You're down by 15 with 7:00 minutes left in the game (1 Viewer)

[Bottomfeeder Sports]

Oh, for actual math.

Assumptions:

45% chance of making 2 point conversion at 7:00 minute mark

Down

9 at seven minute mark, opponent has ball at 20 6% chance of winning

8 at seven minute mark, opponent has ball at 20 - 8% chance of winning

7 at seven minute mark, opponent has ball at 20 - 12% chance of winning

(Since field position is mostly linear I don't believe that changing it will change these calculations.)

.45 * .12 + .55 * .06 = 0.087

0.087 > .08

Now if the team down by 9 tries an onside kick at this point they will likely kill their chances.

Down 9 at seven minute mark, opponent has ball at your 45 - 3% chance of winning

Down 9 at seven minute mark, you have the ball at your 45 - 8% chance of winning

That the success rate will be about 20%

.80 * .03 + .20 * .08 = 0.04

Okay, and if I change the assumptions, you could come to the opposite conclusion. This really isn't that difficult- there is no mathematically correct answer to the question.

Of course. There real probability of making a two point conversion is more like 47%, and if you can run the ball it is 57%. Sure if you give up a long return such that leading team is in field goal range or better the probabilities become 3,4,5 and the available precision is such that the team up 8's 4% is better than the 3.9%. But what percentage of a time is there a 60+ yard return when the probabilities shift in your favor versus the number of kick offs that are for less than 60 yards? So yes there are places on the field where going for one was the right call, but those are unlikely places to be and the differences are smaller (because the chances of winning are decreasing across the board pretty much linearly).

So if you can find probabilities for various kickoff field positions the match could be made more complete, but the end result will not change. The team going for two sooner will have a slightly better chance of an improbable win. NFL coaches seldom go for "slightly better chances" though.

 

[humpback]

I just ran the numbers for pre-2 pt. conversion, and guess what? They show nearly the same exact results as post 2 pt. conversion- teams down by 9 pts. won much more often than teams down by 8, and teams down 17 won more often than teams down 16 or 15. So, blaming these counter-intuitive results on poor strategy or coaching decisions because of the option to go for 2 is completely bunk- the same things happened before going for 2 was even an option.

 

[Ignoratio Elenchi]

Okay, and if I change the assumptions, you could come to the opposite conclusion.

Now we're getting somewhere. It'd be very helpful if you could list whatever assumptions you think would lead to the opposite conclusion.

 

[Maurile Tremblay]

Okay, and if I change the assumptions, you could come to the opposite conclusion.

Now we're getting somewhere. It'd be very helpful if you could list whatever assumptions you think would lead to the opposite conclusion.

There are no assumptions that could lead to the opposite conclusion. That is the point.

humpback says there are some, and I agree with him. (Assume, for example, that the chance of making a two-point conversion when down by 9 is 15%, while the chance of making a two-point conversion when down by 2 is 85%.)The question is whether any such set of assumptions is reasonable. I don't know exactly which assumptions humpback has in mind (or will be able to come up with), but we can't judge their reasonableness very well until we know what they are. So I think it'd be helpful if he were to state them.

 

[humpback]

Okay, and if I change the assumptions, you could come to the opposite conclusion.

Now we're getting somewhere. It'd be very helpful if you could list whatever assumptions you think would lead to the opposite conclusion.

There are no assumptions that could lead to the opposite conclusion. That is the point.

humpback says there are some, and I agree with him. (Assume, for example, that the chance of making a two-point conversion when down by 9 is 15%, while the chance of making a two-point conversion when down by 2 is 85%.)The question is whether any such set of assumptions is reasonable. I don't know exactly which assumptions humpback has in mind (or will be able to come up with), but we can't judge their reasonableness very well until we know what they are. So I think it'd be helpful if he were to state them.

