Sirianni Leads Aggressiveness Index for 2022
NFL Offseason - Nick Sirianni was the most aggressive coach in the league on fourth downs in 2022, using sound analytics -- and a nearly unstoppable fourth-down quarterback sneak -- to significantly improve Philadelphia's chances of winning games.
Sirianni was sixth in Football Outsiders' Aggressiveness Index in 2021 but climbed into the top spot this year. What stood out the most is that Sirianni went for it on 13 of 14 qualifying fourth-and-1 opportunities during the regular season, with the Eagles converting on 10 of these. (They were also 2-for-2 in the postseason.) Sirianni also stood out by going for it on six of 10 qualifying opportunities in "no man's land" between the 34- and 39-yard-line where a head coach has to choose between a short punt, a long field goal, or a chance to convert on fourth down.
The next couple of coaches in Aggressiveness Index may be a surprise to you. Mike McDaniel was extremely aggressive in his first year with the Dolphins, finishing second in AI. McDaniel finished that high despite never going for it in eight qualifying fourth-and-2 opportunities! Arthur Smith of the Falcons finished third, a marked difference from the year before as his AI climbed from 0.89 in 2021 to 1.51 in 2022. The coaches after that are mostly the coaches that were on top of AI in previous years: Kliff Kingsbury fourth, Matt LaFleur fifth, Dan Campbell sixth, Sean McDermott seventh, and Kevin Stefanski eighth.
Robert Saleh of the Jets was the least aggressive coach in the league when compared to his peers, including a decision to go for it on only three of 12 qualifying fourth-and-1s. Other very conservative coaches included Mike Tomlin, who has always been very conservative, and Bill Belichick, whose career has been very strange in this area. As we've written about in the past, Belichick was one of the most aggressive coaches in the league year after year back when the league as a whole was much less aggressive. He inexplicably became much less aggressive on fourth downs around 2012, strange timing because it doesn't fit the two most common explanations for why Belichick became less aggressive. It wasn't timed to the failed fourth-and-2 against Indianapolis on a big Sunday night game in 2009, and it wasn't timed to the departure of Tom Brady which of course made it harder for the Patriots to convert in any situation, much less fourth down.
Football Outsiders introduced the concept of Aggressiveness Index way back in Pro Football Prospectus 2006. The goal was to find a way to rank coaches based on their tendencies on fourth downs in a manner that was easy to understand but accounted for the different rates at which the average coach will choose to "go for it" in different situations. There are other methodologies now for measuring fourth-down aggressiveness, mostly based on win probability analysis: for example, results spit out by Ben Baldwin's fourth-down simulator. Each methodology will have small differences in how it ranks the coaches, but Aggressiveness Index differs from the others by measuring coaches not against what they should do but against the actual decisions made by coaches themselves.
Aggressiveness Index numbers were designed to center around 1.0 and generally describe how much more (or less) likely each coach is to go for it on fourth down compared to his peers; for example, a coach with 1.20 AI is roughly 20 percent more likely to go for it than an average coach in equivalent situations.
You may remember that last year we re-did all the baselines for Aggressiveness Index. Head coaches in the NFL had become so much more aggressive since 2018 that the leaguewide Aggressiveness Index was up to 1.90 in 2021 and not a single head coach in the league came in below 1.0 that season. Based on changes in coach tendencies since 2018, our new baselines expect coaches to go for it more often on fourth-and-short, near the goal line, and near midfield.
Aggressiveness Index excludes obvious catch-up situations: third quarter, trailing by 15 or more points; fourth quarter, trailing by nine or more points; and in the last five minutes of the game, trailing by any amount. It also excludes the last 10 seconds of the first half, and it adjusts for when a play doesn't actually record as fourth-and-short because of one of those bogus delay of game penalties that moves the punter back five yards. Only the regular season is included.
