Going For Two When Down By 14 (1 Viewer)

[Jeff Pasquino]

This came up over the holidays when I was talking football (shocking, I know) with the in-laws.



Should you always go for two when down by 14?

Of course there was this thread from two years ago: LINK

Which I believe showed that since you were getting success on the 2-pointer at <50% you shouldn't do it.



I'm going to push this again, and say you should.



I'll even go another step, and say a team should ALWAYS go for two points - at ANY point of the game.

Here's why:

1. The assumption here is that the goal is to getting a tie. That's a fatal flaw in the argument. You need 15 to win the game, so that's the real goal. Actually, it is AT LEAST 15" to win the game.

2. Playing it safe and kicking twice gets you a tie anyway (98% of the time), so there's no incentive to go for two ever if you play for a tie.

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.Therefore the proper / safe play is to kick the ~99% likely extra point.

Now I'm going to look at it this way.

Assuming that you will get the two TDs, there are 3 outcomes:

Go for 2 and make it after first TD - THEN KICK THE XP AFTER TD #2
Go for 2 and miss after first TD, fail, Go again after 2nd and make it
Go for 2 and miss after first TD, fail, Go again after 2nd and miss itThe bold part of #1 is the key.

In Scenario 1, you have a 42.57% of getting 15 points (and taking the lead). (43% x 99% chance of making the extra point)

In Scenario 2, you have a 24.51% chance of getting exactly 14 points (57% of missing the first 2-PT, then multiply by the 43% for the successful second 2-PT try).

In Scenario 3, you have a 32.49% chance of getting only 12 points and being down by 2. (57% x 57% for two misses).

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

That's over 10% of the time where you take the lead by going for two after the first score.

The other part of the equation is that you will need a third score to win the game regardless (either with time remaining in regulation or in OT). That leads to further digging on just how much worse Scenario 3 is (down by 2) than Scenario 1 (tied).

Ok, fire away.

 

[Sigmund Bloom]

At least one flaw I see in the argument is that you are going to use your best (highest expected rate of success) play on the first attempt. If it doesn't make it, then I think whatever you run on the second attempt is much less likely to work - you could see this in Steelers-Jags - the Jags were all over the fade after it had worked on the first attempt, only to be overturned by penalty.

In other words, the expected rate of success on the 2nd attempt has to be lower when you miss on the first attempt. I would argue for kicking the XP on the first TD and saving your #1 play for the 2 after the second TD if you really want to go for the win, but then the numbers are not as kind - 57% chance you lose the game on that play.

 

[Kirby]

I'll test this theory in my next season of madden.

 

[Jeff Pasquino]

At least one flaw I see in the argument is that you are going to use your best (highest expected rate of success) play on the first attempt. If it doesn't make it, then I think whatever you run on the second attempt is much less likely to work - you could see this in Steelers-Jags - the Jags were all over the fade after it had worked on the first attempt, only to be overturned by penalty.In other words, the expected rate of success on the 2nd attempt has to be lower when you miss on the first attempt. I would argue for kicking the XP on the first TD and saving your #1 play for the 2 after the second TD if you really want to go for the win, but then the numbers are not as kind - 57% chance you lose the game on that play.

Hi Bloom,Yeah, I did think of this, but I didn't want to complicate the math.Even if you assume 10% less success on play #2, it is still favorable to go for it.The second 2-PT play would have to be absolutely terrible to change the math here (~75% unsuccessful, or about twice as bad as the first one), and even then it is close to a wash.Assume that you can do better than 30% (say 33%) on your second best play and the math still says to go for two after the first TD.Keep them coming, I'm interested if there's a big hole I didn't see, but I can't find it.

 

[bridaniel]

I think the key variable is the 2 point success rate. A small change makes a BIG difference. If, for example, the success rate is 37% instead of 43%, and the odds of winning a game in overtime is 50%, then under the "go for 2 no matter what" theory, you have a 48.47% chance of winning the game (all other earlier assumptions unchanged).

Chance of winning with "go for 2 no matter what" theory

(.37*.99)+[(.37*.01)+(.63*.37)]*.5 = .4847

.37*.99 is the chance of winning in regulation

.37*.01 is the chance you make the 2 point try and miss the XP --> tied and heading to OT

.63*.37 is the chance you miss the first 2 point try and make the second --> tied and heading to OT

However, if you were to kick the first XP, and if successful, then kick the second XP (if not, go for 2), then you have a 49.19% chance of winning the game.

Chance of winning if kick first, then kick again or go for 2 if the first kick failed

[(.99*.99)+(.01*.37)]*.5 = .4919

Thoughts?

 

[3nOut]

First of all, this all has to go on the assumption that you WILL score twice. Second, I believe there are too many situations (are you expected to win or are you an underdog, weather conditions, confidence in your defense, time remaining, etc.) where there are exceptions and it may be better to decide to KICK it.

Lastly, you state the following:

I'm going to push this again, and say you should.

I'll even go another step, and say a team should ALWAYS go for two points - at ANY point of the game.

1. You are already going against your own convention. You say ALWAYS go for two points, but then you say KICK on the second try. There is an exception already.2. Maybe most coaches would prefer to go for the tie and win in overtime, which is a 50% chance you win the coin toss. 50% >42.57%. This also assumes you think your team is good enough to win on its first drive. And if you are, why were you playing from behind?

