Going for 2 after scoring a touchdown to get within 8. Yes, there will be math

[Irieguy 1059]

Dr. Beeper, you aren’t a big numbers guy. Your partners in the betting contest are. The ones who actually put together the spreadsheet. You don’t even understand what blacklab did with his spreadsheet. He allowed you to change assumptions so you DON’T have to rely on league averages. You are out of your element and looking silly. You can prove me wrong by posting a response that shows your assumptions, assigns percentages to those assumptions and then use those assumptions and percentages to calculate win percentages. 

Edited November 12, 2019 by Irieguy

[Bartles 769]

I liked what the Panthers did. When you have a player like McCaffrey, nothing wrong with being on the aggressive side of the go-for-2 chart (though they didn't use him on that play). The Eagles have a lot of decent run or pass plays to get 2's, and their defense is also pretty bad which makes it even more urgent to take the lead and get out of the stadium before OT.

But I can see an offensively challenged team with a good defense (Bears e.g.) deciding to kick PATs and rely on the D to continue to set them up for conventional scores, even if that means OT is the best case.

 

[Huckleberry 34657]

7 hours ago, Dr. Beeper said:

I’m a numbers guy and don’t claim to be an expert in football.

...

if I tie this game, I have a better shot than the Packers in OT because I have all momentum.

lulz

[Armybrat 41408]

And the Iowa State coach was absolutely correct to go for 2 at the end of their game with blowU because he has balls and wants to win.

Same with WVU against us last year.

if it doesn’t work, take the loss like a man and not snivel & snot around about it.

Edited November 12, 2019 by Armybrat

[NowThis 931]

4 minutes ago, Armybrat said:

And the Iowa State coach was absolutely correct to go for 2 at the end of their game with blowU because he has balls and wants to win.

Same with WVU against us last year.

if it doesn’t work, take the loss like a man and not snivel & snot around about it.

have to consider game conditions. Suppose Hurts had a twisted ankle and the OU kicker can't make anything greater than 30?

[TXSooner518 2455]

12 hours ago, Dr. Beeper said:

I’d employ analytics in most situations. Starting in the first quarter. I’d let situations dictate whether or not to employ such statistical plays. In that situation that late in the game, take the damn near sure thing, make it a 7 point game, and save, likely, one of your better 2-point conversion plays for later. 

I also don’t like going for 2 if Carolina were to score again in the 4th (assuming the score were 24-23 at that point). Nor do I like ISU going for 2 against OU. Yes, I know - on the road, but in both situations the Panthers and Cyclones would’ve had all the momentum. I would’ve played for OT at that point. 

If you’re in a shootout with OU, or Rodgers, in Norman or Green Bay, you absolutely go for 2. 

I mean, the OU-ISU game was 42-41 at that point in Norman. Also, you think the team with "momentum" wins more often in OT, but in fact it is the better team that wins. I will bet you a lot of money that if you compare the WP of "team that scored last to force OT" vs "team that was favored to win", "team that was favored to win" will be higher.

9 hours ago, Dr. Beeper said:

I’m a numbers guy and don’t claim to be an expert in football. I’ve coached a lot and can manage a game/clock, and clearly this is my personal opinion. 

If I were Rivera, my thoughts would be: I’m 50/50 on any 2-point attempt, I’m 100% on any XP, and if I tie this game, I have a better shot than the Packers in OT because I have all momentum.  

You aren't 100% on an XP. In the NFL, XPs are 93% now after the change. This was in wind and snow, so less than that. Also, if the Panthers are actually 50/50 to hit a 2pc, and more than 50/50 to win in OT then going for 2 is even better than Bob posted, especially since they are more like 90% to make an XP, meaning more like 81% to hit them both. 

8 hours ago, blacklab said:

YES!

Another one I learned in Madden was when down 10 kick a FG as soon as you're in range and go for the onsides. A few NFL and college teams have done it and the announcers were like "WTF ARE THEY DOING"

This was better before they ruined onside kicks, but yeah. Teams always burn the entire clock to get the TD. They need to allow you the option of kicking off or taking it 4th and 15 from your own 35. 

8 hours ago, orangecat92 said:

I agree with this.  The thing about the Iowa state game was I think sooners were totally surprised, but I think the refs stopped the game, after the TD, to give OU time to recover, and take a breath.  OUs defense had been on the field quite a bit.  I think the interception thrown by OU was done pretty early in the drive.  If the refs had not stopped the game, would OU coach have been smart enough to call timeout to give his team time to recover?  

