Comments on Advanced NFL Stats: Going for It on Fourth Down

Love those charts man! What's your secret?

2008-03-12T22:45:00.000-04:00

Love those charts man! What's your secret?

Brian, I do not have a question specifically perta...

2008-03-05T10:59:00.000-05:00

Brian, I do not have a question specifically pertaining to this article, but wanted to get your attention here....
I have sent you an email, but you have indicated that you monitor these posts more frequently, so I wanted to give you a heads up.

Thanks

JKL-I know what you mean by tedious. I would be gr...

2008-03-03T14:16:00.000-05:00

JKL-I know what you mean by tedious. I would be great to have a pay-by-play database, like FO does, but one we could share openly and build on. Sort of a football version of baseball's Retrosheet. I'm working on something that might have some promise.

Anon-I love Herm, but he's afflicted like all the other coaches out there.

I think coaches are thinking of "how far they've come" too. Billick's quote is revealing on many levels. He displays a fan's perspective when he says,"after what we had done to get down there." It doesn't matter if you drove 99 yards with miraculous plays or if you lucked into a turnover. 4th and 1 on the goal line is 4th and 1 on the goal line. The only thing that should really matter is the probability of winning the game, not spoiling a nice narrative about the drive in question.

Herm Edwards, coaching the Jets, did the same thin...

2008-03-03T09:31:00.000-05:00

Herm Edwards, coaching the Jets, did the same thing, down three. I think it was fourth and one or two. With the win or loss in his hands if he goes for it, he takes the field goal and goes for OT. Loss is out of his hands!

I did actually limit the situations, but like I sa...

2008-03-02T13:44:00.000-05:00

I did actually limit the situations, but like I said, I can't find the file (probably deleted it). I think it was probably something like the first 3 quarters, and only if within 10 points either way. I did go through the play by play, but it was tedious and I dont want to do it again.

I agree if you just look at raw 4th down rates, it will be overrepresented by bad teams.

24-I'd like to do some work on that. But first I n...

2008-03-02T13:13:00.000-05:00

24-I'd like to do some work on that. But first I need a play-by-play database. In other words, I need the yards gained/lost on every carry, not just an average for a season or even a game. This off-season I'm working on a way to generate exactly that. But it is a large project.

JKL-Interesting. I've found that coaches aggressiveness on 4th down correlates with losing, or more precisely, desperation. There are some exceptions, like Belichick and Del Rio this year. Otherwise, it's the teams that are frequently behind in the 4th quarter that get aggressive. It makes sense.

It also fits in with the idea that the stronger a team's defense is, the more conservative its offense can be.

Say your defense gives up 30 points every game, and your offense always scores 24. In a very simple world, you could go 0-16. You'd be better off gambling and scoring 34 sometimes and 10 points sometimes. Your total net points would be the same, but you have a chance of winning 5 or 6 games. The point is that over-matched teams should be aggressive from the first kickoff and not wait until the 4th quarter to increase their risk tolerance.

I'd like to isolate the first 3 quarters and just look at those 4th down situations. This would tell us more about the coaches independent from end-of-game desperation. Again, a play-by-play database would make this possible.

Billick's decision was inexcusable. You don't eve...

2008-03-02T09:21:00.000-05:00

Billick's decision was inexcusable. You don't even have to rely on Romer, because he did not have to go through the myriad of possibilities after the decision. The equation is simple--chances of scoring on one play from inside the 1, versus chances of winning in OT. The first one is roughly 70%, the second roughly 50%. Not even close. He wasn't the only one to do that, as I also watched Whisenhunt make the exact same decision at home against SF, and like Billick, his team lost to an inferior team in overtime, when they could have won the game outright against not particularly good front sevens.

Brian, a couple of years ago, I also looked at coaches decisions on 4th and 3 or less situations for the 2005 season. I cannot find the file, so I can't report specifics. However, I do remember one general thing that stood out. I compared the coach's aggressiveness rating with the team's offensive/defensive balance. Thus, teams like Houston and Oakland 2005 were "offensive" teams because though their offenses weren't great, they were relatively better than their defenses. When I did that, and compared it to the coach's aggressiveness from their own 40 to the opponent's goal (teams universally rarely go for it inside their own 40), I found that for "offensive" or "balanced" teams, being too conservative (compared to the league average) or very conservative (for a balanced team) was very costly in terms of actual wins versus pythagorean wins. Coaches like Turner, Capers, and Haslett, as well as Vermeil, severely underperformed their pythag projections, and did not adequately match their aggressiveness with their team situation. Offensive teams that were aggressive in these situations, namely Belichek, but also guys like Shanahan and Fisher, were near or above their pythag projections.

However, for defensively skewed teams, there was no correlation between aggressiveness/conservatism and pythag performance. Being conservative didn't cost these coaches, but being overly aggressive, for the overall group, did not either.

I would be interested if you could try to duplicate a study on the last few seasons using your efficiency numbers, to see if being too conservative on an offensive team is particularly costly, while being overly aggressive on a defensive team is not. Such a finding would be further evidence in support of Romer.

I love reading your blogs. Keep up the great work....

2008-03-02T06:04:00.000-05:00

I love reading your blogs. Keep up the great work.

Some time ago you wrote, that we might find out which was better, the straight ahead running back like Jamal Lewis or boom or bust running backs like Barry Sanders.
Is that something you are working on?