 

[Jeff Pasquino]

I think the key variable is the 2 point success rate. A small change makes a BIG difference. If, for example, the success rate is 37% instead of 43%, and the odds of winning a game in overtime is 50%, then under the "go for 2 no matter what" theory, you have a 48.47% chance of winning the game (all other earlier assumptions unchanged).Chance of winning with "go for 2 no matter what" theory(.37*.99)+[(.37*.01)+(.63*.37)]*.5 = .4847.37*.99 is the chance of winning in regulation.37*.01 is the chance you make the 2 point try and miss the XP --> tied and heading to OT.63*.37 is the chance you miss the first 2 point try and make the second --> tied and heading to OTHowever, if you were to kick the first XP, and if successful, then kick the second XP (if not, go for 2), then you have a 49.19% chance of winning the game.Chance of winning if kick first, then kick again or go for 2 if the first kick failed[(.99*.99)+(.01*.37)]*.5 = .4919Thoughts?

Absolutely the 37% vs. 43% matters.To boil down your thought and to place it aside mine:Scenario 1: Win - 36.63% chanceScenario 2: Tied - 23.68%Scenario 3: Loss (missed twice) = 39.69%So basically if the odds of making the first 2-pointer are higher than the odds of missing two in a row, you should go for two after the first touchdown.In this case, 37% ~ 36.63% < 39.69%, so don't go for it.So now it begs the question - what does the success rate have to be for this to make sense?Assuming 99% XP success, the answer for breakeven (A) is :.99xA = (1-A)*(1-A) A = 38.37%So as long as the odds of making a 2-point conversion in the NFL is > 38.4%, you should do this.

 

[Jeff Pasquino]

First of all, this all has to go on the assumption that you WILL score twice. Second, I believe there are too many situations (are you expected to win or are you an underdog, weather conditions, confidence in your defense, time remaining, etc.) where there are exceptions and it may be better to decide to KICK it.

Lastly, you state the following:

I'm going to push this again, and say you should.

I'll even go another step, and say a team should ALWAYS go for two points - at ANY point of the game.

1. You are already going against your own convention. You say ALWAYS go for two points, but then you say KICK on the second try. There is an exception already.2. Maybe most coaches would prefer to go for the tie and win in overtime, which is a 50% chance you win the coin toss. 50% >42.57%. This also assumes you think your team is good enough to win on its first drive. And if you are, why were you playing from behind?

The absolute statement is that you ALWAYS go for 2 after a touchdown when you trailed by 14 before the TD. Not EVERY touchdown.As for the 42.57%, you win 42.57% of the time PRIOR TO over time. You didn't calculate the 50% of the time where you win when the score is tied and you head to OT. So the true winning percentage by that math is 42.57% + half of 32.49%, or 58.815%, vs. a 41.185% losing percentage.

 

[Captain Spaulding]

Down by 14...no matter how much time is left in the game, you kick extra points. You need 2 TD's to tie and that is the easiest way to get there.

The Steelers were down by 5 points to Jax when Pitt scored a TD midway into the 4th Quarter and tried they're 1st 2 pt attempt. Whole different logic.

 

[Jeff Pasquino]

Down by 14...no matter how much time is left in the game, you kick extra points. You need 2 TD's to tie and that is the easiest way to get there.The Steelers were down by 5 points to Jax when Pitt scored a TD midway into the 4th Quarter and tried they're 1st 2 pt attempt. Whole different logic.

The statement of fact that 2 TDs plus extra points is the easiest way to get there might be correct, but if the object is to win the game then you might be wrong in not approaching the goal differently.I'm not bringing up specific examples - we're keeping this generic. You are the head coach of a team that trails by 14, and you just scored a TD and you believe you're likely to get another one. Go for two or not?Based on what I've read so far, everything points me against convention and to go for 2.

 

[JBPH]

If the other team scores a TD:

18% of the time, they will have the option to match your 2pt attempt and restore the 14 point lead. (Original .42 * .42)

24% of the time, they will try to match you and miss. Your deficit will have been improved. (Original .42 * .58 miss)

58% of the time, you will have missed and they will simply kick an extra point and improve themselves from the position before the TDs happened. Now they are up 15.

24% of the time, going for two improves your standing.

58% of the time, you make things worse for your team.

I think your reasoning would work if it were guaranteed that there would be no other scoring in the game, though, other than your 2 TDs.

 

[Modog814]

If the other team scores a TD:18% of the time, they will have the option to match your 2pt attempt and restore the 14 point lead. (Original .42 * .42)24% of the time, they will try to match you and miss. Your deficit will have been improved. (Original .42 * .58 miss)58% of the time, you will have missed and they will simply kick an extra point and improve themselves from the position before the TDs happened. Now they are up 15.24% of the time, going for two improves your standing.58% of the time, it does not.I think your reasoning would work if it were guaranteed that there would be no other scoring in the game, though, other than your 2 TDs.

Agreed. If you could be certain (or near certain) that the other team isn't going to score then this strategy seems as good as any. Even if you could gaurantee holding them to a field goal (assuming that you'd have enough time to get 2 scores back) it seems like it would work.