Here are the full Aggressiveness Index results for 2022 with the number of qualifying opportunities, gos, and expected gos.
| Aggressiveness Index, 2022 | |||||||
|---|---|---|---|---|---|---|---|
| Rank | Coach | Team | AI | Go | Opp | Rate | Exp |
| 1 | Sirianni | PHI | 1.57 | 28 | 105 | 26.7% | 17.8 |
| 2 | McDaniel | MIA | 1.52 | 19 | 107 | 17.8% | 12.5 |
| 3 | A.Smith | ATL | 1.51 | 13 | 92 | 14.1% | 8.6 |
| 4 | Kingsbury | ARI | 1.46 | 21 | 107 | 19.6% | 14.4 |
| 5 | LaFleur | GB | 1.40 | 19 | 85 | 22.4% | 13.6 |
| 6 | Campbell | DET | 1.32 | 28 | 102 | 27.5% | 21.2 |
| 7 | McDermott | BUF | 1.32 | 11 | 81 | 13.6% | 8.4 |
| 8 | Stefanski | CLE | 1.30 | 23 | 107 | 21.5% | 17.7 |
| 9 | Vrabel | TEN | 1.24 | 12 | 110 | 10.9% | 9.7 |
| 10 | O'Connell | MIN | 1.18 | 12 | 100 | 12.0% | 10.2 |
| -- | Rosburg | DEN-2 | 1.13 | 2 | 14 | 14.3% | 1.8 |
| 11 | Eberflus | CHI | 1.09 | 8 | 88 | 9.1% | 7.3 |
| 12 | McCarthy | DAL | 1.03 | 13 | 104 | 12.5% | 12.6 |
| 13 | Staley | LAC | 1.02 | 17 | 113 | 15.0% | 16.6 |
| 14 | McVay | LAR | 1.02 | 9 | 100 | 9.0% | 8.8 |
| 15 | Taylor | CIN | 1.01 | 13 | 95 | 13.7% | 12.8 |
| 16 | Pederson | JAX | 0.98 | 18 | 103 | 17.5% | 18.3 |
| 17 | Bowles | TB | 0.93 | 17 | 122 | 13.9% | 18.3 |
| 18 | Reich | IND-1 | 0.90 | 5 | 57 | 8.8% | 5.6 |
| 19 | Shanahan | SF | 0.89 | 14 | 104 | 13.5% | 15.8 |
| 20 | Wilks | CAR-2 | 0.84 | 7 | 81 | 8.6% | 8.4 |
| 21 | Reid | KC | 0.83 | 11 | 88 | 12.5% | 13.3 |
| -- | Rhule | CAR-1 | 0.82 | 4 | 34 | 11.8% | 4.9 |
| 22 | Carroll | SEA | 0.81 | 8 | 97 | 8.2% | 9.8 |
| 23 | Rivera | WAS | 0.78 | 12 | 118 | 10.2% | 15.3 |
| 24 | Harbaugh | BAL | 0.76 | 14 | 102 | 13.7% | 18.3 |
| 25 | Daboll | NYG | 0.74 | 7 | 106 | 6.6% | 9.4 |
| 26 | L.Smith | HOU | 0.72 | 9 | 112 | 8.0% | 12.5 |
| -- | Saturday | IND-2 | 0.69 | 6 | 51 | 11.8% | 8.6 |
| 27 | Hackett | DEN-1 | 0.66 | 10 | 118 | 8.5% | 15.1 |
| 28 | McDaniels | LV | 0.65 | 12 | 90 | 13.3% | 18.5 |
| 29 | Tomlin | PIT | 0.59 | 7 | 109 | 6.4% | 11.8 |
| 30 | Allen | NO | 0.54 | 6 | 99 | 6.1% | 11.0 |
| 31 | Belichick | NE | 0.49 | 6 | 109 | 5.5% | 12.2 |
| 32 | Saleh | NYJ | 0.49 | 6 | 112 | 5.4% | 12.3 |
You already knew about Brandon Staley's backpedal from 2022, as he drops from our No. 1 coach in 2021 to the middle of the pack in 2022. Another head coach who dropped significantly was Ron Rivera, who was at 1.34 in 2021 but dropped to 0.78 in 2022.
