Got Stat skills... here's a shot at 500 points

[tto827]

(Mod please link to think tank as well)

The math isn't hard, but I don't have the databases or knowledge to determine the true percentages so I could use some help.

I'm basically positive Tre Mason would have been better off kneeling at the 1 instead of scoring with slightly over a minute left last night. Given the situation, I would not expect him or Malzahn to be thinking of such a thing.

But if someone can determine a somewhat accurate win percentage for Auburn immediately after scoring and compare it to first and goal at the 1 (forcing FSU to take there 2nd timeout), I'll toss you 500 points.

At minimum you need to consider
FSU OT % win
Auburn scoring % first and goal on the 1
Auburn TD percentage first and goal on the 1

If you have the ability to consider all possible options playing out from the kneel down to end of game, that would be fantastic. But even if you just assume any Auburn TD would leave FSU with 1 timeout and about 1:15 left, that would be worth the 500 points to me.

Good luck

If you post in this thread that you are attempting to get a solution... I will give you a 20 minute exclusivity window so that you don't have to worry about someone beating you too it.

[Chi_archie]

are you doing chest or back in the gym?

[ChuckyTheGoat]

If they had been up 2, would have been a big deal. Up 3, tho, is tricky.

Still have to score the TD at some point. And that might have been next play, so only burn 5 seconds.

[warbux]

bout 86.74% auburn

[tto827]

are you doing chest or back in the gym?

If someone could give me semi-accurate numbers for these three things:
FSU OT % win
Auburn scoring % first and goal on the 1
Auburn TD percentage first and goal on the 1

I'd gladly do it myself. But I'd be pulling these numbers out of my ass, so its a waste of time for me to do the math knowing it isn't based on anything.

200 points if someone can give me some sort of statistically backed answer... (has to be somewhat team specific can't just be NCAA averages) of just those three things.

[SamDiamond]

Here's a variable that you can't quantify.

Suppose Mason falls down at the 1.

And, FSU calls a TO.

There is a very real possibility that FSU would allow Auburn to score on the very next play in order to get the ball back.

Which wouldn't have changed anything.

Because of that possibility--- there is no way to calculate the absolute risk-potential success for Mason to fall down at the 1.

It would hinge on how FSU's plays it from that point moving forward.

If you want math...and you want to discount FSU allowing Auburn to score......consider this.

Auburn's FEI for all drives was 75% this year.

To gain 1 yard in 4 tries---

Their chances would be north of 75%.

[Footballtime]

Auburn hasnt been stopped at the 1 all year, 3 downs they get in easy........next play or play after.......99% Auburn scores a TD from there...........Florida state would of had less time, but shit they scored with 13 seconds left anyway..........

[tto827]

FSU with 1 timeout vs. 2 is extremely critical. That would have to be factored in.

As I said, don't bother working out the time burnt off the clock, for the sake of making the math simpler, assume any TD scored after the kneel caused there to be about 3 seconds run off and FSU to take a timeout.

[tto827]

Auburn hasnt been stopped at the 1 all year, 3 downs they get in easy........next play or play after.......99% Auburn scores a TD from there...........Florida state would of had less time, but shit they scored with 13 seconds left anyway..........

Exactly.

FSU burns a TO after the kneel at the 1.. they likely can't afford to take the TO at 46 seconds leaving them with none, because a sack or any miscommunication likely ends the game. Usually takes a min of 6-7 seconds to spike a ball, so things get very interesting.

[The Kraken]

0%

thanks

[tto827]

Closed... Best answer found elsewhere, will post a link later

[yisman]

It makes a big difference starting a drive with 1 timeout rather than two timeouts.

[gui_m_p]

Don't know if all my % are right but here are some insights:

With 1:19 left in 4th (after TD), Auburn was an 79% favorite to win according to a prediction machine;

The chance to Auburn score a TD (1st and goal on 1yd) was around 80%. If they scored you can assume that, with less time on the clock, and FSU with one timeout at max, they would be at least 90% to win the game.

So they win 80%*90% of the games = 72%

Now the 20% games left (Auburn do not score TD): you can assume an overtime game, because after the field goal, wouldn't be time left. In OT, the chances to Auburn win would be around 35%.

So they win 35%*20%= 7%.

7% + 72% = 79%.

I say the chances would be very equal.

