scored every 160.7 receiving yards. Denver tight end Julius Thomas gained 788 receiving yards that year, which means we might have “expected” him to catch 4.9 touchdowns. But Thomas was no average scorer: he finished with 12 touchdowns, or 7.1 more than expected.
In 2013, only Jimmy Graham and Vernon Davis scored more receiving touchdowns relative to expectation than Thomas (Davis scored 13 times on just 850 receiving yards, or 7.7 more than expected, while Graham converted 1,215 receiving yards into 16 touchdowns, or 8.4 more than expected).
Well, as good as Thomas was in 2013, he was even crazier at scoring touchdowns last year. Despite gaining just 489 receiving yards, Thomas again scored 12 touchdowns, 8.9 more than expectation (the league average was 159.7 receiving yards per receiving touchdown). That was the most in the NFL, and only Dez Bryant (1,320/16/+7.7) was within shouting distance of him. [continue reading…]