No, according to professor jon, there is no possible way that this could happen. It's mathematically impossible. The reason why I haven't stated any is because it would just start another 5 pages of back and forth arguing about something that can't be proven. One side would consider different assumptions to be reasonable, the other side would not. All you have to do is tweak some numbers in BFS's example to come to the opposite conclusion (say, assume the odds of winning when down 7 are 10% instead of 12%, or assuming the odds of winning when down 8 are 9% instead of 8%).Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points. People who believe these things will never see eye to eye with people who do not, and vice versa.

 

[jon_mx]

Okay, and if I change the assumptions, you could come to the opposite conclusion.

Now we're getting somewhere. It'd be very helpful if you could list whatever assumptions you think would lead to the opposite conclusion.

There are no assumptions that could lead to the opposite conclusion. That is the point.

humpback says there are some, and I agree with him. (Assume, for example, that the chance of making a two-point conversion when down by 9 is 15%, while the chance of making a two-point conversion when down by 2 is 85%.)The question is whether any such set of assumptions is reasonable. I don't know exactly which assumptions humpback has in mind (or will be able to come up with), but we can't judge their reasonableness very well until we know what they are. So I think it'd be helpful if he were to state them.

No, according to professor jon, there is no possible way that this could happen. It's mathematically impossible. The reason why I haven't stated any is because it would just start another 5 pages of back and forth arguing about something that can't be proven. One side would consider different assumptions to be reasonable, the other side would not. All you have to do is tweak some numbers in BFS's example to come to the opposite conclusion (say, assume the odds of winning when down 7 are 10% instead of 12%, or assuming the odds of winning when down 8 are 9% instead of 8%).Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points. People who believe these things will never see eye to eye with people who do not, and vice versa.

Well I am assuming things like the law of gravity applies equally throughout the game. Assumptions need to be consistent.

 

[Maurile Tremblay]

Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points.

I don't think any of your statements about what other people believe are accurate.

 

[humpback]

Okay, and if I change the assumptions, you could come to the opposite conclusion.

Now we're getting somewhere. It'd be very helpful if you could list whatever assumptions you think would lead to the opposite conclusion.

There are no assumptions that could lead to the opposite conclusion. That is the point.

humpback says there are some, and I agree with him. (Assume, for example, that the chance of making a two-point conversion when down by 9 is 15%, while the chance of making a two-point conversion when down by 2 is 85%.)The question is whether any such set of assumptions is reasonable. I don't know exactly which assumptions humpback has in mind (or will be able to come up with), but we can't judge their reasonableness very well until we know what they are. So I think it'd be helpful if he were to state them.

No, according to professor jon, there is no possible way that this could happen. It's mathematically impossible. The reason why I haven't stated any is because it would just start another 5 pages of back and forth arguing about something that can't be proven. One side would consider different assumptions to be reasonable, the other side would not. All you have to do is tweak some numbers in BFS's example to come to the opposite conclusion (say, assume the odds of winning when down 7 are 10% instead of 12%, or assuming the odds of winning when down 8 are 9% instead of 8%).Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points. People who believe these things will never see eye to eye with people who do not, and vice versa.

Well I am assuming things like the law of gravity applies equally throughout the game. Assumptions need to be consistent.

Yes, to prove something mathematically, assumptions need to be consistent. In an NFL football game, they do not. Let's just agree to disagree.

 

[Bottomfeeder Sports]

One side would consider different assumptions to be reasonable, the other side would not. All you have to do is tweak some numbers in BFS's example to come to the opposite conclusion (say, assume the odds of winning when down 7 are 10% instead of 12%, or assuming the odds of winning when down 8 are 9% instead of 8%).

Those aren't assumptions.

 

[humpback]

Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points.

I don't think any of your statements about what other people believe are accurate.

Several posts in here disagree with you.ETA- only have to look at the one right above this one. And the one above yours. And a couple dozen others earlier.

 

[GregR]

Just because a single individual situation may have a different likelihood doesn't mean that it is invalid to determine the overall likelihood for all situations... nor does it mean that it isn't useful to know the overall likelihood.