 

[Sigmund Bloom]

Pasquino,

I think your analysis is very narrow and ignores the context of the game. You treat the two point conversion like it is strictly a matter of probability - as if when a team goes for two, the ref just gets out a 100 sided die, and if the number rolled is 1-43, its good, and if its 44-100, its not. Luckily for us, there is much more going on than mere probability on the football field.

Another thing that needs to be considered is momentum - by going for two after the first TD, you run the risk of giving the opponent momentum right after you had just seized it. At home, it can take the wind out of the sails of your crowd. On the road, it can cause a hostile crowd to get fired up right after you silenced them. It can similarly affect the players. If you go for 2 and fail, its not like the game will go exactly as it was going to, except you have one less point on the board. The emotional mood of the game does change in the wake of the stop by the opposing defense. Making the two can increase your momentum, but more often than not, you will fail and set your team back.

The reality is that 43% number is a very broad number from that contains a variety of scenarios and matchups. A coach has to look at their personnel, how their team is executing in certain sets, whether certain key players are hot or cold or tired or hurt, the general mood of the team, how they executed certain plays in practice and any other number of scenarios to gauge whether going for two is the right call when its not completely necessary. You are trying to make a very complex decision much simpler than it actually is.

I think there is a great case for going for two at the end of the game when an extra point can tie, saving your A 2 pt play for that play, and asking your guys to summon up everything and win that play to win the game, especially when your D has not been stopping the other team, and you believe in your guys ability to execute. Other than that, I think teams generally go for 2 about as much as they should. If your theory was correct it wouldnt be so common for teams to get hurt badly by missed extra points - its easy to make them up right? Instead we often see teams chasing those points and that missed single point play a crucial role in the outcome. Take the free point unless you have no other choice.

 

[Tau837]

1. Is Jeff Pasquino really Steve Spurrier?

2. What if the team that scores gets a penalty, making the 2 pt conversion occur from the 12 yard line?

3. This reminds me of some Football Outsiders, or Football Prospectus, or similar article that basically made the case that it mathematically makes sense to approach playing offense as if your team will always go for it on 4th down. While the math may work, I'm not convinced it is the right approach.

 

[Jeff Pasquino]

Pasquino,I think your analysis is very narrow and ignores the context of the game. You treat the two point convention like it is strictly a matter of probability - as if when a team goes for two, the ref just gets out a 100 sided die, and if the number rolled is 1-43, its good, and if its 44-100, its not. Luckily for us, there is much more going on than mere probability on the football field.Another thing that needs to be considered is momentum - by going for two after the first TD, you run the risk of giving the opponent momentum right after you had just seized it. At home, it can take the wind out of the sails of your crowd. On the road, it can cause a hostile crowd to get fired up right after you silenced them. It can similarly affect the players. If you go for 2 and fail, its not like the game will go exactly as it was going to, except you have one less point on the board. The emotional mood of the game does change in the wake of the stop by the opposing defense. Making the two can increase your momentum, but more often than not, you will fail and set your team back.The reality is that 43% number is a very broad number from that contains a variety of scenarios and matchups. A coach has to look at their personnel, how their team is executing in certain sets, whether certain key players are hot or cold or tired or hurt, the general mood of the team, how they executed certain plays in practice and any other number of scenarios to gauge whether going for two is the right call when its not completely necessary. You are trying to make a very complex decision much simpler than it actually is.

Bloom,Actually I'm doing exactly the opposite of what you're saying. I'm challenging conventional thought of taking the automatic extra pont and asking everyone to think about the possibility that going for two opens up a better chance for a team to possibly win a game.Is there more to it than probability? Of course. But if the numbers imply that common practice is flawed, why not discuss the validity of adopting a new thought process?43% is a broad number, but you could also make the case that a team should capitalize on their current momentum (scoring a TD) and force the defense to stay out on the field and make them stop them again. If you've just driven 80 yards and picked up 5 yards a carry, including a red zone touchdown, why not punish that weakened defense and get two points? Again, ultimately the goal is to win the game, and just like in college overtime you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game. Aggressive decisions earlier in the game could afford you a more conventional and conservative approach for a victory on the next possession.Probability is probability and nothing is assured, but when the overall odds say you win nearly 60% of the time if you go for 2 after the first score, you cannot ignore it so easily.

 

[Sigmund Bloom]

If you've just driven 80 yards and picked up 5 yards a carry, including a red zone touchdown, why not punish that weakened defense and get two points? Again, ultimately the goal is to win the game, and just like in college overtime you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

I think thats a fine argument for making it a better decision to go for 2 (capping a long, grinding drive), but that's exactly what Im talking about - you only know whether it is the right call in the context of the game. You are trying to make a general rule that is ALWAYS correct (go for 2 down for 14 on the first TD), and I just don't think football can be boiled down to rules that are that simple. You know what you need to do on the final possession with just as much clarity if you take the XP as you do if you make or miss the 2 pt, I dont find that argument compelling as a reason to go for two.

