Fantasy Football - Footballguys Forums
This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
up 7 with 5 mins left....Go for PAT or 2? (1 Viewer)
- Thread starter biggamer3
- Start date
Warrior
Footballguy
All the statistics say that this is completely wrong and it's not close.
biggamer3
Footballguy
has there been studies on this? would love to see it
Bills_Fan11
Footballguy
I'd like to see those too. Considering the other team is never going for 2 and the win when they get within 1 and thus OT is the worst case scenario either way, it can't be a slam-dunk pat.
Ignoratio Elenchi
Footballguy
I, too, am curious to see the statistics Warrior is referencing.
Let's simplify as a starting point. Assume that the probability of successfully converting a 2-pt conversion is 0.40, the probability of successfully converting a PAT is 1.00, and the probability of winning if the game goes to OT is 0.50.
If you kick the PAT, you go up by 8 with 5:00 left. Assume the other team uses most of that time to drive down the field and score a TD:
20% of the time, they'll successfully convert the 2-pt try to tie, and then beat you in OT.
20% of the time, they'll successfully convert the 2-pt try to tie, but you'll beat them in OT.
60% of the time, they'll fail to convert the 2-pt try and you win.
If you go for two:
60% of the time, you'll fail to convert the 2-pt try and will be up by 7 points. Assume they drive down the field and tie the game with a TD. 50% of the time they'll beat you in OT.
40% of the time, you'll successfully convert the 2-pt try and will be up by 9. Assume the best case scenario, that they are unable to make up a 9 point deficit in 5:00 and you win.
So based on this, you win 80% of the time you kick the PAT, and 70% of the time you go for two. This is obviously oversimplified, but I'd be interested to see what other factors or assumptions you might include that would swing the advantage towards going for two.
Let's simplify as a starting point. Assume that the probability of successfully converting a 2-pt conversion is 0.40, the probability of successfully converting a PAT is 1.00, and the probability of winning if the game goes to OT is 0.50.
If you kick the PAT, you go up by 8 with 5:00 left. Assume the other team uses most of that time to drive down the field and score a TD:
20% of the time, they'll successfully convert the 2-pt try to tie, and then beat you in OT.
20% of the time, they'll successfully convert the 2-pt try to tie, but you'll beat them in OT.
60% of the time, they'll fail to convert the 2-pt try and you win.
If you go for two:
60% of the time, you'll fail to convert the 2-pt try and will be up by 7 points. Assume they drive down the field and tie the game with a TD. 50% of the time they'll beat you in OT.
40% of the time, you'll successfully convert the 2-pt try and will be up by 9. Assume the best case scenario, that they are unable to make up a 9 point deficit in 5:00 and you win.
So based on this, you win 80% of the time you kick the PAT, and 70% of the time you go for two. This is obviously oversimplified, but I'd be interested to see what other factors or assumptions you might include that would swing the advantage towards going for two.
packersfan
Footballguy
Depends how good my defense is and how well it's playing that day. If it's the 49ers, for example, and they're throttling the other offense I'll kick the XP and put my trust in them. If it's the Ravens and facing RG3 I'm going for two.
Leroy Hoard
Footballguy
1
Marauder
Footballguy
I think this is the correct methodology for figuring it out but the percentages, especially the 2 pt conversion percentages for both you and your opponent, are going to vary greatly by team. There is no correct answer.
B-Deep
Footballguy
Saw someone reference this on Grantland
http://www.grantland.com/story/_/id/8731709/bill-barnwell-14-moments-make-smile-plus-rest-week-14-news
I had not considered it but I like the go for 2 option
http://www.grantland.com/story/_/id/8731709/bill-barnwell-14-moments-make-smile-plus-rest-week-14-news
I had not considered it but I like the go for 2 option
Ignoratio Elenchi
Footballguy
Right, but you need some kind of baseline to start from. We might determine that an "average" team would be better off kicking the PAT, 80-70%. From there, a coach would have to decide whether his particular situation dictates that the 2-pt conversion is the better option anyway. 70% and 80% are close enough that a coach needs to consider if he's got a great 2-pt play in his pocket, a great rushing QB, a winded defense, whatever that might swing the percentages in his favor. If, on the other hand, we did the math and it came out 80% to 10% or something, then a coach would probably be foolish to onsider it, no matter who his QB is. That's why it helps to have some kind of general baseline. I feel like sometimes people just say, "Impossible to determine, depends on the situation," without really giving it any thought. Like we could ask, "Should a team kick a FG when they're down by 1 at their opponent's 10 yard line with 10 seconds left in the game?" And some people would say, "There's no way to answer that, it depends on your kicker and your opponent's defense and the weather and who has the momentum and..."