If strategy A is successful 55% of the time and B is successful 45%, that tells me I should have some compelling reasons before I go with B. If B is the one that is 55% successful, that tells me I should have some compelling reasons before I go with A.

 

[Bottomfeeder Sports]

Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points.

I don't think any of your statements about what other people believe are accurate.

Several posts in here disagree with you.ETA- only have to look at the one right above this one. And the one above yours. And a couple dozen others earlier.

Average Win probabilities at specific points on the field at specific down and distances at specific times with specific point differentials are not assumptions, but calculation. The one I linked to is based on eight years of data. Is it perfect model - probably not, but it is not just some numbers pulled out of thin air. And the data that feeds this provides consistent results in "Expected Points" to the data used twenty five years ago by Pete Palmer and company so I think the assumption that this 8 year period is more or less reflective of the game is a safe one.

 

[Bottomfeeder Sports]

Just because a single individual situation may have a different likelihood doesn't mean that it is invalid to determine the overall likelihood for all situations... nor does it mean that it isn't useful to know the overall likelihood.If strategy A is successful 55% of the time and B is successful 45%, that tells me I should have some compelling reasons before I go with B. If B is the one that is 55% successful, that tells me I should have some compelling reasons before I go with A.

Well to be fair we are talking about strategies that are almost never successful to begin with. There is no strategy at this point in the game that give you either a 45% or 55% chance or even 25%. Odds are overwhelming you are going to lose, despite the fact that as fans we can recall many similar come from behind victories over the year. The questions is in this situation should a NFL head coach pick the option that gives his team a slightly better chance to win or take the option that will likely not be second guessed on Monday? And also, replace "should" with "would" to determine how the "coaches book" is written, what is drilled in our heads by analysts that confuse the two.

 

[Bottomfeeder Sports]

People think a team will play no differently when up 7, 8, or 9 points.

Most teams up 9 will be much more conservative than teams up by 8 or 7. Most teams up by 8 will be slightly more conservative than teams up by 7. Yes there are exceptions.

 

[Bottomfeeder Sports]

..All you have to do is tweak some numbers in BFS's example to come to the opposite conclusion (say, assume the odds of winning when down 7 are 10% instead of 12%, or assuming the odds of winning when down 8 are 9% instead of 8%)...

The numbers were12%8%6%If X is the probability of winning down 7 at a given point in the game and Y is the probability of winning down by 9, does is make intuitive sense that being down 8 is more like being down by 9 then being down by 7 if making a two point conversion is less than a 50-50 proposition? Thus Z the probability of winning down 8 is always closer to X than Y when two point conversions are less than 50-50? Is there any logic that would make being down by 8 much more like being down by 7 like your new set of assumption require? This is why jon says the actual numbers don't matter,

 

[humpback]

Basically, this comes back to flawed logic IMO- people think they can precisely determine the odds of the next event occuring based only on past data. People think the likelihood of an event happening is the same for all teams under all sets of circumstances. People think a team will play no differently when up 7, 8, or 9 points.

I don't think any of your statements about what other people believe are accurate.

Several posts in here disagree with you.ETA- only have to look at the one right above this one. And the one above yours. And a couple dozen others earlier.

Average Win probabilities at specific points on the field at specific down and distances at specific times with specific point differentials are not assumptions, but calculation. The one I linked to is based on eight years of data. Is it perfect model - probably not, but it is not just some numbers pulled out of thin air. And the data that feeds this provides consistent results in "Expected Points" to the data used twenty five years ago by Pete Palmer and company so I think the assumption that this 8 year period is more or less reflective of the game is a safe one.

Yes, a calculation based on previous events, all of which are independent from this current unique set of circumstances. It would be like saying the stock market will go up X% next month because that's been the historical average gain per month. The reality is we don't know what the chances are of the next 2 pt. conversion being successful are. You can estimate them based on previous attempts, but my estimate is going to account for a lot of other variables as well. Either way, it's still only going to be an unprovable estimate, not statistical fact.