 

[Jeff Pasquino]

If you've just driven 80 yards and picked up 5 yards a carry, including a red zone touchdown, why not punish that weakened defense and get two points? Again, ultimately the goal is to win the game, and just like in college overtime you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

I think thats a fine argument for making it a better decision to go for 2 (capping a long, grinding drive), but that's exactly what Im talking about - you only know whether it is the right call in the context of the game. You are trying to make a general rule that is ALWAYS correct (go for 2 down for 14 on the first TD), and I just don't think football can be boiled down to rules that are that simple. You know what you need to do on the final possession with just as much clarity if you take the XP as you do if you make or miss the 2 pt, I dont find that argument compelling as a reason to go for two.

The last comment is the one that I am starting to question.Are you trying to play to win or play to tie?

You know that if you made the 2-point play you can now win in regulation and you haven't eliminated your ability to send the game to overtime - just hurt your odds of forcing that tie in the first place.

Even if you say that the XPs are locks (100%), you only have a 50% chance of winning the game in overtime.

If you are playing for the win, the 2-point play after the first score - statistically - gives you a decided leg up on winning the game.

 

[Anarchy99]

The issue I had in the old thread (and most likely now as well) is that the averages are just that. Some teams may be more able to (or better consistuted) to get 2-point conversions and others may not. Similarly, some defenses may be more prone to giving up 2-point converstions and others may not give up many.

I HAVE NOT looked into the actual results, but one would think that if the Pats are the ones trying to get a 2-point conversion on the road against the Lions, then sure that probably is a great option. But if it's the Falcons in a snow storm at Lambeau trying it I might not be so quick to recommend it.

 

[CalBear]

The issue I had in the old thread (and most likely now as well) is that the averages are just that. Some teams may be more able to (or better consistuted) to get 2-point conversions and others may not. Similarly, some defenses may be more prone to giving up 2-point converstions and others may not give up many.I HAVE NOT looked into the actual results, but one would think that if the Pats are the ones trying to get a 2-point conversion on the road against the Lions, then sure that probably is a great option. But if it's the Falcons in a snow storm at Lambeau trying it I might not be so quick to recommend it.

I think recommending it at all is a enough of a departure from convention that it's noteworthy, even if there are cases where it doesn't make sense.Surely I would expect a team like San Diego with Tomlinson and Gates to have a high enough probability of getting 2.5 yards to make the math work out. One other factor to consider, though, is that the score differential will affect how the other team plays. For example a team that gets the ball at their own 20 in a tie game with 40 seconds left might take it easy on play calling and be content for OT, while a team down by one in a tie game with 40 seconds left will be going all-out to score. So scoring 15 points doesn't necessarily mean you win; it depends on the game situation.

 

[Anarchy99]

I think the other issue was that many NFL coaches would play for the tie because if they did something unconventional in a crucial moment they could lose their job. I don't agree with that, but that is the practical reality why some folks will play not to lose and hope for the best in OT.

 

[BroadwayG]

Umm, what if they score a field goal? Too many variables to consider. If you're in onside kick mode, you may be on to something.

Making the assumption that the two TDs you score will be the only points scored for the entire rest of the game is pretty radical.

 

[toppshelff]

Somebody was once bold enough to take a strictly statistical look at baseball, and ignore the "intangibles". That seemed to work out pretty well.

I have found too often that people look at this as black or white. Either they ignore the pure probabilities and play completely by the gut (whether risky or safe) while others view the numbers as the all mighty.

I imagine that the best coaches would study the numbers religiously, understand them fully, then make good decisions based on the numbers and include all of the less objective parameters of the game, team, momentum, etc. in their decision.

I believe that the current NFL coaches don't respect or understand the numbers and play it by gut. Unfortunately, that gut is usually "play it safe, don't get fired for going for it and failing".

I've looked at the math before and agree that this is absolutely worth reconsideration. Makes me wonder about what other calls should be analysed. For instance, if you are down by 10 points nearing the end of the game, how do you decide to go for it on on 4th down versus kicking a field goal? Obviously dependent on the distance to go for 1st down, distance for TD, time on clock and field goal distance, but I imagine that a simple formula could be made to help make the decision. Again, the coach needs to make the call on the field with all of the facts, subjective and objective.

 

[Doug Drinen]

This came up over the holidays when I was talking football (shocking, I know) with the in-laws.



Should you always go for two when down by 14?

Of course there was this thread from two years ago: LINK

This, from a little over a year ago, is a more relevant link

 

[3nOut]

Probability is probability and nothing is assured, but when the overall odds say you win nearly 60% of the time if you go for 2 after the first score and you succeed the first time, you cannot ignore it so easily.

edited to add bolded part.

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

Where do you get 60%? And I thought your objective was to win, not NOT to lose or tie. According to your own statistics, if you add the first two scenarios together, you get a 57% chance (more than half) that you DON'T WIN (you either tie or lose by 2), and only 43% chance that you win. Also, there's only a 25% chance that you tie thegame and go into overtime, when you can have a 99% chance for a tie just by kicking.

 

[Pai Mei]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.Therefore the proper / safe play is to kick the ~99% likely extra point.

Now I'm going to look at it this way.

Assuming that you will get the two TDs, there are 3 outcomes:

Go for 2 and make it after first TD - THEN KICK THE XP AFTER TD #2
Go for 2 and miss after first TD, fail, Go again after 2nd and make it
Go for 2 and miss after first TD, fail, Go again after 2nd and miss itThe bold part of #1 is the key.