[BennyBigNuts]

Don't know if all my % are right but here are some insights:

With 1:19 left in 4th (after TD), Auburn was an 79% favorite to win according to a prediction machine;

The chance to Auburn score a TD (1st and goal on 1yd) was around 80%. If they scored you can assume that, with less time on the clock, and FSU with one timeout at max, they would be at least 90% to win the game.

So they win 80%*90% of the games = 72%

Now the 20% games left (Auburn do not score TD): you can assume an overtime game, because after the field goal, wouldn't be time left. In OT, the chances to Auburn win would be around 35%.

So they win 35%*20%= 7%.

7% + 72% = 79%.

I say the chances would be very equal.

Very solid post here.
Except for Auburn winning 35% in OT.
That's way off for sure. You can't tell me FSU is 65-35 in overtime. No chance that's accurate and don't care what ESPN machine spit that number out.
It's wrong.

[sweep]

deem you can come out and play now, problem ~solved..........

[gui_m_p]

Very solid post here.
Except for Auburn winning 35% in OT.
That's way off for sure. You can't tell me FSU is 65-35 in overtime. No chance that's accurate and don't care what ESPN machine spit that number out.
It's wrong.

Thanks!
Do you think chances of Auburn should be higher than 35% in OT? In this case it would be better Tre Mason not to score TD.
Anyway it's real close.

[MoneyLineDawg]

Still think you take the TD and play defense....they should have been extremely aggressive though once FSU was around the redzone......If FSU burns you, atleast you have a minute or so and down 3....The bend and eventually break defense gave them zero chance with 12 seconds left

[yisman]

Auburn had to be at least 45% to win in OT

[tto827]

Credits to Crazy Pete across the street for this answer:

There are some huge practical problems with Mason 'taking a knee at the 1'. He was running at
(nearly) full speed. He cannot instantaneously stop. He really cant slow down much, he doesn't have eyes behind his head to know how close a defender is. Physically, how can he stop? Slam his knee into the ground running full speed? He shreds his leg. Run out of bounds? No, that stops the clock! Maybe do some sort of slide, but he still has to slow down for that, very physically dangerous, and the spot could easily wind up on the 2 or 3.

Run parallel to the end zone? That risks a penalty, like teammates taking their helmet off in celebration without realizing the play was not over (that almost happened in their TD return in the Iron Bowl). And there are drawbacks to telling a player to tank at the one when they are 35 yards away. So, no, a 'knee at the one' was NOT going to happen.


BUT...if this was Madden, and you could push a button and have the ball placed at the one, without worrying about HOW he defies laws of physics and does that, then the question is: Is that the right move in terms of expected value? IMO, the answer is yes. Quick breakdown..
The live in-game odds of FSU winning after Mason's TD was what, 46%? That sounds too high for me (too high helps the 'score a TD' side), so I'll be a bit conservative and say 40% So Auburn wins 60% of time by Mason scoring (54% using live odds).

Okay, Auburn has 1st and goal from 1. Let's assume 50% chance of scoring each play. Let's ignore the fact that if FSU was smart, they'd LET THEM SCORE on 1st down.
So, odds of Auburn scoring on 1st down = 50% . FSU loses a few seconds and a timeout. Chances to win, imo, now drops from 40% to no more than 36%. So half the time they add 4% EV. That's +2% of EV for Auburn.

50% of time, Auburn goes to 2nd down. They score TD 50%, burn 2 FSU timeouts and maybe 10 seconds. Now FSU has to score TD with no timeouts and @ 1 minute. IMO, a wag would be FSU wins 24% at most. That means that they lost 16% EV, this happens 50% x 50% =25%, meaning Auburn picks up another 4% of EV.

OK, 3rd down happens 1 time in 8 (.5 x .5 x .5). They score half the time by hypothesis. They killed the 2 FSU timeouts, and what, maybe 50 seconds, leaving FSU needing a TD with what, @ 25 seconds left? Yikes! That's down to what, 8% at most. So FSU loses 40% -8% = 32% of EV, 1/8 of the time. So Auburn picks up another 4%. That's a net of 2% +4% +4%= +10% EV for Auburn so far.

But now, Auburn is in a -EV situation. Lets say, for simplicity sake, they kick the FG. Now game is tied, with a few seconds left. FSU is a modest favorite in OT, and could score a longshot TD in regulation. I will arbitrarily call Auburn now a 48-52 underdog. That's down from 60-40. That costs Auburn -24%in EV.... BUT...that only happens .5 x .5 x .5 = 1/8 of the time. So, they lose -24% x 1/8 = -3% of EV.