 

[humpback]

People think a team will play no differently when up 7, 8, or 9 points.

Most teams up 9 will be much more conservative than teams up by 8 or 7. Most teams up by 8 will be slightly more conservative than teams up by 7. Yes there are exceptions.

I agree with this logic in general, although it's hard to say to what degree teams will adjust their strategy, and even harder to say if it would have any impact on the outcome. However, others have said that the leading team wouldn't change their strategy.

 

[Bottomfeeder Sports]

Yes, a calculation based on previous events, all of which are independent from this current unique set of circumstances. It would be like saying the stock market will go up X% next month because that's been the historical average gain per month. The reality is we don't know what the chances are of the next 2 pt. conversion being successful are. You can estimate them based on previous attempts, but my estimate is going to account for a lot of other variables as well. Either way, it's still only going to be an unprovable estimate, not statistical fact.

I've acknowledged through out this thread that the possibility exists that the chances of converting at 7 minutes are uniquely poorer for some reason or another than the chances would normally be (which is the best assumption for later in the game you can have at this moment). But these would be the exception and not the rule.I mean if your quarterback sneaked down the steps to locker room that home teams have at modern stadiums to thinking he has at least a change of possession TV timeout when instead the defense scores a TD, and you don't have a fake conversion in your playbook you should probably go for one, but other than situations like this one.

 

[firstseason1988]

I haven't posted this in a few pages but figured I would enlighten folks again.

Madden I got for 2.

Real life I go for 1.

Emotion is the ultimate factor.

/thread

 

[shnikies]

I think we can all agree that's it's better to be down by 7 than 8.

Historical data indicates that (counter-intuitively) it's better to be down by 9 than down by 8.

Go for two

Whatever historical data you have found may be based on human error. It's illogical if a coach uses dominant strategies for both cases, where he down by 9 as well as down by 8, would result in more wins for the teams down by 9. If we're using game theory and/or probabilities for deciding whether going for 1 or 2 is correct then you can't use historical data that might include situations where the coaches didn't. Whether historical data shows it's better to be down by 9 that by 8 or not has no effect on the question. I'm not saying going for one is better, I'm just saying what you said isn't a valid reason.

When you are down by eight you playing for a 50% chance to tie the game. And you better not have left any time on the clock or the other team will have a chance to kick a FG. So even after all this, you then only have a 50% chance to win in overtime. When you are down by 9, you are playing to win the game. You don't have to worry about making the two point conversion or losing in overtime. You may have to make some riskier plays to get there, but if you are successful, you win.

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

 

[Ignoratio Elenchi]

I think we can all agree that's it's better to be down by 7 than 8.

Historical data indicates that (counter-intuitively) it's better to be down by 9 than down by 8.

Go for two

Whatever historical data you have found may be based on human error. It's illogical if a coach uses dominant strategies for both cases, where he down by 9 as well as down by 8, would result in more wins for the teams down by 9. If we're using game theory and/or probabilities for deciding whether going for 1 or 2 is correct then you can't use historical data that might include situations where the coaches didn't. Whether historical data shows it's better to be down by 9 that by 8 or not has no effect on the question. I'm not saying going for one is better, I'm just saying what you said isn't a valid reason.

When you are down by eight you playing for a 50% chance to tie the game. And you better not have left any time on the clock or the other team will have a chance to kick a FG. So even after all this, you then only have a 50% chance to win in overtime. When you are down by 9, you are playing to win the game. You don't have to worry about making the two point conversion or losing in overtime. You may have to make some riskier plays to get there, but if you are successful, you win.

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

Half the time your "score as quickly as possible" strategy is the wrong one. That's why being down by 8 is problematic.

 

[shnikies]

I think we can all agree that's it's better to be down by 7 than 8.

Historical data indicates that (counter-intuitively) it's better to be down by 9 than down by 8.