Overall, aggressiveness in the NFL did decline slightly in 2022. Three coaches in 2021 had an AI higher than any coach in 2022, and four head coaches in 2022 had an AI lower than any coach in 2021.
Some additional notes on this year's Aggressiveness Index:
- Although Philadelphia went for it most frequently in qualifying fourth-and-1s, Cleveland was the team that went for it most often in qualifying fourth-and-1s: 16 out of 20 opportunities.
- We're still not at the point where any coach goes for it on every qualifying fourth-and-1, but we got close. Both Sirianni and Matt LaFleur (11 of 12) went for it on every qualifying fourth-and-1 except one.
- Kliff Kingsbury (5 of 10) and Brandon Staley (7 of 16) were notable for going for it on fourth-and-2.
- Sirianni and Doug Pederson (also 6 of 10) were the coaches who went for it the most in no man's land (from the 34 to the 39) but LaFleur (5 of 7) had the highest rate in that area.
- Four coaches never went for it in no man's land: Dennis Allen, Todd Bowles, Lovie Smith, and Sean McDermott (but with only one qualifying opportunity).
- Five coaches had at least six qualifying fourth-and-2s and never went for it: Mike McDaniel, as noted above, Bill Belichick, Ron Rivera, Lovie Smith, and Steve Wilks.
Thanks to Jim Armstrong, who came up with the idea of Aggressiveness Index and computes it for every season.
Comments
29 comments, Last at 23 Feb 2023, 9:22pm
#8 by theslothook // Feb 22, 2023 - 1:16pm
I've never understood Belichick's anti analytics commentary despite loving Math.
As an aside, I would argue the 4th and 2 decision was the turning point for the league at large with respect to 4th down aggression.
#2 by Meaningful gam… // Feb 22, 2023 - 11:38am
I like to think that we, as intelligent football fans, understand that it's no great honor to be in the bottom quarter of this list. However, it's more defensible when your defense is significantly better than your offense. And it's more broadly acceptable (even if it isn't actually better) when you're a defensive HC rather than an offensive one.
What the hell, Josh? What's your excuse?
#7 by jds // Feb 22, 2023 - 1:16pm
Well, I think that also matches with, relatively speaking, them having a better defense than offense. So they don't think their offense can convert, and they think their defense can get a stop.
Some interesting cases. Riverboat Ron in the past would be aggressive. Maybe this year he had no confidence in the offense, but had confidence in the defense? Same for Harbaugh. And for Belichick, maybe pre-2012 he was aggressive, but since then has built his defense (certainly this year) so that it is stronger than his offense,
This is Aggressiveness Index. I expect these guys know they are not being aggressive, and are fine with it, because this year, with these players, this is not a time to be aggressive.
#10 by StraightCashHomey // Feb 22, 2023 - 2:21pm
"Although Philadelphia went for it most frequently in qualifying fourth-and-1s, Cleveland was the team that went for it most often in qualifying fourth-and-1s: 16 out of 20 opportunities."
Is this a typo (Cleveland went for the most fourth-and-2s?), or is it saying that Philly had the highest rate but Cleveland had the most actual opportunities?
#11 by rpwong // Feb 22, 2023 - 3:07pm
I read it as:
- Philly attempted the highest percentage of 4th-and-1 opportunities (13/14 = 92.9%)
- Cleveland attempted the most 4th-and-1 conversions (16)
It's possible that another team had more opportunities than Cleveland, but attempted fewer than 16.
#12 by Oncorhynchus // Feb 22, 2023 - 3:12pm
It's unfortunate that for all Sirianni was aggressive all-season, his lack of aggression is arguably the biggest factor in the Eagles losing the Superbowl. There's not much difference between being up 3 points versus 6 points versus the Chief, but 3 versus 10 is huge. Sirianni should've gone for it on 4th and 6 or 4th and 3 in the 4th quarter.
