Rams/Cardinals and Pete Prisco

[NKR1978]

Read this and immediately thought of this board.

Football has to be the dumbest sport when it comes to accepting new wY of thinking. Baseball is painfully dumb with the absolutely ridiculous BBWA, but at least teams generally accept statistics (except for untouchable players like Jeter his last season).

With football there just seems like there's this idiotic mentality of "you gotta take the points" even if you're at the 1 and down 10 in the 4th.

This is just a funny exchange:

deadspin.com/pete-prisco-very-adamantly-does-not-understand-probabil-1670188622

 

[Full_Rebar]

No shock that Peter King was on Prisco's side.

 

[NKR1978]

No shock that Peter King was on Prisco's side.

I know. I knew Peter King would support an idiotic decision before I even saw his tweet.

 

[Millhouse]

Read this and immediately thought of this board.

Football has to be the dumbest sport when it comes to accepting new wY of thinking. Baseball is painfully dumb with the absolutely ridiculous BBWA, but at least teams generally accept statistics (except for untouchable players like Jeter his last season).

With football there just seems like there's this idiotic mentality of "you gotta take the points" even if you're at the 1 and down 10 in the 4th.

This is just a funny exchange:

deadspin.com/pete-prisco-very-adamantly-does-not-understand-probabil-1670188622

just another example of meatheads stupidly thinking that the objective is to maximize the amount of time where you have any chance at all. the smart objective is to maximize the chance of winning. if it means some games are decided a few minutes earlier, so be it

herm edwards you play to win the game

you don't play to extend the game

 

[javadoc]

just another example of meatheads stupidly thinking that the objective is to maximize the amount of time where you have any chance at all. the smart objective is to maximize the chance of winning. if it means some games are decided a few minutes earlier, so be it

I haven't followed the statistical models and calculations that underlie these systems, so how accurate and reliable are they? What are the margins of uncertainty? If the model tells you one choice gives you 15% and another gives you 19%, is that significant enough to act on?

I'm more interested in cases where the system tells you that your gut choice is really 25% worse than alternatives, sort of an "aha!" moment. Have there been any such, or are we generally talking small differences?

 

[Millhouse]

you don't even need statistical models here, this one is common sense.

if you convert a 4th and 1, the drive you would need to win the game is 40 yards shorter.

think of it this way. say you're at the 40 yard line with 1 second left and you're down 2. ref gives you two choices. you can throw the hail mary. or you can flip a coin - heads you can win with a FG attempt, tails the game is over.

everyone would say flip the coin. 50% coin flip times 50% chance of hitting the FG is 25% chance of winning. There is no way you have a 1 in 4 chance of a hail mary.

the 4th and 1 call is kind of like that coin flip. you're taking a 50/50 chance in order to save 40 yards later. the only difference is the 4th and 1 coin flip is a couple minutes earlier than the hypothetical one

no one is using a statistical model here to figure out how to extend the game - my point is that their goal is all wrong. extend vs win.

if a coach has a model that tells him that a FG actually does increase their win probability, i'm all ears. but they're not even asking that question. it's just about delaying definitively losing, which isn't a good objective

 

[NKR1978]

I haven't followed the statistical models and calculations that underlie these systems, so how accurate and reliable are they? What are the margins of uncertainty? If the model tells you one choice gives you 15% and another gives you 19%, is that significant enough to act on?

I'm more interested in cases where the system tells you that your gut choice is really 25% worse than alternatives, sort of an "aha!" moment. Have there been any such, or are we generally talking small differences?

In this situation, even his "feelings" should have told him to go for it. There were 6 minutes to go in the game, they hadn't done anything on offense in the 2nd half. This was their only decent drive of the 2nd half (believe it was the only drive that moved past the 50), so why would he think he could put another 80 yard drive together? If they converted they would just have to move 40 yards instead of 80 to kick a makable FG.