Go for two

Whatever historical data you have found may be based on human error. It's illogical if a coach uses dominant strategies for both cases, where he down by 9 as well as down by 8, would result in more wins for the teams down by 9. If we're using game theory and/or probabilities for deciding whether going for 1 or 2 is correct then you can't use historical data that might include situations where the coaches didn't. Whether historical data shows it's better to be down by 9 that by 8 or not has no effect on the question. I'm not saying going for one is better, I'm just saying what you said isn't a valid reason.

When you are down by eight you playing for a 50% chance to tie the game. And you better not have left any time on the clock or the other team will have a chance to kick a FG. So even after all this, you then only have a 50% chance to win in overtime. When you are down by 9, you are playing to win the game. You don't have to worry about making the two point conversion or losing in overtime. You may have to make some riskier plays to get there, but if you are successful, you win.

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

Half the time your "score as quickly as possible" strategy is the wrong one. That's why being down by 8 is problematic.

If there is enough time to score twice when you're down by 9 then why isn't there enough time to score twice when you're down 8? The only difference between being down by 8 or being down by 9 is that when you're down 8 you have a chance to tie the game in one possession.

 

[Bottomfeeder Sports]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

 

[GregR]

...Yes, a calculation based on previous events, all of which are independent from this current unique set of circumstances. It would be like saying the stock market will go up X% next month because that's been the historical average gain per month. The reality is we don't know what the chances are of the next 2 pt. conversion being successful are. You can estimate them based on previous attempts, but my estimate is going to account for a lot of other variables as well. Either way, it's still only going to be an unprovable estimate, not statistical fact.

The math cannot give you a better answer than the set of inputs (beliefs on the likelihood of each factor) you give it. This is no different than a coach who isn't using any of the math consciously. He still has to make a decision. He still has to base it on his beliefs about the likelihood of each factor being successful. Whether he bases that on historical trends or on his gut doesn't matter. Both math and the coach have to work with what they have to work with.But where the math and the coach differ is that if the math is given the coach's beliefs as input, it will show which is the best answer for that set of beliefs. The coach may not be correct when he combines his beliefs to come to a decision.To use your stock market example, math can't necessarily tell you whether the stock market will continue to go up after a historical trend, anymore than your brain can. But what math can do is tell you that when you believe the stock market will go up by that type of trend, which course of investing is optimal based your beliefs.Which brings us back to MT's request. For a set of assumptions on the odds of a 2 pt succeeding, etc, the math can tell us which course of action is optimum for that set of assumptions. So what set of assumptions do you believe are the most reasonable for the various factors involved for an average pair of NFL teams? Other people are saying a reasonable set of beliefs won't give going for 1 as the optimal strategy. If you disagree, give what you think the odds are for the various factors involved for an average pair of NFL teams.

 

[shnikies]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

 

[humpback]

...Yes, a calculation based on previous events, all of which are independent from this current unique set of circumstances. It would be like saying the stock market will go up X% next month because that's been the historical average gain per month. The reality is we don't know what the chances are of the next 2 pt. conversion being successful are. You can estimate them based on previous attempts, but my estimate is going to account for a lot of other variables as well. Either way, it's still only going to be an unprovable estimate, not statistical fact.

The math cannot give you a better answer than the set of inputs (beliefs on the likelihood of each factor) you give it. This is no different than a coach who isn't using any of the math consciously. He still has to make a decision. He still has to base it on his beliefs about the likelihood of each factor being successful. Whether he bases that on historical trends or on his gut doesn't matter. Both math and the coach have to work with what they have to work with.But where the math and the coach differ is that if the math is given the coach's beliefs as input, it will show which is the best answer for that set of beliefs. The coach may not be correct when he combines his beliefs to come to a decision.To use your stock market example, math can't necessarily tell you whether the stock market will continue to go up after a historical trend, anymore than your brain can. But what math can do is tell you that when you believe the stock market will go up by that type of trend, which course of investing is optimal based your beliefs.Which brings us back to MT's request. For a set of assumptions on the odds of a 2 pt succeeding, etc, the math can tell us which course of action is optimum for that set of assumptions. So what set of assumptions do you believe are the most reasonable for the various factors involved for an average pair of NFL teams? Other people are saying a reasonable set of beliefs won't give going for 1 as the optimal strategy. If you disagree, give what you think the odds are for the various factors involved for an average pair of NFL teams.