#13 by Aaron Brooks G… // Feb 22, 2023 - 3:33pm
Guy went for it on 4th-5 from midfield, and went for it deep in his own territory in a tie game in the NFCCG, and you're carping about punting on 4th-3 from deep in your own end?
I know there's a school of thought that you should never punt, but that's a similarly-irrational display of high-variance behavior. (i.e., that knife cuts both ways)
#14 by Oncorhynchus // Feb 22, 2023 - 5:14pm
I don't think you should never punt it. There are definitely times to punt. But the 4th quarter of the Super Bowl should almost never qualify.
4th&5 at KC 44 in the Super Bowl was aggressive - but also analytically sound with an expected win probability (E[WP]) of +3.2 (according to Ben Baldwin). I would've supported it, even if it didn't work out.
As for 4th&3 at the PHL 32: it was 2 yards closer to the 1st down, 24 yards further from the end zone, and 30 minutes closer to the end of the game. It was an even more analytically sound decision to go for it (E[WP] = +4.6). I would've supported it, even if it didn't work out. I wouldn't have supported it, even if Toney muffs the punt and the Eagles recover.
As for the field goal at 4th and 6, that's a bit more equivocal. But the Eagles lost to the Chiefs in 2021 in the same fashion. Hurts was playing awesome, the defense wasn't and the difference was Sirianni choosing his kickers' legs over Hurts' arms and legs. You don't beat Patrick Mahomes and Andy Reid with field goals.
#15 by carlosla // Feb 22, 2023 - 9:50pm
In the treat of a lifetime, I was lucky enough to be at this year’s SB.
To paraphrase a guy who used to write abt this game, when the Iggles punted on that 4th qtr 4th down from their 32, I turned to my buddy and said “Game over.”
So sad!
#19 by Aaron Brooks G… // Feb 23, 2023 - 10:37am
If you can't keep the Chiefs from scoring TDs, then the extra three points from that TD (There's a solid chance Philly doesn't go for 2 to go up by 4 after their last TD; you want the option of Reid settling for a FG) don't matter. Remember, they intentionally bypassed a very good TD chance for a near-certain FG.
I'm dubious that you would have felt differently had a go-for-it failed, and this isn't all just post-factualism.
#24 by Oncorhynchus // Feb 23, 2023 - 12:02pm
You can be dubious all you like. If we were chatting in game I would've said the same thing. I'm a data analyst by profession and Eagles fan by birth. For example, I will argue with other Eagles fans that the only poor draft pick Howie made in recent years was Jalen Reagor. That's an argument for the process based on my understanding that drafting a player is essentially a bet. Every other draft pick that didn't work out were good bets at the time they were made. Some long odd bets he made were very good (e.g. Jordan Mailata and Jalen Hurts)
Same thing for game calls. It's about risk/reward - with two kinds of rewards to consider. The first reward is true for every offensive possession it's either points or maintaining position. The second reward you get as a consequence of winning the first reward: that's winning the game. This is weighted differently for different games and so it should also change your risk calculus. There is no higher reward than winning the Super Bowl. You risk and win and they'll build a statue outside the Linc. You risk and lose and they'll bitch about you WIP all off-season (but they're going to do that anyway - it's Philadelphia.)
So yeah, "they intentionally bypassed a very good TD chance for a near-certain FG", and that's my problem. I think your words convey you have some concept of probability at play here. I just don't think you're thinking deeply about the relative rewards. I think going up by 10 points is way better than going up by 6 points, especially in the 4th quarter of a game you're dominating by time of possession. Even if your defense can't keep the Chiefs from scoring a TD, your offense can play defense on the next possession by grinding the clock. Here's the other thing: the "rewards" of failure for either choice are not equal. Going for it and winding up short of the line to gain is better than kicking it and missing. Barring a negative-yardage play on going-for-it, you're looking at giving the Chief's the ball somewhere between the 10 and 15 yard line versus giving the Chief's the ball at the 33. So if you believe your defense can hold the Chief to a FG attempt (and so going up by 6 maintains your lead), then really you're saying you think you can keep the Chiefs from gaining about 40 yards. So if you really believe in your defense, then you should also go for it because by going for it, you've spotted your defense 18 yards.