I especially like the part about telling the offense you have confidence they will get the two yards.  As the underdog, I think the object should be to end the game in regulation.  Why give the favorite another chance with the game tied?  Make them uncomfortable.

It was actually great for ISU that they stopped to review. The play clock was running down and ISU was unprepared to run a play and were about to get a delay of game, so they would have had to run the play from the 7. If there was less time left, Riley def should have taken a timeout if ISU wasn't going to get a DOG, but with 30 seconds left, I'm OK saving the 3 timeouts to maximize the chance of answering with a GWFG.

1 hour ago, NowThis said:

have to consider game conditions. Suppose Hurts had a twisted ankle and the OU kicker can't make anything greater than 30?

Sure, but neither of those was the case. The point is, you should default to going for 2 when you score a TD in the 4th when down 14, and only kick if there's a specific reason to, instead of the opposite. It's a closer decision when scoring a TD down 1, and you should rely more exclusively on the game. Momentum is a consideration, but drastically behind things like which team is better, which team is home, critical injuries, red zone offense/defense proficiencies, kicker strength/weakness, etc.

[Armybrat 41408]

Another example from DKR was the 1964 game with pig.
Ernie Koy scored with just over a minute left in the 4th to bring the Horns within one (13-14), but Marvin Kristynik’s 2 point conversion pass to Hix Green bounced on the turf. Cost UT a consecutive MNC.

I had a hot date who flew in from Fort Lauderdale for the weekend, so she consoled me properly.

[DaggerHorns 3170]

18 minutes ago, Armybrat said:

I had a hot date who flew in from Fort Lauderdale for the weekend, so she consoled me properly.

Further analysis would prove that your chance of getting laid would plummet to <50% had the horns won the game.

Horns win: Going out with the boys afterwards and crushing beers accompanied with victory shots would lead to >97% whiskey dick thus lowering sexual performance and probability of intercourse. The chance of the girl ditching would also lead to >50% of no showing for an impending fuck fest postgame. Chance of fucking —> 48.5%
 

Horns lose: Going back to hotel, bypassing post game booze fest 99%. Accompanied by out of town Floridian looking for a postgame round of “hide the sausage” —> 70%. Chance of fucking —> 69.3%

[LTtxfan 50005]

15 hours ago, blacklab said:

And if you need further proof it's the right thing to do, let's look at the game of the century, and what they guy with the name on our stadium decided to do: (from wikipedia)  James Street scrambled for a touchdown on the first play of the fourth quarter. Longhorns coach Darrell Royal had decided before the game to go for two after the Longhorns' first touchdown to avoid a tie, and Street dove into the end zone to make it 14–8. This was an unusual decision since most teams would have delayed the decision for a two-point conversion until the next touchdown (though clearly mathematically correct, since this would have given the Longhorns a later chance to tie in case of a failed two-point try).

• Here's James Street talking about that 2pt play (at 1:28 mark)
• And the 2pt play (at 4:00 mark)

 

Edited November 12, 2019 by LTtxfan

[DanTheHorn 5079]

14 hours ago, WBT said:

You should also go for 2 when you score to cut the lead to 4.

https://fivethirtyeight.com/features/when-to-go-for-2-for-real/

He only took into consideration winning and losing and not being fired or ridiculed for being perceived to be a moron. Remember perception is reality regardless of what the truth really is in any scenario. That is why coaches don't do it. We have are own dumb ass going for it 4th because the numbers say so. Yet we don't buy his numbers.

[DanTheHorn 5079]

1 hour ago, Armybrat said:

Another example from DKR was the 1964 game with pig.
Ernie Koy scored with just over a minute left in the 4th to bring the Horns within one (13-14), but Marvin Kristynik’s 2 point conversion pass to Hix Green bounced on the turf. Cost UT a consecutive MNC.

I had a hot date who flew in from Fort Lauderdale for the weekend, so she consoled me properly.

Avatar makes sense now.

[JBJ 3827]

The rules are simple:

1) When down, only kick an XP when it gets you into a position to win instead of tie. (-3, -7, -14), but not at -10 because a 2PC makes it one-possession.

2) When up, only go for two when it gets you into a position where the opponent ties instead of wins. (1, 5, 12), but again not at 8 because an XP makes it a two-possession game.