In Scenario 1, you have a 42.57% of getting 15 points (and taking the lead). (43% x 99% chance of making the extra point)

In Scenario 2, you have a 24.51% chance of getting exactly 14 points (57% of missing the first 2-PT, then multiply by the 43% for the successful second 2-PT try).

In Scenario 3, you have a 32.49% chance of getting only 12 points and being down by 2. (57% x 57% for two misses).

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

That's over 10% of the time where you take the lead by going for two after the first score.

I'm not really sure how one derives that math. For example you say there is ~43% chance to convert. How solid is that? How much does that take into effect the quality of the OL, DL, the QB, the weather, how tired your OL is, how healthy your OL is, etc. Its really ~43% independent of these factors? I'm skeptical because to me football is blocking and tackling. I'd think WHO you have up front to block can greatly impact your odds.

 

[CalBear]

Probability is probability and nothing is assured, but when the overall odds say you win nearly 60% of the time if you go for 2 after the first score and you succeed the first time, you cannot ignore it so easily.

edited to add bolded part.

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

Where do you get 60%? And I thought your objective was to win, not NOT to lose or tie. According to your own statistics, if you add the first two scenarios together, you get a 57% chance (more than half) that you DON'T WIN (you either tie or lose by 2), and only 43% chance that you win. Also, there's only a 25% chance that you tie thegame and go into overtime, when you can have a 99% chance for a tie just by kicking.

The chance for Tie+Win under the go-for-two scenario is 24.51%+42.57%=67.08%. That is, 67% of the time you either win or tie. If you tie, you revert to the other scenario; 12.25% of the time you tie and win in OT, 12.25% of the time you tie and lose in OT, so you wind up winning 54.82% of the time, as opposed to 50% if you go for the tie. (50/50 chance in OT).

 

[Pai Mei]

I think one could argue that if the head coach feels his OL is dominating the DL, he might as well press that advantage by going for 2 every time. In that scenario the conversion rate might be higher than ~43% though.

 

[Sigmund Bloom]

If you've just driven 80 yards and picked up 5 yards a carry, including a red zone touchdown, why not punish that weakened defense and get two points? Again, ultimately the goal is to win the game, and just like in college overtime you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

I think thats a fine argument for making it a better decision to go for 2 (capping a long, grinding drive), but that's exactly what Im talking about - you only know whether it is the right call in the context of the game. You are trying to make a general rule that is ALWAYS correct (go for 2 down for 14 on the first TD), and I just don't think football can be boiled down to rules that are that simple. You know what you need to do on the final possession with just as much clarity if you take the XP as you do if you make or miss the 2 pt, I dont find that argument compelling as a reason to go for two.

The last comment is the one that I am starting to question.Are you trying to play to win or play to tie?

You know that if you made the 2-point play you can now win in regulation and you haven't eliminated your ability to send the game to overtime - just hurt your odds of forcing that tie in the first place.

Even if you say that the XPs are locks (100%), you only have a 50% chance of winning the game in overtime.

If you are playing for the win, the 2-point play after the first score - statistically - gives you a decided leg up on winning the game.

Your statement was:

you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

If you take the XP, you know that you need to get a TD+2 to win, and a TD+1 to tie. I dont see how going for 2 helps you "know what you need to do" any more than goes for 1 does.

 

[Bri]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.

Not right manfirst is 2pt in general

For the 2nd, you made it more specific, the second is a second 2pt attempt right after another, that doesn't happen as often and would be a different %.

 

[KnowledgeReignsSupreme]

Football commentary chart agrees with you (I disagreed before I read the chart since it seemed counterintuitive) assuming you're in the 2nd half. He doesn't have a 1st half 2pt chart, but he does discuss going for 2 (in general) in the 1st half by using the Dallas - Washington game from last year as an example.

BTW, the row you are looking for on the chart is -8 since that's the point differential at the time you're making the decision to go for 2.

 

[KnowledgeReignsSupreme]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.

Not right manfirst is 2pt in general

For the 2nd, you made it more specific, the second is a second 2pt attempt right after another, that doesn't happen as often and would be a different %.

Huh?

 

[Sigmund Bloom]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.

Not right manfirst is 2pt in general

For the 2nd, you made it more specific, the second is a second 2pt attempt right after another, that doesn't happen as often and would be a different %.

Huh?

its the same point i originally made to pasq, the expected success rate on the 2nd attempt is different than the overall average (and very likely lower, especially when the first was unsuccessful)

 

[KnowledgeReignsSupreme]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.

Not right manfirst is 2pt in general

For the 2nd, you made it more specific, the second is a second 2pt attempt right after another, that doesn't happen as often and would be a different %.

Huh?

its the same point i originally made to pasq, the expected success rate on the 2nd attempt is different than the overall average (and very likely lower, especially when the first was unsuccessful)

Why?

 

[Liquid Tension]

I apologize in advance that I have not done the math YET to verify what you are saying is right, but one factor that is very important is WHEN you take the lead. My take is having the lead at the end of the 2nd score is the key. In other words how much time is left on the clock when you take a lead? If you take the lead with 1 minute left in the game going for 2 is stupid when an extra point ties the game. The reason is that you are giving the other team 4 downs to only get a game winning FG.