#26 by Aaron Brooks G… // Feb 23, 2023 - 2:50pm
So yeah, "they intentionally bypassed a very good TD chance for a near-certain FG", and that's my problem. I think your words convey you have some concept of probability at play here. I just don't think you're thinking deeply about the relative rewards.
I'm talking about the Chiefs, who had 2nd-1 from the Philly 2 yard line with 1:36 left. They kneeled on their next play -- intentionally bypassing both the TD and the 1st down opportunity, in favor of setting up a 27-yard FG chance with 11 seconds left.
If Philly scored that TD instead of a FG, the Chiefs have 2nd-1 from the 2, down 3, with 1:36 left. I suspect their win percentage would be pretty high in that situation.
#29 by Oncorhynchus // Feb 23, 2023 - 3:38pm
I'm talking about the Chiefs, who had 2nd-1 from the Philly 2 yard line with 1:36 left. They kneeled on their next play -- intentionally bypassing both the TD and the 1st down opportunity, in favor of setting up a 27-yard FG chance with 11 seconds left.
Um... yeah? Because possession matters more than points in a tie game - especially with one minute left. I'm not sure what you're arguing for here. This is completely different than the Eagles decision to kick with 1 minute left in the 3rd quarter.
If Philly scored that TD instead of a FG, the Chiefs have 2nd-1 from the 2, down 3, with 1:36 left. I suspect their win percentage would be pretty high in that situation.
So what you're saying is, that in a different universe, things would be completely the same? That's completely nonsensical. The Eagles still could've lost if they got the TD. They also could've won if they went for it and missed. We have no idea how the rest of the game would've played out. I'm simply talking about the merits of the decision at the time it was made.
#22 by MuddyTiger // Feb 23, 2023 - 10:55am
Absolutely should be part of the equation. Coaching decisions on offensive aggressiveness will presumably be very different depending on who you are playing against, so it’d be interesting to see the evaluation of which defenses are played against more aggressively or least aggressively.
#17 by Noahrk // Feb 23, 2023 - 9:21am
Having watched a lot of the Dolphins I'm obviously not surprised by McDaniel. The one that sticks to mind, he went for it against Baltimore in the first quarter from his own 37 and that drive ended in a touchdown. There would have been no comeback without that decision, which no one aside from Sirianni would have made, and Sirianni likely only because he has a can't-fail play. Like Staley before, McDaniel faced a lot of backlash from the fanbase for being so aggressive, but unlike Staley he didn't have a rotten conversion rate, which hopefully will mean he'll have the fortitude to keep it up next year.
#27 by Aaron Brooks G… // Feb 23, 2023 - 2:52pm
And the injuries to his remaining QBs?
\Tua also got injured multiple times, and the Shanahan tree seems to like sending out guys who are crippled so they can sustain further injury.
\\That ended RG3's career, and may yet gather up Tua and Purdy.
#28 by Noahrk // Feb 23, 2023 - 3:04pm
I don't remember the others and I bet you don't, either. Blaming the head coach after one year of data screams of bias. If I recall correctly this all started with Shanahan, but McDaniel is running a very different offense. I don't know what you think they do, anyway. Maybe football is a violent sport with a high injury rate and sometimes weird patterns can be found if one looks hard enough.
#31 by LionInAZ // Feb 23, 2023 - 9:22pm
Isn't it time to start evaluating the results of all this aggressiveness? Which coaches provide more success by being aggressive, and which ones suck? Do some do better not being aggressive? Aggressiveness by itself is not a success rate, just a playcalling tendency.