Fight or Bend the Knee?

[K-S-U-Wildcats!]

what kind of term is "bend the knee"?  a weird one.  just say kneel, my goodness.

No. Winter is coming.

[star seed 7]

I don't care about either decision

[DQ12]

Really, I think that situation is one where there's really "no bad decision."  Either decision he makes still results in an overwhelming statistical probability that ISU wins.  I think we're talking about moving the odds of k-state winning from maybe like 2% to 3%. 

[Trim]

Really, I think that situation is one where there's really "no bad decision."  Either decision he makes still results in an overwhelming statistical probability that ISU wins.  I think we're talking about moving the odds of k-state winning from maybe like 2% to 3%. 

I think kneeling out was clearly the correct choice, but the reality is ISU was losing that game no matter what, somehow.  They suck, and they were playing K-State.

[catastrophe]

Really, I think that situation is one where there's really "no bad decision."  Either decision he makes still results in an overwhelming statistical probability that ISU wins.  I think we're talking about moving the odds of k-state winning from maybe like 2% to 3%. 

I think kneeling out was clearly the correct choice, but the reality is ISU was losing that game no matter what, somehow.  They suck, and they were playing K-State.

Yep. Let the stats say what they say, but ISU always had a 100% chance of losing the game. That's just the kind of team they are. I mean, even Cantele nailed a clutch long FG in cold weather to win it. The same guy who got benched just last year.

[stunted]

I didn't read the whole post because the 80+ yards is made up, there is no way that punter is unloading a 50 yard plus punt in that wind at that time in the game. The first down and touchdown percentages are also fraudulent because they aren't based what actually happened at that specific game to that point. Disregarding all of this obviously using nfl stats for two average college teams is obviously flawed.

50+ yards, 80+ yards, whatever



you can count on mir to discount logic and statistics. you are really an educator? this is how you figure situations with imperfect information, and on average it would be the best decision to make. a lot better than "it is definitely x because it's just my opinion" which is really stupid logically. there's a reason analytics is winning in sports. if you can find better numbers, it would be interesting to see how it affects the result.

using average nfl stats for 2 average teams sounds about right. it is a similar enough game, and i have to ballpark it anyways. but here's how it looks when i swing everything in favor of kneeling:



4 touchdowns more. and this assumes they're fumbling 1 out of every 33 carries when they're going to be safer with the football. fumbles is a high variance stat, while fumbling a lot earlier will make it more likely it happens, assuming one is going to fumble this much is gambler's fallacy. nfl top 10 rush attempts average 1.2% fumbles.

also if we're going off how the game is going to that point, taking out warren's long run, he's still averaging 5 yards a carry. isu making a first down at 50% seems pretty low, but ksu getting a td 50% of the time from the 50 is way too high. the best teams score touchdowns 66% of the time in the redzone (from any yard line), and now you have the clock against you without timeouts. ksu does well in the redzone, but it's not because of passing.

it's all a close decision anyways. to be 100% sure of a decision based on some guess, makes you a dumbass.

[catastrophe]

Look man, I'm pretty sure ISU is not firing the chart. They just let Rhoads go because he relied on a laminated piece of paper (which probably agrees with your analysis) instead of paying any attention whatsoever to what was happening in the rough ridin' game.

The guy didn't even realize his shoe was untied until KSU scored with 40 seconds left in the game.

[Kat Kid]

I didn't read the whole post because the 80+ yards is made up, there is no way that punter is unloading a 50 yard plus punt in that wind at that time in the game. The first down and touchdown percentages are also fraudulent because they aren't based what actually happened at that specific game to that point. Disregarding all of this obviously using nfl stats for two average college teams is obviously flawed.

