The Two Point Conversion When Eight Down

[srjunkacct]

[ QUOTE ]
Is the 42% backed with data? I can see it at mid-field, but it seems high for the 2 yard line.

[/ QUOTE ]

I'm pretty sure it is, though the sample size (of actual 2-point conversions) is probably not more than a couple hundred attempts or so. I don't know what it is if all 4th-and-goal attempts from the 2 are counted.

I always thought the NFL 2-pt conversion should be from the 3-yard line, the way it is in high school and college.

Edit: Found these stats from a newsgroup posting:

http://groups.google.com/group/rec.sport...4af501cc6ff7ac4


1996: 44/92
1997: 47/109
1998: 41/105
1999: 31/84
2000: 31/85

Total from 1996-2000: 194/475, 40.8%

[ALawPoker]

While I completely agree that it is correct to go for two in the spot David highlights, I want to point something out. Yes, the logic here is very simple to understand. Football coaches aren't idiots; they'd probably realize the mistake in a heartbeat if it was brought to their attention. But you have to realize, this is just one specific example in a football game that is filled with countless mathematically influenced decisions. I really don't see it as surprising that coaches make this mistake, as I don't see the situation as one that is significant enough to warrant prior consideration. To ask the question "how can coaches not understand why it's correct to go for two in this spot" is really asking the question "how can coaches not make a mathematically optimal decision in every spot."

The answer is that you don't want them doing that. Coaches aren't math wizards, and math wizards wouldn't make good football coaches. Some coaches understand the science of the game better than others, but few (probably none) can be expected to make the mathematically optimal decision all the time. If a coach was constantly thinking in mathematical terms, and not in the language that he understands best (raw football observation), he would make far more errors. Either by mis-calculation or, more often, the opportunity loss of not picking up on something (or being as in flow with the game as he'd like to be) because he is so engrossed in a practice that does not come as naturally to him. His "raw observation" is not optimal, and this example is one where math would do him better, but in general a football coach will make better decisions when he bases them on other things.

If my point is not clear, consider a poker player who can't do relatively easy math to save his life, but he has amazing instincts and reading ability. You could point to somewhat obvious math errors and say "how could he make this mistake" but you'd be discounting all the equity he gained on a good read or an instinctual fold. If he spent some of his focus on the math, he may calculate errors, or if nothing else he would miss opportunities to use his other skills. So while you wish he didn't fold his 52/48 pot equity some of the time, you should understand that it is actually to his long-term best interest to make this mistake. Ideally you could teach him the math to go along with his instincts, but in the case of football, I think you lose more than you gain by expecting your coach to factor it in near perfectly.

I don't really see this situation as a huge deal, so as to say "wow, how could they miss THIS." It adds 5%, and occurs fairly infrequently. If I'm the owner of a football team I can probably find a more critical mathematical inefficiency to beat into my coach's brain in between seasons. And I'm certainly not gonna bother distracting him with all of them. So in short, thinking he should or wishing he would make the right decision in this spot is somewhat myopic.

[gusmahler]

Where are people getting the 96% or even 98% success rates for 1-point conversions?

This season, there have been 767 1-point conversion attempts. 761 have been successful. That is a 99.2% success rate.

Last season it was 1091 out of 1104 or 98.8%

(Stats taken from adding up the individual numbers of all the kickers at www.nfl.com/stats)

[gusmahler]

Denny Green has been villified by many "pundits" for going for two in a game last week.

Situation: Cards down 31-13 with 10 minutes to go. Cards score a TD and have to decide what to do. Denny decided to go for 2 and failed, leaving the game at 31-19. So when the Cards scored later, the score was only 31-26 (they went for 1 because a 2-point conversion is meaningless when there is a 6 point differential).

The critics say that, if he had gone for 1 the first time, it would have been 31-20. Then, when the Cardinals scored the second time (with 1:49 remaining), they could have gone for two THEN and gotten the score to be 31-28.

Basically, the critics say that going for 2 with 2 minutes left is better than going for 2 with 10 minutes left, even though you end up with the exact same score if the 2 point conversion failed.

[KreellKeiser]

I'm not gonna read replies. But basically:

You have a 42% chance to win (as part of the whole use of the strategy) if you go for two and make it: 41% for making 2points and making the extra point after the second touchdown and then another 1% for when you miss the extra point and go into overtime (this will happen 2% of the time and you will win 50% of that).
x=(.42)(.98)+(.02)(.5)=42%

Now if you miss the first extra point: you then have to make the second extra point. This will give you your 50-50 shot in overtime. In this case you would win 12% of the time (as part of the whole use of the strategy)
y=(.58)(.42)(.5)=12%

So in total, x+y=54% to win if you adobt David's strategy.

If you do not, then you have the following possibilities to win: you make 2 extra points and win in overtime, you miss the first extra point then make a 2 point conversion and then win in overtime. (obviously if you make the first extra point and the miss you lose).
z=(.98)(.98)(.5)+(.02)(.42)(.5)=48%

Clearly you have a much better chance ot win using David's strategy rather than the conventional wisdom, assuming the numbers used are accurate. However there are other important factors such as how good is you defense? (if the score is 14-0 that's a big difference than if the score is 49-35).

[andyfox]

"Being a great baseball manager is about a lot more than just lineup decisions and decisions made during games."

No doubt. So give us a few months to learn the ropes. I also happen to be a good manager of people.

[Pudge714]

Even if we assume OT isn't exactly 50/50 it's not like it will end up being 60/40 or 70/30 it will be more realistically something like 51/49 especially since you only need to kick a FG and that the coinflip plays a very big role in who wins.

[Eponymous]

[ QUOTE ]
Denny Green has been villified by many "pundits" for going for two in a game last week.

Situation: Cards down 31-13 with 10 minutes to go. Cards score a TD and have to decide what to do. Denny decided to go for 2 and failed, leaving the game at 31-19. So when the Cards scored later, the score was only 31-26 (they went for 1 because a 2-point conversion is meaningless when there is a 6 point differential).

The critics say that, if he had gone for 1 the first time, it would have been 31-20. Then, when the Cardinals scored the second time (with 1:49 remaining), they could have gone for two THEN and gotten the score to be 31-28.

Basically, the critics say that going for 2 with 2 minutes left is better than going for 2 with 10 minutes left, even though you end up with the exact same score if the 2 point conversion failed.

[/ QUOTE ]
This example is exactly why football coaches will not exploit edges like the one proposed here that may win one extra game every several years (because of the infrequency in which the situation presents itself). The media are about as clueless when it comes to math, probability, and logic as the football coaches are. Even if the owner had a wise and insightful owner, he probably would not be able to withstand the pressure of the media and public to keep a coach who consistently makes "bad" decisions (he needs to sell tickets and merchandise, after all). Even if the owner was willing to withstand that pressure, the coach would not be able to have that good a read on his owner's thinking, no matter how much confidence he tells him he has in him. (votes of confidence are given all the time, often right before a firing)

[armPitt]

ok, so we've sufficiently answered the question, so i'd like to ammend it.

should you conventionally go for two, or do a fake extra point attempt?

[ALawPoker]

[ QUOTE ]
ok, so we've sufficiently answered the question, so i'd like to ammend it.

should you conventionally go for two, or do a fake extra point attempt?

[/ QUOTE ]

It depends which alternate reality we're dealing with: If your team is the maverick team for the mathematical efficiency, or if everyone has sort of come to anticipate it.