Again, this is a giant waste of time, but I gave 2 simple tweaks to the numbers in one of the posts above that would say going for 1 would be the optimal strategy.

 

[Bottomfeeder Sports]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

 

[shnikies]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

So it's better to be down by 2 than tied? I did not know that.

 

[Ignoratio Elenchi]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

So it's better to be down by 2 than tied? I did not know that.

wat

Step away shnikes, this isn't going well for you.

 

[Bottomfeeder Sports]

Again, this is a giant waste of time, but I gave 2 simple tweaks to the numbers in one of the posts above that would say going for 1 would be the optimal strategy.

You replaced numbers calculated from historical data with numbers pulled out of you and you call them "simple tweaks"? I think Greg and Maurile and everyone else wants to know what game situations allows your team that is down by 7 have a much less than historical average probability of winning while at the same time your team down by 8 have a much more than historical average probability of winning? What reasonable factors would create these situations?

 

[shnikies]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

So it's better to be down by 2 than tied? I did not know that.

wat

Step away shnikes, this isn't going well for you.

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

 

[Ignoratio Elenchi]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

So it's better to be down by 2 than tied? I did not know that.

wat

Step away shnikes, this isn't going well for you.

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

No, of course not, because it's not correct.

 

[shnikies]

That's the incorrect strategy that I'm sure skewed the results. When you're down by 8, the correct strategy is to score as quickly as possible. You should assume you won't get the 2 point conversion and thus need another possession. So, tell me how being down by 9 is better than being down by 8?

If you are going to fail to make the two point conversion either way how is being down 9 worst than 8?And if you do luck out and get the tie with enough time on the clock to score again you are still more likely to lose! (The team with the ball at the end of a tied game has the better than 50-50 chance of winning.)

Edited to add the "still".

So, how is being down 9 better than being down by 8?

In some cases being down 8 is for all practical purposes the same as being down 9. In some cases being down 8 is for all practical purposes the same as being down 7. The proper strategy down by 7 is different from the proper strategy down by 9. Being down 9 or 7 you know which is the proper strategy. Being down 8 you have no clue. Ignorance might be bliss, but it is not a winning strategy.

So it's better to be down by 2 than tied? I did not know that.

wat

Step away shnikes, this isn't going well for you.

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

No, of course not, because it's not correct.

 

[Ignoratio Elenchi]

I know you're having a hard time with this, which is why I suggested you let it go.

 

[humpback]

Again, this is a giant waste of time, but I gave 2 simple tweaks to the numbers in one of the posts above that would say going for 1 would be the optimal strategy.

You replaced numbers calculated from historical data with numbers pulled out of you and you call them "simple tweaks"? I think Greg and Maurile and everyone else wants to know what game situations allows your team that is down by 7 have a much less than historical average probability of winning while at the same time your team down by 8 have a much more than historical average probability of winning? What reasonable factors would create these situations?

Well, since numbers calculated from historical data have zero impact on the teams chances at winning today, yes, they are simple tweaks.What reasonable factors would lead to a team historically being more likely to win when behind by 9 than when behind by 7? I don't know, but it isn't the 2 point conversion.

 

[shnikies]

I know you're having a hard time with this, which is why I suggested you let it go.

You have made so many good points in this argument. Very well done.

 

[Bottomfeeder Sports]

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

You ignore the benefits of being down 7 vs 8 are greater than the benefits of being down 8 vs 9.

 

[shnikies]

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

You ignore the benefits of being down 7 vs 8 are greater than the benefits of being down 8 vs 9.

Like I said earlier, I'm not trying to weigh in on the original question of whether going for 2 is the right play. I'm simply debunking the idea that being down 9 is better than being down by 8. That's simply not true.