So, IMO, if Mason could have taken a knee at the 1...WHICH IS NEARLY IMPOSSIBLE TO DO..., Auburn increases their EV by around 7%. And that's making conserative assumptions. It's probably closer to 8 or 9%
This applies to Madden. It does NOT apply to real life. But, people arguing that they should have gone for the TD, assuming it was POSSIBLE to do that, are wrong imo.

[Wulfman14]

Tto can ya spare 1000 points. Standard 20%. 1% on the outstanding balance added on every week. I got crushed at 5 dimes because of the bengals . I owe a lot of points though so up to you.

[Wulfman14]

As for the answer to this one dunno I'd have to look into it. Doubt I'd come up with anything better then crazy Pete. I forgot all that stuff even though I studied multivariable calculus and diff eq. You don't use it , it goes away. Lol

[konck]

Tto can ya spare 1000 points. Standard 20%. 1% on the outstanding balance added on every week. I got crushed at 5 dimes because of the bengals . I owe a lot of points though so up to you.

[Wulfman14]

No way to generate points with no poker.and I don't only play aa , kk, qq and sit for 2hours in the dailies every single day fruit loop. I'm not retired and ya know.....old

[tto827]

Check points forum Wulfy... tried to but can't PM you

It's a lot of work without statistically valid numbers backing it, Pete did it as well as anyone without large databases could.

[InTheDrink]

Tto can ya spare 1000 points. Standard 20%. 1% on the outstanding balance added on every week. I got crushed at 5 dimes because of the bengals . I owe a lot of points though so up to you.

wulfer where we at pally?

[Wulfman14]

wulfer where we at pally?

i havent forgottten drinker. got shit on this weekend in the bowl games man. sbr took away my PMs. will be back in no time brotha.

[InTheDrink]

No worries hausen

Hausen what happened to your pm's?

[Wulfman14]

ask sbr forum

[InTheDrink]

Lou check in pal

[VegasInsider]

tto - just saw this posted on Deadspin. Great read.

http://www.slate.com/articles/sports/sports_nut/2014/01/auburn_florida_state_bcs_title_game_would_the_auburn_tigers_have_won_the.html

[tto827]

Thanks VI.

Cool but awful numbers. Using NFL based stats is useless for time comparisons since the clock stopping after first downs makes 1:20 at least 1:50 (assuming they get 3 first downs.. usually requires at least that much to march down the field)

Also.. 35% for Auburn to punch it in from the 1 on any given down...

At 50% he says the numbers are about even and with his lack of accounting for "pace" differences (he says he accounted for them, but I'm willing to bet nowhere near enough) and its a toss up... meaning if he actually worked it out more accurately, he'd probably find what the other's found... couple percentage points better off taking the knee.

Again to the people who will cry for no reason, taking the knee was impractical here, but statistically the correct decision.

He also states that he ignores the way's that Auburn doesn't score.. well what fukkin good does that do? Along with absolutely 0 acknowledgement of the value of burning the TO.

He skips thru the part where all the real work is.. and much like Glisan just says "my numbers show ..........."

[yisman]

Mathematically, Auburn's chances of winning were clearly higher had they been able to have first and goal at the 1.

How much higher you can debate, but definitely higher than scoring and giving FSU 1:19 and two timeouts to answer.

Keep in mind FSU's team total before the game was 40, because its offense was viewed as having the clear edge over Auburn's defense. That's more relevant to me than what had occurred in the game until that point.

Three cracks at scoring from the 1, and a field goal if you're somehow stopped 3 times. That just is better than scoring right away.

[sweep]

gotta love Wullfy thread bombin' for betpoints


Drinker- gimme

[Wulfman14]

gotta love Wullfy thread bombin' for betpoints


Drinker- gimme

BDF after a horrible slate of bowl games and undisciplined gambling. cash points everything wiped out. BDF, now I know how you feel sweep

[wrongturn]

Don't agree it is statistically better to knee on 1 yard. The analysis counts the chance of Auburn punching in on 1st and 2nd attempts while concluding the overall winning chance is actually getting better after each failed attempt. This does not make sense. The analysis should only calculate the chance of punch in on 3rd attempt, because 1st and 2nd attempts should be simple knee to burn clock, to be fair to FSU's intelligence. Don't think it is better choice at all.