 

[Millhouse]

In this situation, even his "feelings" should have told him to go for it. There were 6 minutes to go in the game, they hadn't done anything on offense in the 2nd half. This was their only decent drive of the 2nd half (believe it was the only drive that moved past the 50), so why would he think he could put another 80 yard drive together? If they converted they would just have to move 40 yards instead of 80 to kick a makable FG.

i've posted it 10,000 times but i wish everyone would read thinking fast and slow. big theme in it is how when faced with a difficult question (what maximized our win probability), people will substitute an easier one (what maximizes the amount of time we have any shot at all)

 

[javadoc]

you don't even need statistical models here, this one is common sense.

if you convert a 4th and 1, the drive you would need to win the game is 40 yards shorter.

think of it this way. say you're at the 40 yard line with 1 second left and you're down 2. ref gives you two choices. you can throw the hail mary. or you can flip a coin - heads you can win with a FG attempt, tails the game is over.

everyone would say flip the coin. 50% coin flip times 50% chance of hitting the FG is 25% chance of winning. There is no way you have a 1 in 4 chance of a hail mary.

the 4th and 1 call is kind of like that coin flip. you're taking a 50/50 chance in order to save 40 yards later. the only difference is the 4th and 1 coin flip is a couple minutes earlier than the hypothetical one

no one is using a statistical model here to figure out how to extend the game - my point is that their goal is all wrong. extend vs win.

if a coach has a model that tells him that a FG actually does increase their win probability, i'm all ears. but they're not even asking that question. it's just about delaying definitively losing, which isn't a good objective

Understand, was hijacking slightly to get your take on the statistics for tougher choices, since we see reported numbers from time to time.

 

[Millhouse]

Understand, was hijacking slightly to get your take on the statistics for tougher choices, since we see reported numbers from time to time.

i don't know if the uncertainty matters much because there are only two outcomes

if you're investing, you have a whole continuum of outcomes and risk. if the goal is to minimize the losing margin, kicking sure thing fgs might be the right strategy (although not here)

there is some model risk either way - having no model and going with your gut has it's own uncertainty

 

[Finwad32]

i don't know if the uncertainty matters much because there are only two outcomes

if you're investing, you have a whole continuum of outcomes and risk. if the goal is to minimize the losing margin, kicking sure thing fgs might be the right strategy (although not here)

there is some model risk either way - having no model and going with your gut has it's own uncertainty

I like the thought process. My question is this: At what point in the game, time wise, is it necessary to start taking these risks?
Would you go for 4th & 1 from the 40 with 10 min. left?
Would you do it half way through the 3rd?
Is the same timeframe applicable to the first half; even in a tie game in the 2nd quarter would you risk the field position with two minutes before half for a go ahead score?

I'm not trying to poke holes, I like what I hear. Just wondering how aggressive is too aggressive if you adhere to this philosophy.

 

[Millhouse]

I like the thought process. My question is this: At what point in the game, time wise, is it necessary to start taking these risks?
Would you go for 4th & 1 from the 40 with 10 min. left?
Would you do it half way through the 3rd?
Is the same timeframe applicable to the first half; even in a tie game in the 2nd quarter would you risk the field position with two minutes before half for a go ahead score?

I'm not trying to poke holes, I like what I hear. Just wondering how aggressive is too aggressive if you adhere to this philosophy.

the more time you have left, the less likely it is that a failure to convert will clinch defeat

i'd always go for it on 4th and 1 at the one unless you're ahead and only need 3 to end the game

maybe i'd be willing to go for a 4th and 6 in that spot at the end of a game where earlier, I'd just kick the FG because it might not come down to a very desperate situation at the end

 

[OttoinGrotto]

I like the thought process. My question is this: At what point in the game, time wise, is it necessary to start taking these risks?
Would you go for 4th & 1 from the 40 with 10 min. left?
Would you do it half way through the 3rd?
Is the same timeframe applicable to the first half; even in a tie game in the 2nd quarter would you risk the field position with two minutes before half for a go ahead score?

