Did you watch the Cowboys/Seahawks game? When they lined up to kick the PAT, did you think they were making a strategic mistake?
Agreed. Even further, I’d say your chance of winning becomes effectively 0% if you miss the 2pt regardless when you try for it.
After kicking the PAT, then somehow getting the ball back and driving for a game-tying TD, the defense will be more tired out and possibly easier to score the 2pt against They’ve just (unsuccessfully) defended two drives in a row with little break in between. Maybe they’re a little more demoralized, feeling like overtime is inevitable, so they’re a fraction of a second slower on the 2pt attempt as the last play of regulation.
It seems so intuitively obvious to me that the most fundamentally sound strategy is to put off any “make or break” decisions until the last possible moment. Since the 2pt is “make or break,” I would always want to delay it until the last possible moment. I would think this would be some kind of law or philosophy in game theory – even if it were a discredited one --but a quick google hasn’t turned anything up.
The closest I could find is the concept of the Last Responsible Moment in software development: A strategy of not making a premature decision but instead delaying commitment and keeping important and irreversible decisions open until the cost of not making a decision becomes greater than the cost of making a decision.
I absolutely would rather know with 0:00 left, but I’d take :17 over 2:17 in a heartbeat, yes.
Much like I view a “successful” season as one where the last game you play is meaningful – I got that from Wellington Mara – a game is “successful” when the last drive you have is meaningful. If the last time you have the ball you are down by 2 scores, that’s not meaningful, that’s garbage time.
I concur with this approach. First go for 1 to make it a “one-score game.” And then worry about what comes afterwards. If you are so lucky as to score yet another TD, then, at that point, you try the 2-point PAT.
It’s firmly a case of “Cross that bridge when you get to it.”
Well, you are going wrong in the way that I had guessed, and what I wrote didn’t seem to change your mind. I understand where you’re coming from, but the arguments you make pertain to what you enjoy as a fan, and what seems intuitive to you, rather than any case that your side of the argument has a greater chance of winning the game.
But let’s just say I’m right for a moment, and going for the 2 point conversion first gives you some non-negligble higher chance to win. Let’s say if you do it my way, you have a 3% greater chance to win. Would you take a 3% higher chance of winning against the possibility of suffering 2 extra minutes of garbage time if you lose?
If not, would 10% make a difference? 20%? When does it change over?
To me, I’m not sentimental about the whole thing. I’d take a 0.0001% greater chance to win over any sort of desire to make the game interesting until the end. And I think that’s what a coach should be basing their decision on. I could kind of see how, as a fan, you’d prefer leaving it up in the air until the last possible moment, but surely it doesn’t make sense for a coach to make decisions based on that. The coach should be focused solely on maximizing the team’s win chances.
Your other point - where the defense will be easier to score a 2 point conversion on after the last TD because they’re more worn down, is a valid point and makes sense, since it relates to actual chances of winning rather than just being what feels intuitive. I still feel like the advantages of going for 2 first outweigh it (especially the penalty issue), but you could make a case with the wearing down argument. It would be hard to empirically prove either way since it’s a pretty close decision in either case. It’s certainly not the case that the side who advocates the early 2 point conversion is totally wrong or crazy - that reaction stems from your intuitive sense of how it should go down, rather than your logical analysis of the winning percentages of each strategy.
This is the conventional wisdom in football. Coaches and commentators will talk about “extending the game” and similar things. But if your goal is to maximize your chances of winning the game, then the idea of extending the game is literally irrelevant.
There are reasons to care about things other than increasing your winning chances. If you put extra value on continuing to have winning chances later in the game, then delaying important decisions might make sense. Coaches seem to put value on that, and on not losing by larger margins, or looking stupid by making a non-standard decision that goes wrong.
It’s a bit similar to deciding what your strategy should be as a big underdog. If you’re only concerned about winning the game, then you should use a high-risk strategy to increase your variance. This will increase your chances of being blown out, decrease your expected scoring margin, but also increase your chances of winning the game.
I was actually thinking of asking something like this as a hypothetical, to demonstrate the point of separating the intuitive from the logical.
Something like “would you rather have a 55/45 chance of winning, but if you win it’s a small win, but if you lose it’s a blowout? Or would you take 50/50 with a more normal range of outcomes?”
Intuitively, it feels worse to be blown out, obviously, but if your only concern is maximizing your win chances, then winning or losing should be all that matters.
One of the reasons that Belichick is so good is that he makes these sorts of decisions all the time, because he has all the job security in the world, and he’s very strategy minded. He doesn’t mind doing what seems unconventional if he thinks it’ll give him even a sliver of a greater win chance, even if it gets him all sorts of criticism when it fails.