 

[kegger]

And still,

The overwhelmingly correct answer is go for 1 till you have to go for 2.

-------

The "math" guy's want everyone to believe there math, but leave all other

factors out of the decision, like what the other team can do now, or how it

will affect the thought process of either team, emotion plays a much larger

part to big time sports then many here want to believe, let alone the "logic"

that disagrees with them.

Ever hear of momentum swing, choking under pressure, playing conservative, ect..

(None of that is even being factored in)

 

[Ignoratio Elenchi]

Oh god

 

[shnikies]

Oh god

Do you ever wonder if your posts serve any purpose?

 

[Ignoratio Elenchi]

Oh god

Do you ever wonder if your posts serve any purpose?

No.

 

[Bottomfeeder Sports]

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

You ignore the benefits of being down 7 vs 8 are greater than the benefits of being down 8 vs 9.

Like I said earlier, I'm not trying to weigh in on the original question of whether going for 2 is the right play. I'm simply debunking the idea that being down 9 is better than being down by 8. That's simply not true.

Fair enough. But you challenged a poster consistently arguing for going for two early based on having knowledge of how to play the rest of the game. Guys on both sides of the debate have stated that teams down 9 win more often than teams down 8, both with and without a two point option. If you believe them and there is any quantity of data to make this more than just some random fluctuation then what is your explanation? Mine is coaches being more conservative up by 9 than 8, and more aggressive down by 9 than 8 either way.

 

[shnikies]

The only difference between being down 8 and being down by 9 is that after you score the touchdown you can tie the game. If you don't get the 2 point conversion you're down by 2, in the same situation as if you were down 9 initially. Saying being down by 9 is better than being down by 8 is the same as saying you'd rather be down by 2 points than tied. Does that help you understand?

You ignore the benefits of being down 7 vs 8 are greater than the benefits of being down 8 vs 9.

Like I said earlier, I'm not trying to weigh in on the original question of whether going for 2 is the right play. I'm simply debunking the idea that being down 9 is better than being down by 8. That's simply not true.

Fair enough. But you challenged a poster consistently arguing for going for two early based on having knowledge of how to play the rest of the game. Guys on both sides of the debate have stated that teams down 9 win more often than teams down 8, both with and without a two point option. If you believe them and there is any quantity of data to make this more than just some random fluctuation then what is your explanation? Mine is coaches being more conservative up by 9 than 8, and more aggressive down by 9 than 8 either way.

My argument is that the data is based on the fact that coaches incorrectly play the game like they only need one possession if they're down by 8. I'm guessing more often than coaches go for the extra point first when they're down by 15, does that mean it's the correct decision? As for the the thread's original question, I answered it doesn't matter. There are pros and cons for both and there are too many unknown variables.

 

[Bottomfeeder Sports]

My argument is that the data is based on the fact that coaches incorrectly play the game ...

They coach to coach again before they coach win. Whether that is incorrect depends on your perspective.

... I'm guessing more often than coaches go for the extra point first when they're down by 15, ...

In the NFL they don't "more often" go for one, they have never gone for two (unless I missed it recently).

...There are pros and cons for both and there are too many unknown variables.

The point that I have mostly made in this thread is that it seem difficult to believe that all of those unknown variables in all of these games dictated going for one. All games that were lost except this year's Denver game versus Miami which the decision was at 3 minutes and not 7. Coaches don't think twice going for one here and they lose every time. In the NFL going for one is not the default position but so far the only option. That has to be wrong.So if the correct is what is the default choice absent the answers for the unknown variables then it is to go for two at 7 minutes. That has been established in various ways in this thread, including but not limited to statistical research. Of course some of those unknowns could have values that change the answer. What several have asked are what these realistic game situations that would make going for one the better choice?

Humpback offered that maybe the odds of winning in some particular game down 8 is roughly the equivalent to being down 7 and both are much, much better than being down 9. But how do you get there without having the two point conversion being a much better than 50-50 chance? Which, of course defeats the argument of not taking the chance to begin with.