Going for 2 is an option at the end of the game with NO time left on the clock because then it is a win or a loss. Although, the numbers would say that if you are not successful more than 50% of the time (assuming making the XP), then you should go into OT. Of course if the defenses are getting torched you might feel a greater than 50% chance. But under no circumstance should you go for 2 with time left on the clock (depending on timeouts).

 

[Michael Fox Fan]

Down by 14...no matter how much time is left in the game, you kick extra points. You need 2 TD's to tie and that is the easiest way to get there.The Steelers were down by 5 points to Jax when Pitt scored a TD midway into the 4th Quarter and tried they're 1st 2 pt attempt. Whole different logic.

The statement of fact that 2 TDs plus extra points is the easiest way to get there might be correct, but if the object is to win the game then you might be wrong in not approaching the goal differently.I'm not bringing up specific examples - we're keeping this generic. You are the head coach of a team that trails by 14, and you just scored a TD and you believe you're likely to get another one. Go for two or not?Based on what I've read so far, everything points me against convention and to go for 2.

I'm too lazy to read the entire thread. But basically, your "analysis" says that going for 2-pts after TD #1 = 42.57% chance of winning the game in regulation + 24.51% * 1/2 odds of winning in overtime = 54.82% overall chance of winning. [note that I'm assuming a 50-50 shot at winning in overtime]Using similar logic, by kick 2 XPs you have ~98% chance of reaching OT. 98% * 1/2 = 49% chance of winning.So again, using your assumptions you gain a 5-6% increase in odds of winning the game overall. [which frankly is the only thing coaches should care about]That said, your assumptions have a huge impact on this analysis. It doesn't take much of a shift....in odds of 2-pt conversion success...to make it a bad decision to go for 2.

 

[cstu]

Down by 14...no matter how much time is left in the game, you kick extra points. You need 2 TD's to tie and that is the easiest way to get there.The Steelers were down by 5 points to Jax when Pitt scored a TD midway into the 4th Quarter and tried they're 1st 2 pt attempt. Whole different logic.

The statement of fact that 2 TDs plus extra points is the easiest way to get there might be correct, but if the object is to win the game then you might be wrong in not approaching the goal differently.I'm not bringing up specific examples - we're keeping this generic. You are the head coach of a team that trails by 14, and you just scored a TD and you believe you're likely to get another one. Go for two or not?Based on what I've read so far, everything points me against convention and to go for 2.

While the goal is to win the game, there's more than math involved. I think you have to consider the psychology of the players as well. If you fail on the conversion and are down 8 points, it seems like it would be harder for the team to get fired up for the TD when they'd still have to get a conversion just to tie, especially knowing they already failed on the first one. Mathematically it makes sense to go for 2 down 14 but there's a good reason coaches don't do it.

 

[CalBear]

I apologize in advance that I have not done the math YET to verify what you are saying is right, but one factor that is very important is WHEN you take the lead. My take is having the lead at the end of the 2nd score is the key. In other words how much time is left on the clock when you take a lead? If you take the lead with 1 minute left in the game going for 2 is stupid when an extra point ties the game. The reason is that you are giving the other team 4 downs to only get a game winning FG.

Going for 2 is an option at the end of the game with NO time left on the clock because then it is a win or a loss. Although, the numbers would say that if you are not successful more than 50% of the time (assuming making the XP), then you should go into OT. Of course if the defenses are getting torched you might feel a greater than 50% chance. But under no circumstance should you go for 2 with time left on the clock (depending on timeouts).

I don't see how having time left on the clock makes any difference. Whether you're tied or ahead by 1, you lose if the other team goes down the field and scores.

 

[JBPH]

1. The assumption here is that the goal is to getting a tie. That's a fatal flaw in the argument. You need 15 to win the game, so that's the real goal.

The fatal flaw continues to be that there is absolutely no guarantee you need 15 to win the game. You need one more than your opponent, which may or may not be 15.

 

[Jeff Pasquino]

The issue I had in the old thread (and most likely now as well) is that the averages are just that. Some teams may be more able to (or better consistuted) to get 2-point conversions and others may not. Similarly, some defenses may be more prone to giving up 2-point converstions and others may not give up many.I HAVE NOT looked into the actual results, but one would think that if the Pats are the ones trying to get a 2-point conversion on the road against the Lions, then sure that probably is a great option. But if it's the Falcons in a snow storm at Lambeau trying it I might not be so quick to recommend it.

I agree with you to a point, but at some point you have to recognize that the math is screaming at you that, if you attain at least the NFL average, you have a much higher chance of winning if you go for two. If that's the case, you should be going for two more often in this scenario and adapt to it (i.e. get more plays for 2-pt scenarios).

 

[Jeff Pasquino]

This came up over the holidays when I was talking football (shocking, I know) with the in-laws.



Should you always go for two when down by 14?

Of course there was this thread from two years ago: LINK

This, from a little over a year ago, is a more relevant link

Excellent link Doug, thanks.The first post there summarized what I was trying to originally say - that the math says to go for 2, and that your odds of winning do go up.