Ignoratio Elenchi
Footballguy
Decent writeup but I think he's making the same mistake everyone else does in these types of discussions - he's overweighting the good that can come from going for 2, and overweighting the bad that can come from kicking the PAT. "Had the Ravens gone for two with the 27-20 lead, they would have gone up by nine points with less than five minutes to go."
No, they might have gone up by nine points. More often than not, they'll fail the try and go up by only seven points.
"As it turned out, of course, the Ravens kicked the extra point and weren't able to stop the Redskins from scoring a touchdown and getting a two-point conversion to tie the game. When Baltimore punted in overtime and allowed a long punt return, the game was over. Had the Ravens attempted the two-point conversion and failed, the game would have gone exactly the same way."
Right, as it turned out, Washington ended up making the 2-pt try and then winning in overtime. But a lot of times, Washington won't make the 2-pt try. And a lot of times even if they do, Baltimore will win in OT anyway. He seems to be saying that there'd be no difference between being up by 8 and being up by 7. That's obviously not the case.
Washington may be the best team in the league at 2-pt conversions, as the author believes, and still their conversion rate will be nowhere near the conversion rate for PATs. When you ignore the actual probabilities involved in each scenario, you end up creating narratives where it sounds like going for two is the smart play, a conclusion that may not be supported by properly considering all the possibilities.
Neil Beaufort Zod
Footballguy
If RG III is such a weapon going for two, why would you let them try it for the win? If you miss and they're down seven, who's to say they don't go for two and try for the win?
Ignoratio Elenchi
Footballguy
Right. I think his argument is that "coaches never do that," which is basically true. You just assume that a team down by 7 will kick the PAT to tie and go to OT, because that's what they (almost) always do. But if you're Shanahan, down by 1 after scoring a TD, and you think RG3 can convert a 2-pt try more than half the time, you're probably better off putting the ball in his hands rather than kicking the PAT and taking your chances in overtime. He's one of the few coaches in the league I might expect to actually do that if he was faced with that situation.
Because that never, ever happens.
Bills_Fan11
Footballguy
Yah, I think the "it depends..." line is usually a weak cop-out for defending status quo decisions; as if coaches have some superhuman "feel for the game" that fans are just too dumb to understand; when in reality they're just doing the conservative thing they've always done because that's what their mentor used to do and it's what all their peers do. Take the Saints kicking a FG with 1:50 left in the 1st half on 4th&2 from the NYG 7. What is the best case scenario you're hoping transpires the last 2 minutes of this half? You get the 3, stop a QB who is great in the 2minutes with your terrible defense, and net +3pts before halftime. Great. What about all the alternatives though?A) Get 3, give up 3.B) Get 3, give up 7.C) Go for it, get 0, give up 7.D) Go for it, get 0, give up 3.E) Go for it, get 0, give up 0.F) Go for it, get 7, give up 0.G) Go for it, get 7, give up 3.H) Go for it, get 7, give up 7. The fact that 'B' happened is irrelevant - this is not 20/20 hindsight. Note that 'C' is very unlikely given the Giants need to go 93 yards for a TD. Also note that getting a 1st down without a TD takes valuable time off the clock and/or makes NYG use time outs. Also note that 'D' is preferable to 'B' as -3 is better than -4; while 'A' and 'E' are the same. There is absolutely no way kicking isn't a terrible decision there, but ya know, coaches have to be coaches and do what they're used to.
Last edited by a moderator:
Ignoratio Elenchi
Footballguy
Are you sure about that? What probabilities are you using to make this statement?1) If the Saints kick the FG, they then kick off to the Giants with 1:50 left. The Giants get the ball at, on average, their own 23 yard line or so. They need to pick up ~44 yards to get into FG range, or 77 yards for a TD. What is the probability that the Giants score 0 points in that scenario?
2) If the Saints go for it, a number of things can happen. For one, they can fail to pick up the first down. Then the Giants get the ball at their own 6 or so. They need to pick up ~61 yards to get into FG range, or 94 yards for a TD. They don't get the three points but the Giants are also much less likely to put points on the board as well before halftime. What is the probability of this occurring?
3) Or, they can score a TD on the 4th down play, and then they're in the same position as (1) except they're +7 points instead of +3. What is the probability of this occurring?
