## Oops, I Almost Won the Game

The Lions almost accidentally beat the Titans. In the new OT format, the Lions were in the position where a FG ties the game and continues it, but a TD ends the game with a win. Head coach Jim Schwartz said he intended to try to draw the Titans off side on the critical 4th and 1, but the ball was snapped accidentally.

DET faced a 4th and 1 on the TEN 6. Let's look at the FG option first. A FG ties the game, but gives the ball to TEN in a sudden death format, which means a 0.43 WP (they'd win 43% of the time). FG attempts are good 97% of the time, so the FG option is worth a total of:

0.97 * 0.43 = 0.42 WP

Conversions on 4th and 1 are good 68% of the time. A 1st and goal at the 5 results in a TD 60% of the time and result in a FG attempt 28% of the time. The math gets a little tricky here because the TD has a WP of 1, but the FG gives us a 0.43 WP, very similar to what we calculated above.

A successful conversion would mean a total WP of:

0.60 * 1 + 0.28 * 0.43 = 0.72 WP

So the total WP for the go for it option is:

0.67 * 0.72 = 0.49 WP

Not all 4th and 1s are created equal. DET faced a fairly long yard to go, so the 0.49 WP number for going for it might be on the high side. But even if I use the conversion rate for 4th and 2 (55%), it would mean nearly a break-even decision (0.39 WP vs. 0.42 WP).

Note that the probabilities of TDs and FGs in these situations are somewhat speculative. There aren't nearly enough OTs in the new format for empirically-based inferences. We have to use examples from other situations to estimate them.

### 22 Responses to “Oops, I Almost Won the Game”

1. Chase Stuart says:

I like to use the gut test. If I was a Titans fan, what would I have wanted? The answer for me is to have Detroit line up to kick the FG. Especially with Megatron, you have to think the Lions odds of scoring a TD were higher than average (I think they led the NFL in red zone TD rate last year, too).

2. Joshua Perry says:

So, anyone criticizing Schwartz for his decision is a dick, right?

3. Wheell says:

I'd rather criticize him for his (marginally) bad decision to go for one last week. Going for two and finding out whether you need one more score or two early is underrated.

4. Geoff says:

No criticism of the Titans not going for it on the first possession of OT?

5. Jeff says:

This is a good illustration of the limitations of WPA since I think most everyone would agree the odds of Tennessee scoring if they got the ball back were fairly high, and much higher than the WPA formula (which is based on an *average* game situation, not one with such dominant offenses) assumes. This would tilt the Lions' odds even more toward going for it on 4th down.

6. Brian Burke says:

Jeff, I don't disagree, but why do you say the odds of TEN scoring were much higher than the WP formula?

7. James says:

That was not a long yard on the 4th and 1. There was a measurement after the 3rd down completion to LeShoure and it was a half yard short at most, looked more like a foot.

8. SportsGuy says:

James is right. The sticks came out to measure before that play and the gap was clearly less then a full yard. I figured about 18 inches.

I just didn't like the play call. I wonder if there was an option for Hill at the line.

9. Tarr says:

Chase Stuart's approach is actually an instructive one, because it inverts any built-in risk averse biases you may have. In the famous go-for-it decision Belichick made in the Pats/Colts game a few years back, the Colts fans in attendance were all petrified when the offense stayed out there for 4th down.

10. Tarr says:

Chase Stuart's approach is actually an instructive one, because it inverts any built-in risk averse biases you may have. In the famous go-for-it decision Belichick made in the Pats/Colts game a few years back, the Colts fans in attendance were all petrified when the offense stayed out there for 4th down.

11. James says:

Jeff, Tenn is currently 29th in Yards/Pass and 31st in Yards/Carry. I don't care how bad the Lions defense is, Tenn is in no way a dominant offense.

Even if you factor in strength of schedule as NE and SD are Top Ten in both defensive categories (and there's a cause/effect question there), Tenn O vs Det D is at best average, making the WPA analysis appropriately representative.

12. Jeff says:

Bryan/James - I guess there are some arguments against this, but my main point was that Detroit had shown little ability to stop Tennessee's offense. TN scored on 5 of their 10 offensive possessions, and were in field goal range for 7 of the 10 (missed 2 FGs). So my sense is that if Det kicks the FG, there is a significant chance (maybe 60% +/-) that they score on the next possession, plus some small chance of them winning the game on a later possession even if they don't score on the first possession.

Anyway, I guess my more general point isn't that the 43% number (or 57% for TN to win) you use is necessarily wrong. You make good points on Tennessee and maybe I am overestimating their chances. But it does seem like a problem with the WPA is that it based on an average offense vs an average defense, and a lot of times (esp in games where the offenses are scoring a lot of points) the real percentages may be substantially different than the WPA formulas.

13. Gershon says:

Brian: Aren't the WPA numbers in this situation flawed because it doesn't factor in the probability of a tie? If the Lions kick the FG, they may not even get the ball back even if the Titans don't end up scoring...

14. Mike says:

Of course, now the sports media are speculating Schwartz intended to go for it and was just covering his behind during the press conference saying it was a miscommunication.

I'm not sure if I buy that story, but if anyone would do that it's probably Schwartz.

15. LarryinLA says:

It looked pretty clear to me that the play was botched. Only the center and QB moved at the snap, and the exchange looked odd to me also, with Hill ending up tucking the ball, which is unusual for QB sneak, I think. That seems consistent with the idea that the ball wasn't supposed to be snapped when it was.

16. Anonymous says:

Why isn't anyone factoring in that there was about 6 minutes left. If they kick the FG, by the time they get the ball back again, they might not have enough time to mount a drive so the game ends in a tie.

17. Bob says:

Brian,

Love the site. Here you are treating every 4th down conversion as having a 1st down at the 5 yard line, but a ton of conversions will have better field position, or even score TDs.

Between this fact and the fact that the 4th and one was more like 4th and inches as people have mentioned, I think your WP for going for it is extremely conservative.

18. Jeff Clarke says:

I was watching on Red Zone and barely saw this after shaking my head in amazement that the Chiefs went for it on 4th and 1 on the 43 in OT. Obviously, this was the right call and they eventually won. Two straight correct calls (I didn't realize this was inadvertent) with the game on the line.

19. Anonymous says:

Brian, can you do an analysis of the Steelers 4th and 1 from like their own 30 vs. the Raiders? It was similar to the Pats 4th and 2 call only that the game was tied. Really ballsy by Tomlin

20. Eddie says:

Someone already did that analysis and it was by far the right call to go for it.

http://www.footballperspective.com/mike-tomlin-makes-aggressive-unconventional-call-converts-no-one-cares/

21. Anonymous says:

An odd effect of the new format is that it makes ties much, much more likely. (I expect to see one every couple of seasons under this format.)

The odds of winning, per se, are clearly far lower than 43% after kicking a field goal. OT was more than half over already by that point. Even if you count a tie as half a win, that doesn't automatically leave WP at 43% (although it might have no significant impact on WP-- I haven't worked through it, but my gut hunch is that counting a tie as half a win would actually raise the WP of a field goal slightly, since the other team is unlikely to get two possessions to your one).

Anyway, tie probabilities need to be accounted for to analyze the situation fully.

22. Brian Burke says:

Thanks for all the great comments. Agree with most everything. Ties used to be so rare, the model ignored the possibility. Can't do that anymore.

All the 4th down analyses I do take the conservative side of the go-for-it option, because that's usually the unconventional option. The numbers should be:

P (punt/kick) ~= x
P (go for it) ~=> y