Edited November 12, 2019 by JBJ

[TXSooner518 2455]

36 minutes ago, DanTheHorn said:

He only took into consideration winning and losing and not being fired or ridiculed for being perceived to be a moron. Remember perception is reality regardless of what the truth really is in any scenario. That is why coaches don't do it. We have are own dumb ass going for it 4th because the numbers say so. Yet we don't buy his numbers.

Most of Herman's go/no go decisions are good, though he does punt on 4th and shortish from opp territory too much, it's just that many of the playcalls have been beyond silly. 

[XYZ 2154]

15 hours ago, blacklab said:

Kevin and Chad on the radio today were discussing the choice by Ron Rivera to go for 2 when Carolina scored  on Green Bay to make it 24-16. Chad had no idea why they did it and thought it was stupid. He's wrong. It's the right thing to do.

If we assume GB doesn't score, Carolina scores another TD as the clock is running out, overtime is a 50/50 win/loss, current nfl average for 2 point conversions is 47.5% and current XP% this year is 93.3%. here's how it works out:

Kick XP on first

• make first, miss second, lose 100% of the time (happens 6.7%)
• Miss first
  • miss 2pt conversion (52.5%) - lose 100% (happens 93.3% x 52.5% = 49% )
  • make 2pt conversion (47.5%) go to overtime, lose 50% of the time (happens 44%)

So if you kick the first XP you're winning 44% of the time and losing 56%  of the time.

Go for 2 on first TD

• Make it (47.5%),
  • now you can kick xp on second and win (47.5% x 93.3% = 44% win)
  • miss xp on second, go to overtime and have a 50/50 chance of winning. (47.5% x 6.7% x 50% =  2% win 2% loss)
• Miss it (52.5%)
  • make 2 pointer on second TD, go to OT, win 50%  (52.5% x 47.5% x 50% = 12% win 12% loss)
  • miss 2 pointer on second TD lose (52.5% * 52.5% = 28% loss)

So if you go for 2 you will win 58% of the time and lose 42% of the time

And if you need further proof it's the right thing to do, let's look at the game of the century, and what they guy with the name on our stadium decided to do: (from wikipedia)

There’s a problem with your analysis, blacklab. You are equating “statistics” with “probability”. For example, just because the average success rate of the two-point conversion is 47.5%, that doesn’t mean that the probability of that team executing a successful two-point conversion, against that opponent, at that point in that game was 47.5%.

[SwanderedTalent 5761]

The data doesn't allow for the impact that getting two, getting one, or getting nothing have on the opponent's behavior, which is unknowable and where human intuition comes in. The data/the charts are valuable as inputs, so that you have some logical guardrails to intuition, but I wouldn't trust them as the automatic arbiter of what the correct decision is a specific situation. 

Edited November 12, 2019 by SwanderedTalent

[blacklab 12064]

You are correct, it's not. But it is a good starting point. The good Dr made the same argument when he said not to use NFL averages.

I guess I could take the next month or two and run a few billion simulations of the two teams to get the exact number, but it's going to be around that number.

Which is why I made the spreadsheet to show results across a range of possibilities. If the percentage is higher, then it makes going for two an even better option and if the percentage drops below 37% there will be some cases where if you have the greatest kicker of all time it makes more sense to kick. But if the chance of them making it is above 36% then they should go. The NFL is unlike college in there is parity. Most teams are pretty close in the level of talent they have with a few outliers making the difference between the best and worst offenses and defenses. I can expand the spreadsheet to see what it looks like, but it's going to be more of the same.

[Huckleberry 34657]

1 minute ago, blacklab said:

You are correct, it's not. But it is a good starting point. The good Dr made the same argument when he said not to use NFL averages.

I guess I could take the next month or two and run a few billion simulations of the two teams to get the exact number, but it's going to be around that number.

Which is why I made the spreadsheet to show results across a range of possibilities. If the percentage is higher, then it makes going for two an even better option and if the percentage drops below 37% there will be some cases where if you have the greatest kicker of all time it makes more sense to kick. But if the chance of them making it is above 36% then they should go. The NFL is unlike college in there is parity. Most teams are pretty close in the level of talent they have with a few outliers making the difference between the best and worst offenses and defenses. I can expand the spreadsheet to see what it looks like, but it's going to be more of the same.