I'm not trying to poke holes, I like what I hear. Just wondering how aggressive is too aggressive if you adhere to this philosophy.

I think we have people on the absolute extreme end that love kicking and punting.

I'd love to see what someone on the absolute extreme end of "*** it, we're going for it" looks like. I think it would be fun.

 

[690West]

I think we have people on the absolute extreme end that love kicking and punting.

I'd love to see what someone on the absolute extreme end of "*** it, we're going for it" looks like. I think it would be fun.

Over the long run the "never punt guy" would have a better win percentage. Problem is that coaches don't get "the long run" leash to make correct choices. Or at least that's their perception and thus meathead decisions week after week

 

[NKR1978]

I think we have people on the absolute extreme end that love kicking and punting.

I'd love to see what someone on the absolute extreme end of "*** it, we're going for it" looks like. I think it would be fun.

Aren't their some high school coaches that do it?

I don't punt inside my opponent's 40 on NCAA14 and get chastised for it by Herbstreit.

 

[Scooch]

When it comes to NFL coaches always playing to extend the game rather than taking the best chance to win, I always think of starting pitchers in baseball playoffs.

Team X is in a five game series down 2 games to 1 and can either start their ace on 3 days rest in game 4 (let's assume said pitcher tends to be less effective on short rest), or start another pitcher on full rest in game 4 and give their ace full rest to pitch in game 5. I have heard baseball fans passionately advocate for starting the ace on short rest with the logic of "you have to win game 4 to get to game 5!".

Well, no duh.

But to win the series you have to win 2 games, and typically letting two pitchers go on full rest maximizes your chance to win each game.

NFL coaches who play to not lose for longer while ignoring their best chance to win make me insane.

 

[RF2044]

i've posted it 10,000 times but i wish everyone would read thinking fast and slow. big theme in it is how when faced with a difficult question (what maximized our win probability), people will substitute an easier one (what maximizes the amount of time we have any shot at all)

Is that the Kaneman / Teleb collaboration?

 

[Bayside44]

When it comes to NFL coaches always playing to extend the game rather than taking the best chance to win, I always think of starting pitchers in baseball playoffs.

Team X is in a five game series down 2 games to 1 and can either start their ace on 3 days rest in game 4 (let's assume said pitcher tends to be less effective on short rest), or start another pitcher on full rest in game 4 and give their ace full rest to pitch in game 5. I have heard baseball fans passionately advocate for starting the ace on short rest with the logic of "you have to win game 4 to get to game 5!".

Well, no duh.

But to win the series you have to win 2 games, and typically letting two pitchers go on full rest maximizes your chance to win each game.

NFL coaches who play to not lose for longer while ignoring their best chance to win make me insane.

But leaving your best pitchers on ice to be like Matt Williams is baseball meatheadedness.

Different conversation...

 

[KaiserUEO]

this is why its good to be a Giants fan.

we love our punters.

despite the numerous SuperBowl wins, quite often the punter is the highlight of the day.

craig morton and joe pisarcek cant get it going?? watch this, Dave Jennings is about to loft a 45yard punt with a hang time of 3 minutes that brings rain...

Simms goes 3 & out on his own 10?? watch this, Sean Landeta is about to launch a 75 yd punt.

Eli stalls on the opposite 45 or 50?? watch this, Jeff Feagles is about to stick them inside the 5.

of course more times then not, they have a great D...

 

[OttoinGrotto]

Over the long run the "never punt guy" would have a better win percentage. Problem is that coaches don't get "the long run" leash to make correct choices. Or at least that's their perception and thus meathead decisions week after week

I'd like to see some exhibitions where the coaches have an agreement that the only kicks either will attempt are kickoffs, just to see how that changes things.

 

[Millhouse]

Is that the Kaneman / Teleb collaboration?

Kahneman yes. Taleb no.