So what are the real world examples of why teams chose not to go for two? I've offered numerous, but they all reflect poorly on the team or coach in question. What else do we have? Emotionally fragile players?

 

[shnikies]

My argument is that the data is based on the fact that coaches incorrectly play the game ...

They coach to coach again before they coach win. Whether that is incorrect depends on your perspective.

... I'm guessing more often than coaches go for the extra point first when they're down by 15, ...

In the NFL they don't "more often" go for one, they have never gone for two (unless I missed it recently).

...There are pros and cons for both and there are too many unknown variables.

The point that I have mostly made in this thread is that it seem difficult to believe that all of those unknown variables in all of these games dictated going for one. All games that were lost except this year's Denver game versus Miami which the decision was at 3 minutes and not 7. Coaches don't think twice going for one here and they lose every time. In the NFL going for one is not the default position but so far the only option. That has to be wrong.So if the correct is what is the default choice absent the answers for the unknown variables then it is to go for two at 7 minutes. That has been established in various ways in this thread, including but not limited to statistical research. Of course some of those unknowns could have values that change the answer. What several have asked are what these realistic game situations that would make going for one the better choice?

Humpback offered that maybe the odds of winning in some particular game down 8 is roughly the equivalent to being down 7 and both are much, much better than being down 9. But how do you get there without having the two point conversion being a much better than 50-50 chance? Which, of course defeats the argument of not taking the chance to begin with.

So what are the real world examples of why teams chose not to go for two? I've offered numerous, but they all reflect poorly on the team or coach in question. What else do we have? Emotionally fragile players?

So you're saying in less than 7 minutes your odds of winning are better if you get the ball back, score a touchdown or kick a field goal then get the ball back again and score a TD if you got the field goal first or vice versa than it is to get the ball back, score a touchdown, convert a 2 point conversion and win in overtime? Plus there's a chance of getting the ball back and scoring to win in regulation. I'm going to agree to disagree.

 

[Goldalene]

Easy decision, you go for two.

 

[Bottomfeeder Sports]

So you're saying in less than 7 minutes your odds of winning are better if you get the ball back, score a touchdown or kick a field goal then get the ball back again and score a TD if you got the field goal first or vice versa than it is to get the ball back, score a touchdown, convert a 2 point conversion and win in overtime? Plus there's a chance of getting the ball back and scoring to win in regulation. I'm going to agree to disagree.

Why do we assume that the two point conversion at 7 minutes is an automatic miss and the two point conversion at the end of the game is automatic? I'm saying that the difference between being down 7 vs 8 is much bigger than the difference being down 8 vs 9. This is both intuitive and reflected in statistical research. Unless one 2 point conversion is more likely than another, this difference is what matters. Your earlier posts that strategically you have to treat down by 8 as two possession game suggest that even you think being down 8 is more like being down 9 than it is 7. But in order to argue you a point you just claimed wasn't your point at all you tossed that aside?

 

[shnikies]

So you're saying in less than 7 minutes your odds of winning are better if you get the ball back, score a touchdown or kick a field goal then get the ball back again and score a TD if you got the field goal first or vice versa than it is to get the ball back, score a touchdown, convert a 2 point conversion and win in overtime? Plus there's a chance of getting the ball back and scoring to win in regulation. I'm going to agree to disagree.

Why do we assume that the two point conversion at 7 minutes is an automatic miss and the two point conversion at the end of the game is automatic? I'm saying that the difference between being down 7 vs 8 is much bigger than the difference being down 8 vs 9. This is both intuitive and reflected in statistical research. Unless one 2 point conversion is more likely than another, this difference is what matters. Your earlier posts that strategically you have to treat down by 8 as two possession game suggest that even you think being down 8 is more like being down 9 than it is 7. But in order to argue you a point you just claimed wasn't your point at all you tossed that aside?

Whatever you said, I agree to disagree.

 

[Ignoratio Elenchi]