Football fans love to jump on coaches who do the unconventional thing. They like an easy target for blaming. So as soon as a coach makes any decision that strikes fans as unconventional or counterintuitive, and it fails, they get a way overblown undeserved amount of criticism. NFL coaches largely play to avoid making unconventional decisions in this way, because while it very well may be the correct decision, and given their team the best chance to win, when it fails it may cost them their job. So they keep their heads down and make the wrong decision. This is why, as an example, NFL coaches are so wrong so consistently on when to go for it on fourth down in marginal situations.
Suppose, when the Cowboys scored with 2:17 left to pull within 9 points, they had all 3 timeouts left. Would that make a difference in the XP decision?
I say it does. SenorBeef offers some interesting points in the decision-making process, but all of them (I think) depend on a recovery of an onside kick at some point. If the Boys had all of their timeouts left (plus the 2-minute warning), then I think the correct strategy is to kick the extra point. You would then hope that your defense can stop the Rams without a first down, and you get the ball back, down 8, with somewhere around 2 minutes to play, and a chance to drive for the game-tying TD/2-point conversion.
However, if you go for 2 with 2:17 left, and fail, then you MUST recover an onside kick at some point to have a chance to win the game, because you are still down 2 scores.
You haven’t established that there is a higher chance to win by going for the 2 first. Regardless what you do, you need:
[list=#]
[li]A 2pt conversion (slightly higher % chance my way as the defense is more tired)[/li][li]A next TD (higher % chance my way since you have more time/timeouts to spend while doing it compared to saving time for another drive)[/li][li]If you miss the 2pt, another FG after that (0% chance my way, your way is technically still alive)[/list][/li]I’m saying that the higher success rates of #1 (very small) and #2 (significant) in the conventional strategy add up to much higher than the advantage your way has in #3.
This would be my strategy if I were coaching or playing, not just as a fan. I mean, it’s not like I would do it your way while playing Madden but then want my NFL team to do it my way in a real game.
I think my way actually maximizes the chances to win, while your way makes it less likely to win. For example, down by 2 scores with little time left my offensive players might mentally check out because they’re facing hopeless odds.
I’m a Giants fan. The Giants have been big underdogs in three different Superbowls that they won, and the way they won them is exactly opposite to what you advocate here. They employed low-risk ball-control strategy to reduce variance and rode that strategy to victory all three times. The one time they didn’t do that while underdogs in the Superbowl they got massacred by the Ravens.
It’s worth noting that for the first of those three underdog championships, the architect of that game-winning low variance approach was Bill Belichick.
But if you wait, get the ball back, get another TD, and then miss the 2 point, then you also must recover an onside kick (if you’ve left yourself any time). Missing the 2 point forces the onside kick, regardless of when you try the 2 point.
Absolutely. If you believe that the chances of actually converting the 2 point is higher after the second TD than after the first, then waiting is probably correct, since the other advantages we’re discussing by going first are small.
I probably shouldn’t have used “high risk” there, as it’s not necessarily that you want to increase the risk on individual events. The point is to increase variance, which you can do by reducing the number of possessions in the game, as the Giants did.
I’m saying that no matter what you do, you need at least one more touchdown. My position is that your chances to score that next TD are significantly higher if you can focus all your resources (game time remaining, timeouts, emergency plays you’ve been cultivating for years but haven’t used them yet so nobody has film on them) into that one single drive than if you need to save time for another drive after that.
My position is that the difference in chance to succeed on that next TD drive – which you need no matter when you go for the 2pt – positively dwarfs any trivial advantage you have if you went for the 2pt first and missed.
Picture 30 seconds left, no timeouts, you’re about to attempt an onside kick, and you need a FG to win. What’s your chance to win? Not much higher than a couple %; Beef said 3%. Sure, let’s go with that. 3% higher than the 0% if you waited until time expired before missing the 2pt.
You would have had a much higher than 3% extra chance to succeed on the touchdown you just scored if you didn’t need to save that 30 seconds for this onside kick / FG drive.
Yes, the overall chances of getting that touchdown may be higher than in the 3% case, but that doesn’t matter. It’s not comparing the right things.
Go for 2 on the first TD:
Make the 2-point, score a single drive TD, game is tied
Miss the 2-point, win with the 3% miracle (or whatever percentage you want to assign)
Go for 1 on the first TD:
Score a single drive TD, make the 2-point, game is tied
Score a single drive TD, miss the 2-point. Lose.