Yep. If you have a guaranteed 100% chance of making the extra points then the break-even point on 2-point success rate is 38.2% assuming a 50% chance of winning in overtime. No need to expand the spreadsheet, just use this one:

https://docs.google.com/spreadsheets/d/1tAlVjQr7eNLhn3l_Ix5wn3ymR9xtWw4Nh_EndOCzoNE/edit#gid=0

And mess around with the inputs on the left.

[blacklab 12064]

6 minutes ago, SwanderedTalent said:

The data doesn't allow for the impact that getting two, getting one, or getting nothing have on the opponent's behavior, which is unknowable and where human intuition comes in. The data/the charts are valuable as inputs, so that you have some logical guardrails to intuition, but I wouldn't trust them as the automatic arbiter of what the correct decision is a specific situation. 

So Green Bay gets the ball back with 4 minutes left and up 8,7 or 6.

Are they going to do something besides trying to get first downs and run out the clock?

If Carolina stops them are they going to do something besides trying to score a touchdown?

[gsoda3 3362]

11 hours ago, youdunnf'dup said:

I don’t mean to always bring these analytics discussions back to madden, but I swear the top level madden players have been ahead of the curve on this one for a while now. 

for sure.  the more simulations you run the better data you have.  works like that everywhere i can think of.

[Irieguy 1059]

15 minutes ago, SwanderedTalent said:

The data doesn't allow for the impact that getting two, getting one, or getting nothing have on the opponent's behavior, which is unknowable and where human intuition comes in. The data/the charts are valuable as inputs, so that you have some logical guardrails to intuition, but I wouldn't trust them as the automatic arbiter of what the correct decision is a specific situation. 

Nah. Going for 2 is game theory optimal. Your opponent can’t do anything to change that. Seriously, what change could your opponent make after the 2 point that makes the 2 a bad decision? I’m genuinely curious. 

[SwanderedTalent 5761]

Just now, blacklab said:

You are correct, it's not. But it is a good starting point. The good Dr made the same argument when he said not to use NFL averages.

I guess I could take the next month or two and run a few billion simulations of the two teams to get the exact number, but it's going to be around that number.

Which is why I made the spreadsheet to show results across a range of possibilities. If the percentage is higher, then it makes going for two an even better option and if the percentage drops below 37% there will be some cases where if you have the greatest kicker of all time it makes more sense to kick. But if the chance of them making it is above 36% then they should go. The NFL is unlike college in there is parity. Most teams are pretty close in the level of talent they have with a few outliers making the difference between the best and worst offenses and defenses. I can expand the spreadsheet to see what it looks like, but it's going to be more of the same.

No, I personally would not change my mind based on a trillion simulations-- I already accept that the data will look counter-intuitive to me in places and that operating entirely by intuition is a suboptimal approach, that I would have to overcome instinct in some situations where the data says the instinct is wrong. 

The amount of trust I would place in the data on my clipboard would be informed by who was on the opposing sideline. So to your other question:

1 minute ago, blacklab said:

So Green Bay gets the ball back with 4 minutes left and up 8,7 or 6.

Are they going to do something besides trying to get first downs and run out the clock?

If Carolina stops them are they going to do something besides trying to score a touchdown?

I don't know Green Bay's staff well enough to have an opinion on what they might do. My response was more hypothetical than addressing this specific game, but I would expect Green Bay to be relatively more aggressive the smaller the margin. Ahead eight, it might be dive-dive-dive-punt (it's the NFL). Ahead seven, same (expecting that I would go for the tie and OT if I scored a TD). Ahead six, I think they would be more aggressive, no? It's no longer a full one-possession game since my making the XP is practically assumed. 

[SwanderedTalent 5761]

2 minutes ago, Irieguy said:

Nah. Going for 2 is game theory optimal. Your opponent can’t do anything to change that. Seriously, what change could your opponent make after the 2 point that makes the 2 a bad decision? I’m genuinely curious. 

All I can offer here is a hypothetical, which makes me feel a little naked where everyone else is covered up in data. But-- I think your opponent is less likely to engage the prevent offense if you're within six. That may, or may not, make it less likely for you to get the ball back-- I think an overly conservative offense (the kind a team will run to protect a lead) is easier to stop than a "regular" offense. 

This doesn't really feel like a very well-explained position so I'll drop it and just reiterate I'm not staking a position pro or con regarding Rivera's decision. In my gut, I don't like it, but I'm not saying I know or believe it's wrong. 