4) Or, they can pick up the first down, run some more time off the clock, and then score a TD, which is better than (1). What is the probability of this occurring?
5) Or, they can pick up the first down, run some more time off the clock, and then kick a FG, which is better than (1). What is the probability of this occurring?
6) Or, they can pick up the first down, run some more time off the clock, and then turn it over on downs/clock runs out/whatever. Both teams score 0. What is the probability of this occurring?
Or, some other low-probability event can occur (Saints fumble, Giants return it for a TD; Saints kick FG, kickoff to Giants, Giants fumble the return, Saints recover and kick another FG; etc.) Factor those in however you'd like.
Once you have all those probabilities and their associated net point differential, you can weight them all and demonstrate whether or not kicking the FG in that situation is the right call, as you claim. I'm not 100% sure but I think David Romer would probably disagree with you. It's certainly not clear that kicking the FG was "absolutely the right move." At least show your work if you're going to make a claim like that.
SacramentoBob
Footballguy
.'zandbak said:
Bills_Fan11
Footballguy
I thought somebody would bring up the Denver-SD Hoculi game, or a late season Minny-NO game when Tice was on his way out the door...forgot about that one. I think the Bears tried it vs the Packers in the 90's too.But I mean really...that's 4 out of God knows how many (tying PAT's obviously aren't memorable). It's got to be something like ~98%.
Last edited by a moderator:
Bills_Fan11
Footballguy
Yeah, a historically bad defense which may or may not be in a prevent stopping a historically great 2minute drill QB from getting into FG range "being the most likely scenario by far" is a bit, ummm...questionable I would say. Particularly when you weigh it against the likelihood of a a good offense picking up 2yards and all the benefits associated with that (and/or field position if the 4th down fails)...But asking someone who is talking about a different game to show their work might be a bit much.
Let's do this with variables:
Assume that the probability of your team successfully converting a 2-pt conversion is A, the prob of the other team converting is B, the probability of successfully converting a PAT is 1.00, and the probability of winning if the game goes to OT is 0.50.
If you kick the PAT, you go up by 8 with 5:00 left. Assume the other team uses most of that time to drive down the field and score a TD:
(0.5B) of the time, they'll successfully convert the 2-pt try to tie, and then beat you in OT.
(0.5B) of the time, they'll successfully convert the 2-pt try to tie, but you'll beat them in OT.
(1-B) of the time, they'll fail to convert the 2-pt try and you win.
Total prob of winning = 0.5B + (1-B) = 1 - 0.5B
If you go for two:
1-A of the time, you'll fail to convert the 2-pt try and will be up by 7 points. Assume they drive down the field and tie the game with a TD. 50% of the time they'll beat you in OT.
A of the time, you'll successfully convert the 2-pt try and will be up by 9. Assume the best case scenario, that they are unable to make up a 9 point deficit in 5:00 and you win.
Total = 0.5(1-A) + A = 0.5+0.5A
For it to be worth it to go for two 0.5(1+A) > 1-0.5B ==> A > 1-B
So if you assume the probs are equal, the break even point is 0.5.
Assume that the probability of your team successfully converting a 2-pt conversion is A, the prob of the other team converting is B, the probability of successfully converting a PAT is 1.00, and the probability of winning if the game goes to OT is 0.50.
If you kick the PAT, you go up by 8 with 5:00 left. Assume the other team uses most of that time to drive down the field and score a TD:
(0.5B) of the time, they'll successfully convert the 2-pt try to tie, and then beat you in OT.
(0.5B) of the time, they'll successfully convert the 2-pt try to tie, but you'll beat them in OT.
(1-B) of the time, they'll fail to convert the 2-pt try and you win.
Total prob of winning = 0.5B + (1-B) = 1 - 0.5B
If you go for two:
1-A of the time, you'll fail to convert the 2-pt try and will be up by 7 points. Assume they drive down the field and tie the game with a TD. 50% of the time they'll beat you in OT.
A of the time, you'll successfully convert the 2-pt try and will be up by 9. Assume the best case scenario, that they are unable to make up a 9 point deficit in 5:00 and you win.
Total = 0.5(1-A) + A = 0.5+0.5A
For it to be worth it to go for two 0.5(1+A) > 1-0.5B ==> A > 1-B
So if you assume the probs are equal, the break even point is 0.5.
SacramentoBob
Footballguy
Dom Capers did it with Houston.