In both scenarios, you force OT by making a 2-point conversion and scoring a focused, single-drive TD. In the first scenario, you also have the miracle chance. In the second scenario, you don’t. In the second scenario, scoring that second TD is overall more likely; that means you’re more likely to lose by 2 instead of by 9, but you’re also slightly less likely to win the game.
Like you said though, it’s possible the 2-point is more likely to be successful after the second TD than the first, which would mean the chances to force OT aren’t the same in the two scenarios, which could overwhelm the miracle chance.
You can do all those things if you successfully run a 2 point conversion first, too.
“But what if you fail the 2 point conversion? Then you need to hurry up on your next drive and can’t pull out all the stops for one last drive”
… but in this example, if you go for it second, and fail the two point conversion, you lost the game.
You’re essentially acknowledging being forced into a losing/long shot/desperate situation if you go for 2 on the first TD and fail, while not weighing that against the fact that if you go for 2 and fail on the second TD, you lose. You’re counting the upside in one scenario against the downside in another scenario.
There’s a hidden bias I think you’re not picking up on, which is that you’re weighing a (known) failed two point conversion against an (unknown) 50/50 two point conversion. That’s not a fair comparison. Obviously an unplayed two point conversion is more appealing than a failed one.
Consider:
You go for the 2 point conversion early. You succeed. You pull out all the stops for one last drive to run out the clock and then either kick the xp (tie) or go for another 2 point conversion (win)
You go for the 2 point conversion early. You fail. You make a desperate long shot attempt to make 3 scores to come back.
You go for the XP early. You pull out all the stops for one last drive to run out the clock and go for a 2 point conversion to tie as time expires. You succeed. (tie game)
You go for the XP early. You put out all the stops for one last drive to run out the clock and go for a 2 point conversion to tie as time expires. You fail. You lose.
You appear to be saying that 3 is better than 2. Well, yes, it is. But those aren’t comparable outcomes. You can’t choose to compare the downsides of going for the two pointer early and failing to the benefits of going for the two pointer later and succeeding. Obviously it’s better for your two point conversion to succeed than to fail.
So the comparison is between 1 and 3 (which have the same upside of being able to have that focus on one drive, with the additional benefit of being able to choose to tie or win at the end) and 2 and 4 (in which 2 has at least a slim chance over 4’s zero chance).
I’ve gotta run for today. I’ll be around later tonight to continue this.
Is the logical conclusion to what your saying that you should pretty much never kick a PAT barring extreme circumstances? (Down by 1, game almost over.)
I already gave a scenario in which you’d kick the XP - if you committed a penalty on the 2 point conversion and had to re-try it as a more difficult conversion from further out. That’s part of what going for 2 first allows you. You can always fall back on the PAT on the first attempt if you botch the 2 pointer with a penalty - after all, changing the XP distance from 25 to 35 yards is much less of a big deal than taking a 2 point conversion from 2 yards to 7 or 12), which is part of the flexibility that going for two first allows. If you get a penalty on the 2 point conversion after waiting for the second TD, you have no choice - you have to attempt the 7 or 12 yard conversion.
It’s not a big swing either way in terms of expected winning percentage, I don’t think, so you could decide to kick the XP for small reasons. If for whatever reason you thought your chances of making the 2 point conversion were notably higher the second time (maybe your RB is gassed because he just ran a 90 yard TD and you think your best chance at a conversion is with him) then it would probably swing things mildly in favor of going for the PAT first.
But generally going for the 2 pointer first will be correct more often than not, I think.
No, I mean in general. Is the logical conclusion that you should always go for 2, even in the first quarter? That way you leave even more time to make up for missed 2pt attempts, right?
Before they moved the PAT back to a snap from the 15 yard line (making it effectively a 32 yard try), conversion of a one-point PAT was 99+%, and conversion of two-point tries was at around 50%. From an expected-points POV, it was pretty much a wash, though I think that risk-averse coaches likely preferred a close-to-sure thing over a 50/50.
However, the change to the PAT has taken the success rate down several percentage points (the conversion rate has been around 94% for the past three seasons). Assuming that the conversion rate for 2-point tries is still above 47%, it would actually make long-term sense to try for 2 as often as possible (though certain game situations and point spreads would still call for the one-point play).
The only issue that answers the question is whether or not there is a significant difference in likelihood of scoring a 2-pt. conversion the first time or the second time. All other arguments are essentially meaningless in comparison.
No, it isn’t, because early in the game you’re generally trying to maximize the points you score (and minimize the opponent’s), and late in the game the score dictates the value of different possible scoring. But this was discussed in the old thread too, so I’m out