 

[runthebone 8377]

One part is assuming that you have multiple equally successful 2 point conversion plays.  But aside from that, it's math. 

[blacklab 12064]

I see your argument, but Carolina has no control over what GB does and does not know what they are going to do.

But, there are some people like me who work for Carolina. They may have stats on what GB does when up x points with y minutes left in the game and use that as part of their go for two or kick XP formula. 

And, I'd argue that if you know GB will go run up the middle three times if they have an 8 point lead, it's way better for you then taking a chance on Aaron Rodgers getting a first down or two throwing the ball.

You know you have to stop them and get the ball back. What you're arguing makes the calculation way more complex, but probably doesn't move the needle significantly in one way or the other, and I'd bet big money it doesn't overcome a 63-50 difference that going for two does.

[SwanderedTalent 5761]

Just now, blacklab said:

And, I'd argue that if you know GB will go run up the middle three times if they have an 8 point lead, it's way better for you then taking a chance on Aaron Rodgers getting a first down or two throwing the ball.

that's a really good point, I do want the most conservative response I can get from the opponent's offense because it will be easier to stop. I said as much in my response to Irieguy, which probably means I'm seeing this logically only where it fits my preconceived notions

[StruggleBus 3922]

17 hours ago, Agent 13 said:

None of the above calculations are accurate unless your particular team performs exactly at the NFL average.

And what happens to these fancy mathermatics if Jason Garrett is your head coach? 

[blacklab 12064]

2 minutes ago, StruggleBus said:

And what happens to these fancy mathermatics if Jason Garrett is your head coach? 

You're going to lose 98% of the time. 

[blacklab 12064]

Chad actually replied to my tweet, and basically just assumes you're going to miss the 2pt 100% of the time.

 

[TXSooner518 2455]

People just can't grasp the concept of actually being able to WIN THE GAME IN REGULATION having value.  BUT WHAT IF YOU LOSE. Also for some reason BUT WHAT IF YOU MISS only applies to the 2pc, but not the XP, which is now missed on a reasonable basis. So infuriating that people who make a living talking sports can't grasp it. 

You can pick one of two games:
1) we flip a coin and if it is heads I pay you $100 and if it's tails you pay me $90

2) we roll a die. If it is 1 or 2 I pay you $80, if it is 3 or 4 no one pays anything, if it is 5 or 6 you pay me $100

Coaches literally choose game 1 because they get paid half the time instead of 1/3 of the time.

Edited November 12, 2019 by TXSooner518

['stache 22143]

1. The statistics on 2-pt conversion makes is based on how they are usually are used, which is rarely, and most often in desperation circumstances. Teams all have a play or two they consider their best chance, and even still they are under 50%. If teams did them more often, my guess is that the % would be lower since the roster of "best" plays would become a little more diluted.

2. My guess is that the weather played a role in that decision more than statistics. There wasn't snow on the ground yet based on the highlights I saw but something was definitely blowing in because the end of the game had snow on the field.

3. Even from a statistical analysis, the % of making the extra point now that it is harder than a chip shot is less than the 97% when there is any kind of weather. The highlight of the TD in question showed some sideways snow and the ground within a few minutes was visibly white, which tells me there was not only cold but wind involved, and even if not visible on the field at the time, there likely was some snow/ice/moisture on the turf. That drops the % below 97%, probably by a significant margin. If it's a close call based on the stats in the OP, the weather may tip the scales in favor of the 2 based more on the weather. I'd be interested to see if Carolina ends up in that same scenario in a dome or clear weather game to see if they would make the same call. 

[Pimphand 6370]

The real head scratcher was Ron Rivera going for 2 when he was down 8 this Sunday in Green Bay.... Like WTF????

[TXSooner518 2455]

4 minutes ago, Pimphand said:

The real head scratcher was Ron Rivera going for 2 when he was down 8 this Sunday in Green Bay.... Like WTF????

Yeah!  We should discuss that for 90 posts or so!

[Don Johnson 4847]

Related/unrelated but not going to start a new thread. 

I've always felt if you are going to have to go for 2 at some point, you should do it sooner than later.  For example, if you are down by 15 and score with 5 minutes left, I'd rather go for 2 after the first TD than the second.  My logic is that you have one 2 point conversion to make and you either are going to make it or not. 

If I'm not going to make it, I'd rather know this sooner than later and I might have enough time to compensate.  I'd rather find out its going to take 3 scores with 5 minutes to go, than find that out with 30 seconds.