 

[Jeff Pasquino]

1. Is Jeff Pasquino really Steve Spurrier?2. What if the team that scores gets a penalty, making the 2 pt conversion occur from the 12 yard line?3. This reminds me of some Football Outsiders, or Football Prospectus, or similar article that basically made the case that it mathematically makes sense to approach playing offense as if your team will always go for it on 4th down. While the math may work, I'm not convinced it is the right approach.

I'm questioning it as well in general, but the numbers are screaming go for it. Aside from "gut instinct", which may be wrong (maybe because we never see this happen), why wouldn't you agree with the odds?Part of the point of this thread is to question conventional thinking and see if it really does make sense to try the 2-pt play.

 

[Jeff Pasquino]

Probability is probability and nothing is assured, but when the overall odds say you win nearly 60% of the time if you go for 2 after the first score and you succeed the first time, you cannot ignore it so easily.

edited to add bolded part.

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

Where do you get 60%? And I thought your objective was to win, not NOT to lose or tie. According to your own statistics, if you add the first two scenarios together, you get a 57% chance (more than half) that you DON'T WIN (you either tie or lose by 2), and only 43% chance that you win. Also, there's only a 25% chance that you tie thegame and go into overtime, when you can have a 99% chance for a tie just by kicking.

As for the 42.57%, you win 42.57% of the time PRIOR TO over time. You didn't calculate the 50% of the time where you win when the score is tied and you head to OT. So the true winning percentage by that math is 42.57% + half of 32.49%, or 58.815%, vs. a 41.185% losing percentage.

This is where the "nearly 60%" comes from.50% of 32.49% (50% chance of winning if tied - heading to OT) plus the 42.57% (odds you are up by 1).

ETA: Whoops. OK - there's the error. It isn't 50% of 32.49%, it is 50% of 24.51%.

The correct answer is 42.57% + 12.505% = 55=%. So you win 10% more often (55% vs. 45%).

Sorry for the confusion.

 

[Jeff Pasquino]

Probability is probability and nothing is assured, but when the overall odds say you win nearly 60% of the time if you go for 2 after the first score and you succeed the first time, you cannot ignore it so easily.

edited to add bolded part.

So the sum of the scenarios:

24.51% chance you tied the game.

32.49% chance you trail by two.

42.57% chance you are ahead by one.

Where do you get 60%? And I thought your objective was to win, not NOT to lose or tie. According to your own statistics, if you add the first two scenarios together, you get a 57% chance (more than half) that you DON'T WIN (you either tie or lose by 2), and only 43% chance that you win. Also, there's only a 25% chance that you tie thegame and go into overtime, when you can have a 99% chance for a tie just by kicking.

The chance for Tie+Win under the go-for-two scenario is 24.51%+42.57%=67.08%. That is, 67% of the time you either win or tie. If you tie, you revert to the other scenario; 12.25% of the time you tie and win in OT, 12.25% of the time you tie and lose in OT, so you wind up winning 54.82% of the time, as opposed to 50% if you go for the tie. (50/50 chance in OT).

Correct / close enough - nice catch. Still says go for 2, so I think we just need to accept the math at this point and debate the "Gut instinct" now.

 

[Jeff Pasquino]

If you've just driven 80 yards and picked up 5 yards a carry, including a red zone touchdown, why not punish that weakened defense and get two points? Again, ultimately the goal is to win the game, and just like in college overtime you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

I think thats a fine argument for making it a better decision to go for 2 (capping a long, grinding drive), but that's exactly what Im talking about - you only know whether it is the right call in the context of the game. You are trying to make a general rule that is ALWAYS correct (go for 2 down for 14 on the first TD), and I just don't think football can be boiled down to rules that are that simple. You know what you need to do on the final possession with just as much clarity if you take the XP as you do if you make or miss the 2 pt, I dont find that argument compelling as a reason to go for two.

The last comment is the one that I am starting to question.Are you trying to play to win or play to tie?

You know that if you made the 2-point play you can now win in regulation and you haven't eliminated your ability to send the game to overtime - just hurt your odds of forcing that tie in the first place.

Even if you say that the XPs are locks (100%), you only have a 50% chance of winning the game in overtime.

If you are playing for the win, the 2-point play after the first score - statistically - gives you a decided leg up on winning the game.

Your statement was:

you want to know what you need to do on your next (and possibly final) possession to win or at worst tie he game.

If you take the XP, you know that you need to get a TD+2 to win, and a TD+1 to tie. I dont see how going for 2 helps you "know what you need to do" any more than goes for 1 does.

This is why you need to go for 2 the first time - if you miss it, you have the opportunity to tie the game after then 2nd TD with a 2-PT conversion and take your chances in OT. If you settle for the XP after the first TD, you miss the chance to win (~42% of the time). Yes 57+% of the time you'll be down by 8, but you took the chance of greatly increasing your odds of winning.

If you do as you state (kick XP first, then decide to go for 2 if you want to win then) the odds are now against you (43% vs 57%), so then the numbers say don't do it. That's the difference. You gave up on the possibility of getting the 2-PTer to tie, which is a realistic scenario even if it is only your second best play for it.

The difference lies in the opportunity of winning at a higher rate if you go for 2 after the first TD.