50+ yards, 80+ yards, whatever



you can count on mir to discount logic and statistics. you are really an educator? this is how you figure situations with imperfect information, and on average it would be the best decision to make. a lot better than "it is definitely x because it's just my opinion" which is really stupid logically. there's a reason analytics is winning in sports. if you can find better numbers, it would be interesting to see how it affects the result.

using average nfl stats for 2 average teams sounds about right. it is a similar enough game, and i have to ballpark it anyways. but here's how it looks when i swing everything in favor of kneeling:



4 touchdowns more. and this assumes they're fumbling 1 out of every 33 carries when they're going to be safer with the football. fumbles is a high variance stat, while fumbling a lot earlier will make it more likely it happens, assuming one is going to fumble this much is gambler's fallacy. nfl top 10 rush attempts average 1.2% fumbles.

also if we're going off how the game is going to that point, taking out warren's long run, he's still averaging 5 yards a carry. isu making a first down at 50% seems pretty low, but ksu getting a td 50% of the time from the 50 is way too high. the best teams score touchdowns 66% of the time in the redzone (from any yard line), and now you have the clock against you without timeouts. ksu does well in the redzone, but it's not because of passing.

it's all a close decision anyways. to be 100% sure of a decision based on some guess, makes you a dumbass.

Actual stat from the day:

44 carries and 8 fumbles.

[stunted]

Actual stat from the day:

44 carries and 8 fumbles.

https://en.wikipedia.org/wiki/Gambler%27s_fallacy#Reverse_fallacy
http://www.amazon.com/Fooled-Randomness-Hidden-Markets-Incerto/dp/0812975219

[DQ12]

Actual stat from the day:

44 carries and 8 fumbles.

https://en.wikipedia.org/wiki/Gambler%27s_fallacy#Reverse_fallacy
http://www.amazon.com/Fooled-Randomness-Hidden-Markets-Incerto/dp/0812975219

I think it's absurd to assume that fumbling is perfectly odds based. 

[michigancat]

I didn't read the whole post because the 80+ yards is made up, there is no way that punter is unloading a 50 yard plus punt in that wind at that time in the game. The first down and touchdown percentages are also fraudulent because they aren't based what actually happened at that specific game to that point. Disregarding all of this obviously using nfl stats for two average college teams is obviously flawed.

50+ yards, 80+ yards, whatever



you can count on mir to discount logic and statistics. you are really an educator? this is how you figure situations with imperfect information, and on average it would be the best decision to make. a lot better than "it is definitely x because it's just my opinion" which is really stupid logically. there's a reason analytics is winning in sports. if you can find better numbers, it would be interesting to see how it affects the result.

using average nfl stats for 2 average teams sounds about right. it is a similar enough game, and i have to ballpark it anyways. but here's how it looks when i swing everything in favor of kneeling:



4 touchdowns more. and this assumes they're fumbling 1 out of every 33 carries when they're going to be safer with the football. fumbles is a high variance stat, while fumbling a lot earlier will make it more likely it happens, assuming one is going to fumble this much is gambler's fallacy. nfl top 10 rush attempts average 1.2% fumbles.

also if we're going off how the game is going to that point, taking out warren's long run, he's still averaging 5 yards a carry. isu making a first down at 50% seems pretty low, but ksu getting a td 50% of the time from the 50 is way too high. the best teams score touchdowns 66% of the time in the redzone (from any yard line), and now you have the clock against you without timeouts. ksu does well in the redzone, but it's not because of passing.

it's all a close decision anyways. to be 100% sure of a decision based on some guess, makes you a dumbass.

how did you calculate ISU first down %?

[stunted]

how did you calculate ISU first down %?

this stat was sort of a guess, i couldn't find anything for this. i noticed isu had 21 firsts, 5 td's  and was 5/11 on 3rd. also 5 turnovers (but this is way higher than average). but even that, 26 conversions, 11 fails is still 70% converting a first or touchdown.

i had that bend/break post, and i had them converting first downs at 80%+ and it seemed to check out, i dont remember if i got that number from anywhere.

[Kat Kid]

Actual stat from the day:

44 carries and 8 fumbles.

https://en.wikipedia.org/wiki/Gambler%27s_fallacy#Reverse_fallacy
http://www.amazon.com/Fooled-Randomness-Hidden-Markets-Incerto/dp/0812975219

Game conditions and ability of players are obviously major inputs.  It isn't random chance.  But carry on.

[michigancat]

how did you calculate ISU first down %?

this stat was sort of a guess, i couldn't find anything for this. i noticed isu had 21 firsts, 5 td's  and was 5/11 on 3rd. also 5 turnovers (but this is way higher than average). but even that, 26 conversions, 11 fails is still 70% converting a first or touchdown.

i had that bend/break post, and i had them converting first downs at 80%+ and it seemed to check out, i dont remember if i got that number from anywhere.