Yet no one ever does it that way.  

[Huckleberry 34657]

13 minutes ago, Don Johnson said:

Related/unrelated but not going to start a new thread. 

I've always felt if you are going to have to go for 2 at some point, you should do it sooner than later.  For example, if you are down by 15 and score with 5 minutes left, I'd rather go for 2 after the first TD than the second.  My logic is that you have one 2 point conversion to make and you either are going to make it or not. 

If I'm not going to make it, I'd rather know this sooner than later and I might have enough time to compensate.  I'd rather find out its going to take 3 scores with 5 minutes to go, than find that out with 30 seconds.

Yet no one ever does it that way.  

You are 100% correct but should be prepared for the guys who don't understand. This was a long thread on shaggy at one point, I believe.

[jimmyjazz 47205]

18 hours ago, blacklab said:

Kevin and Chad on the radio today were discussing the choice by Ron Rivera to go for 2 when Carolina scored  on Green Bay to make it 24-16. Chad had no idea why they did it and thought it was stupid. He's wrong. It's the right thing to do.hat they guy with the name on our stadium decided to do: (from wikipedia)

One rock solid rule of football strategy:  if Chad Hastings thinks it's a good idea, it's not, and vice versa.

[jimmyjazz 47205]

DISCLAIMER:  not a video game guy

 

Is Madden grounded enough in reality to consider it as a useful database?  I get the statistics being floated in this thread, but I'm surprised Madden is accurate enough to use as evidence.

[Irieguy 1059]

27 minutes ago, Pimphand said:

The real head scratcher was Ron Rivera going for 2 when he was down 8 this Sunday in Green Bay.... Like WTF????

You are out of your element Donny. 

[blacklab 12064]

Pretty much. The scoring is exactly the same so the same rules apply. The FG/XP percentage is probably way off for bad players but other than that it's a fine simulator for this.

[Beau Vine 55867]

2 hours ago, XYZ said:

There’s a problem with your analysis, blacklab. You are equating “statistics” with “probability”. For example, just because the average success rate of the two-point conversion is 47.5%, that doesn’t mean that the probability of that team executing a successful two-point conversion, against that opponent, at that point in that game was 47.5%.

So you ignored the chart he posted where he didn't use the average, but rather every single probability between 37% and 57% and it gave the exact same conclusion?

[Beau Vine 55867]

2 hours ago, SwanderedTalent said:

The data doesn't allow for the impact that getting two, getting one, or getting nothing have on the opponent's behavior, which is unknowable and where human intuition comes in. The data/the charts are valuable as inputs, so that you have some logical guardrails to intuition, but I wouldn't trust them as the automatic arbiter of what the correct decision is a specific situation. 

Then just choosing to kick an extra point like conventional wisdom has advocated for years, also doesn't allow for the impact that getting two, getting one, or getting nothing have on the opponent's behavior, which is unknowable and where human intuition comes in.  So why are we accepting that move as the standard?

[Beau Vine 55867]

41 minutes ago, Huckleberry said:

You are 100% correct but should be prepared for the guys who don't understand. This was a long thread on shaggy at one point, I believe.

You should also be prepared for their entire argument being, "Yeah, but if you miss that 2-pointer, the game is over.  Kicking the extra point keeps you in the game."  They'll repeat that over and over and then eventually resort to "I know what I'm talking about because I ACTUALLY PLAYED FOOTBALL."

[TXSooner518 2455]

Just now, Beau Vine said:

You should also be prepared for their entire argument being, "Yeah, but if you miss that 2-pointer, the game is over.  Kicking the extra point keeps you in the game."  They'll repeat that over and over and then eventually resort to "I know what I'm talking about because I ACTUALLY PLAYED FOOTBALL."

Ha yep. EXTEND THE GAME!  Better to have a 20% chance to win later than a 40% chance to win but find out immediately you lost! Also prepare for lots of talk of "keep it a one-possession game" except that until you attempt the 2pc, you don't actually know if it's a 1 or 2 possession game. It is obvious that more knowledge is better than less, except EXTEND THE GAME.

[ChiTownDoc 21345]

35 minutes ago, Beau Vine said:

 "I know what I'm talking about because I ACTUALLY PLAYED FOOTBALL."

Sorry about your CTE, bro. 