 

[Jeff Pasquino]

So here's the math, and why I think it is a flawed view:

2-PT Conversion Rate in the NFL is ~43% (based on older thread, may have changed).
That means you have to try again ~57% of the time after the second touchdown to tie, and that's again ~43% likely to score.

Not right manfirst is 2pt in general

For the 2nd, you made it more specific, the second is a second 2pt attempt right after another, that doesn't happen as often and would be a different %.

Huh?

its the same point i originally made to pasq, the expected success rate on the 2nd attempt is different than the overall average (and very likely lower, especially when the first was unsuccessful)

Why?

The assumption here is that your best 2-PT play would be exausted on the first try for 2-PTs.I believe that's flawed logic. If head coaches realized that they needed more 2-PT plays, they'd find some more plays and practice this scenario more often.

By the way, teams have > 1 play for most down/distance scenarios during the game to keep their options open. For example, 4th and 1. Some teams will put in the "big" formation and rush it, some will have the RB leap over the pile, some will QB sneak, and some even throw out of the formation. The point is, 4th and short comes up often so they practice that more often - because it is important. 2-PT plays are viewed as less important and therefore rarer, so they aren't prioritized.

If someone pointed out that they should reprioritize, playbooks and practices would adjust.

 

[Jeff Pasquino]

I apologize in advance that I have not done the math YET to verify what you are saying is right, but one factor that is very important is WHEN you take the lead. My take is having the lead at the end of the 2nd score is the key. In other words how much time is left on the clock when you take a lead? If you take the lead with 1 minute left in the game going for 2 is stupid when an extra point ties the game. The reason is that you are giving the other team 4 downs to only get a game winning FG.

Going for 2 is an option at the end of the game with NO time left on the clock because then it is a win or a loss. Although, the numbers would say that if you are not successful more than 50% of the time (assuming making the XP), then you should go into OT. Of course if the defenses are getting torched you might feel a greater than 50% chance. But under no circumstance should you go for 2 with time left on the clock (depending on timeouts).

LT,I wouldn't recommend going for 2 when down by 1 with 1 minute remaining. That's not the scenario here at all.

In this scenario, you're either down by 8 or 6 after the first TD (and you went for 2). Then you are tied or down by 2 after the second 6-pt TD, so you either kick the XP for the lead or go for 2 again to tie.

 

[Jeff Pasquino]

I'm too lazy to read the entire thread. But basically, your "analysis" says that going for 2-pts after TD #1 = 42.57% chance of winning the game in regulation + 24.51% * 1/2 odds of winning in overtime = 54.82% overall chance of winning. [note that I'm assuming a 50-50 shot at winning in overtime]Using similar logic, by kick 2 XPs you have ~98% chance of reaching OT. 98% * 1/2 = 49% chance of winning.So again, using your assumptions you gain a 5-6% increase in odds of winning the game overall. [which frankly is the only thing coaches should care about]That said, your assumptions have a huge impact on this analysis. It doesn't take much of a shift....in odds of 2-pt conversion success...to make it a bad decision to go for 2.

I agree with you, but having a better chance of winning is having a better chance of winning, is it not?Again, I'm flying in the face of convention, I realize this. But if the numbers are screaming that you should consider it, I believe that you should consider it.BTW, the numbers are just the NFL average. If you are a proficient team at 2-PT plays (as one would really have to assume if you practice this more than most teams) then the odds should also go up.

 

[trader jake]

I'd like to subscribve to your newsletter Jeff

[hijack] Also, if you get some time Jeff...what does the math say about going for 2 when a team is up by 1 and scores a touchdown? I've always believed that in the 4th quarter of a game the offensive team should go for 2 in this situation. If they play it safe and kick the extra point they are up by 8 (one possession). If the try doesn't work they are up by 7, but if they 2 point try works they are up by 9 (2 possessions).[/hijack]

 

[Jeff Pasquino]

1. The assumption here is that the goal is to getting a tie. That's a fatal flaw in the argument. You need 15 to win the game, so that's the real goal.

The fatal flaw continues to be that there is absolutely no guarantee you need 15 to win the game. You need one more than your opponent, which may or may not be 15.

But you do concede that you need AT LEAST 15 to win, correct?I simplified the argument for ease of the math, which may or may not be the appropriate thing to do, but it does show the case for going for 2.I'll say it a different way - going for 2 after the first TD gives you a better chance of having 14 or 15 points after two touchdowns than if you don't go for two after the first score. If you kick then go for 2, you have 15 points 43% of the time and 13 points 57% of the time. Going for 2 and then kicking or going for 2 again gives you 15 points ~42.5% of the time, 14 points ~24.5% of the time, and 12 points ~33% of the time. So you're at 14 or 15 points more often if you go for 2 the first time.

 

[Jeff Pasquino]

I'd like to subscribve to your newsletter Jeff

[hijack] Also, if you get some time Jeff...what does the math say about going for 2 when a team is up by 1 and scores a touchdown? I've always believed that in the 4th quarter of a game the offensive team should go for 2 in this situation. If they play it safe and kick the extra point they are up by 8 (one possession). If the try doesn't work they are up by 7, but if they 2 point try works they are up by 9 (2 possessions).[/hijack]

Good question.According to the link provided earlier in this thread ( Chart ) it says go for 2 in the 4th quarter if you're up by 1.