It's a pretty major factor of your calculation and almost impossible to accurately estimate, given the unique scenario and number of factors - plus you can't use their previous first down numbers to justify what you came up with just like you can't use the fumble numbers during the game to expect a fumble.

[EMAWican]

[stunted]

It's a pretty major factor of your calculation and almost impossible to accurately estimate, given the unique scenario and number of factors - plus you can't use their previous first down numbers to justify what you came up with just like you can't use the fumble numbers during the game to expect a fumble.

changing the first down 10% in either direction only changes touchdowns by 2.

when you don't have the stats, and it seems normal (21 first downs, 5 touchdowns and 5/11 3rd down, it's not like 8 fumbles), you can use it more confidently. if anything because of turnovers, first down % should be higher. something to do with standard deviation or bell curves

[stunted]

Game conditions and ability of players are obviously major inputs.  It isn't random chance.  But carry on.

it was never random chance.

anyways using 8 fumbles is off when he had 1 in 21 carries before. before his last fumble, he was 1.9% for the season (his last fumble was unknown beforehand obv), now he is at 2.4% (he only had 3 fumbles before this game).

here it is with data improved. you could argue for more fumbles, i could argue for better 1st down %.

[michigancat]

Game conditions and ability of players are obviously major inputs.  It isn't random chance.  But carry on.

it was never random chance.

anyways using 8 fumbles is off when he had 1 in 21 carries before. before his last fumble, he was 1.9% for the season (his last fumble was unknown beforehand obv), now he is at 2.4% (he only had 3 fumbles before this game).

here it is with data improved. you could argue for more fumbles, i could argue for better 1st down %.

IMO all the numbers are skewed because KSU is both placing extra emphasis on stopping the run and creating turnovers. (granted, ISU should have been placing emphasis on preventing turnovers, too). The first down numbers might get more predictable when passing is included as an option, but then the clock situation completely changes.

I don't think you're necessarily wrong, but I don't think the data you're using is sufficient to confirm that you're right, either.

[catastrophe]

Where is the Will Geary factor? I don't see it on here.

[Kat Kid]

The hilarious part that we are obscuring with this nonsense diversion is that Iowa State very likely could've run out the clock without even risking getting tackled.

[wetwillie]

When the origami rose is around there is always a risk.

[K-S-U-Wildcats!]

The hilarious part that we are obscuring with this nonsense diversion is that Iowa State very likely could've run out the clock without even risking getting tackled.

Yeah, I'm confused. So could they have knelt out the clock or would they have had to punt? Does it depend? On what?

[stunted]

Game conditions and ability of players are obviously major inputs.  It isn't random chance.  But carry on.

it was never random chance.

anyways using 8 fumbles is off when he had 1 in 21 carries before. before his last fumble, he was 1.9% for the season (his last fumble was unknown beforehand obv), now he is at 2.4% (he only had 3 fumbles before this game).

here it is with data improved. you could argue for more fumbles, i could argue for better 1st down %.

IMO all the numbers are skewed because KSU is both placing extra emphasis on stopping the run and creating turnovers. (granted, ISU should have been placing emphasis on preventing turnovers, too). The first down numbers might get more predictable when passing is included as an option, but then the clock situation completely changes.

I don't think you're necessarily wrong, but I don't think the data you're using is sufficient to confirm that you're right, either.

yea can never know for certain with these things, but one thing that is for sure is that it is close either way. one argument for kneeling and punting is the loss can't be pinned on a coaching decision.

wonder how vegas does their thing. i really like this stuff

The hilarious part that we are obscuring with this nonsense diversion is that Iowa State very likely could've run out the clock without even risking getting tackled.

comparing the situations in the thread topic = nonsense diversion

[Rage Against the McKee]

The hilarious part that we are obscuring with this nonsense diversion is that Iowa State very likely could've run out the clock without even risking getting tackled.

Yeah, I'm confused. So could they have knelt out the clock or would they have had to punt? Does it depend? On what?

They would have had to punt, but they could have just punted the ball out of bounds, giving K-State 1 play to go 80 or so yards.

[Trim]

The hilarious part that we are obscuring with this nonsense diversion is that Iowa State very likely could've run out the clock without even risking getting tackled.

Objectively, yes, and they should've tried.  But they'd have mumped it up.