[SHOOTER12 8886]

On 11/11/2019 at 7:48 PM, Neonmoon said:

I was told there would be no math 

Lol, they lied

[Not a cat 5808]

8 hours ago, Irieguy said:

Nah. Going for 2 is game theory optimal. Your opponent can’t do anything to change that. Seriously, what change could your opponent make after the 2 point that makes the 2 a bad decision? I’m genuinely curious. 

 

 

Here's an example- you score to get within one with 90 seconds left and opponent has 3 timeouts.  If you kick the XP, it's a tie game and likely the opponent will play a little cautiously going for a game winning field goal rather than risk a turnover or giving you the ball back.  Let's say it's 25% they get it because you know the coach will turtle to preserve the tie.

If you go for 2 and make it they are going balls to the wall for a field goal or td and their chances go to 50%.

So with an XP (let's just assume 100%) you chances of winning are 37.5% (75% chance of OT, even chance of winning in that OT).  With a 2 pt (assume 50%) you have a 50% chance of losing because you didn't make it, and within the 50% of the time you make it you still lose half of those for an overall win rate of 25%.

I'm not saying the charts are wrong, I'm just saying that there are situations where your opponents reaction, if you can predict it,  should factor in.  Kind of like advanced game theory.

[Bullneck 19528]

Leach posited that you should go for it on 4th down a lot.  

Oh, and this:

 

Edited November 13, 2019 by Bullneck

[Irieguy 1059]

2 hours ago, Not a cat said:

 

 

Here's an example- you score to get within one with 90 seconds left and opponent has 3 timeouts.  If you kick the XP, it's a tie game and likely the opponent will play a little cautiously going for a game winning field goal rather than risk a turnover or giving you the ball back.  Let's say it's 25% they get it because you know the coach will turtle to preserve the tie.

If you go for 2 and make it they are going balls to the wall for a field goal or td and their chances go to 50%.

So with an XP (let's just assume 100%) you chances of winning are 37.5% (75% chance of OT, even chance of winning in that OT).  With a 2 pt (assume 50%) you have a 50% chance of losing because you didn't make it, and within the 50% of the time you make it you still lose half of those for an overall win rate of 25%.

I'm not saying the charts are wrong, I'm just saying that there are situations where your opponents reaction, if you can predict it,  should factor in.  Kind of like advanced game theory.

You are arguing against a point that I didn’t make. And I agree with you in the scenario you posted. But in the scenario that started the thread there is no change to opponent game plan whether you are down 6 or 8. 

Btw, a good model will include changes to opponent behavior like the situation you outline. But your hypothetical is not really that interesting. It’s play to win or play to tie. One decision. The Carolina situation is much more interesting and complex because it’s multiple decisions and the 1st decision influences the 2nd. 

[Valmy77 13626]

12 hours ago, DanTheHorn said:

He only took into consideration winning and losing and not being fired or ridiculed for being perceived to be a moron. Remember perception is reality regardless of what the truth really is in any scenario. That is why coaches don't do it. We have are own dumb ass going for it 4th because the numbers say so. Yet we don't buy his numbers.

I buy his numbers. I am totally on board with the amount of times Tom goes for it. I have my criticism but that is certainly not one of them.

That is just called being a 21st century head coach and not living in the archaic past.

[TheFlyingBoat 913]

1 hour ago, Valmy77 said:

I buy his numbers. I am totally on board with the amount of times Tom goes for it. I have my criticism but that is certainly not one of them.

That is just called being a 21st century head coach and not living in the archaic past.

This. Hell a if Ingram doesn't have butter fingers we probably beat LSU after going for two on the final score and call Herman a genius for going for it on 4th. Fact of the matter is most of us here are clueless and judge based on the results we see not on the soundness of the decision and the process behind it.

[Not a cat 5808]

7 hours ago, Irieguy said:

You are arguing against a point that I didn’t make. And I agree with you in the scenario you posted. But in the scenario that started the thread there is no change to opponent game plan whether you are down 6 or 8. 

Btw, a good model will include changes to opponent behavior like the situation you outline. But your hypothetical is not really that interesting. It’s play to win or play to tie. One decision. The Carolina situation is much more interesting and complex because it’s multiple decisions and the 1st decision influences the 2nd. 

You asked for "a change could your opponent make after the 2 point that makes the 2 a bad decision. " I gave you one.  Admittedly not the most interesting one, but it's the most obvious one I see at the end of a lot of games.