re: To defer or not to defer? LSU Coin Toss Results (Kelly Era)

quote:

---

and not a single person has been able to provide any data to show that it was a poor decision.

---

Let me start off by saying that there are probably several people with more formal training in math and statistics that I could have who could explain this better.

In short though, I think you could do (at least) two types of analysis to answer that question - an historical (based on actually results) and a predictive probability.

Looking at the historical data, I think you’d need to add certain criteria to remove certain variables and get an accurate measure of the effect of taking then ball vs deferring, and not simply pick “all
of LSU’s games”.’ You have to factor for the relative strength of the teams (probably only selecting games where then spread was within so many points), as well as halftime and final scores.


If you’re going with a more probability based predictive approach, you could use things statistics like the average number of possessions or averages on time and length of drives.

quote:

---

Let me start off by saying that there are probably several people with more formal training in math and statistics that I could have who could explain this better.

---

which is why I've posted links to analytical studies on the coin toss.
unfortunately, all of them are from the NFL and I wish they were college based. And one of them was limited to only playoff games from 2002-2019 to narrow it down to as close to equal teams as possible.

even those studies show that as far as possessions and points, it comes out to an average of less than 1 point per game difference.
and winning percentages show an extremely negligible difference of .007 in favor of receiving rather than deferring over that 15 year period.

that's not my research, that's from actual stat and analytic nerds.

although, using just the LSU games since Kelly has been here, the percentages are pretty much in line with what the NFL studies show.

NFL the winning % is .007 in favor of receiving.
LSU the winning % is .051 difference also in favor of receiving.

NFL the percentage of getting back to back scoring possessions is around 12%.
LSU the percentage is around 15%.


there is no statistical evidence to say that either one is significantly different than the other as a blanket answer.

so there is no "poor decision", just personal preference.

Anytime the other team wins the toss they defer so that should tell BK what he should do

quote:

---

Anytime the other team wins the toss they defer so that should tell BK what he should do

---

Except for the 5 times (~30%) where this didn’t happen, of course.

quote:

---

You disagree with Bill Belichick on game strategy?

---

That depends, what’s his record as a head coach without Tom Brady?

quote:

---

Looking at the historical data, I think you’d need to add certain criteria to remove certain variables and get an accurate measure of the effect of taking then ball vs deferring, and not simply pick “all of LSU’s games”.’ You have to factor for the relative strength of the teams (probably only selecting games where then spread was within so many points), as well as halftime and final scores.

---

It’s virtually impossible for college football because of the lack of parity. In the NFL, where there is parity, there’s no statistically significant difference between taking the ball and deferring. Meaning that the results fell within the range that would be expected, given the sample size, if the coin toss decision had no impact at all.

If I flip a coin 1,000 times and yell “Hallelujah” every time, and 510 of those flips turn up heads, I could try to argue that my ritual gave me a 2% edge (51% to 49%). But the reality is that there’s a 27.4% chance that I get heads at least 510 times in 1,000 coin flips. My edge didn’t come from yelling “Hallelujah,” it was just random luck.

One thing I do find a bit interesting - in the Saban comments that have been floating around these threads, Saban says that the wind is one of the biggest factors in his decision to receive or defer. If there’s any legitimate wind, he wants it at his back in the 4th quarter. So he likely has to receive so he can pick a side in the 3rd quarter.

I find it interesting because domes are much more common in the NFL than in college, meaning wind should be much more of a factor in the college game. If you buy into Saban’s logic, that would mean electing to receive is actually a stronger decision in college than in the NFL. And as we’ve established, there’s already no advantage either way at the pro level.

At the end of the day it’s a gut decision every coach has to make that has very little, if any, effect on outcomes.

quote:

---

which is why I've posted links to analytical studies on the coin toss. unfortunately, all of them are from the NFL and I wish they were college based. And one of them was limited to only playoff games from 2002-2019 to narrow it down to as close to equal teams as possible. even those studies show that as far as possessions and points, it comes out to an average of less than 1 point per game difference. and winning percentages show an extremely negligible difference of .007 in favor of receiving rather than deferring over that 15 year period. that's not my research, that's from actual stat and analytic nerds. although, using just the LSU games since Kelly has been here, the percentages are pretty much in line with what the NFL studies show. NFL the winning % is .007 in favor of receiving. LSU the winning % is .051 difference also in favor of receiving. NFL the percentage of getting back to back scoring possessions is around 12%. LSU the percentage is around 15%. there is no statistical evidence to say that either one is significantly different than the other as a blanket answer. so there is no "poor decision", just personal preference.

---

The more I think on this question, the more I realize that trying to use historical data to prove or disprove the hypothesis regarding deferring the option being the strategy that gives you the best chance of winning isn’t the proper way to approach the question.

When it boils down to it, I think there are certain fundamental issues that go into the question - including what the actual question should be.

1- the majority of scoring in the game is done when possessing the ball. Therefore, having more possessions > fewer possessions, is generally true but not absolute

2- the “value” of each possession in the game is fluid, and based on a number of factors.

3- the game is timed, and the amount of time is finite

4- the amount of time remaining in both the half and the total game affects the relative value of each possession

5- teams have more influence over the time remaining in the game when possessing the ball.

6- whether you are leading or trailing, and the score differential, also effect the relative value of each possession.

7- the value of a possession starts off as the same for both teams, but changes throughout the course of the game based on score, time remaining, and who is winning or losing.


8- the relative value/need for possession affect the game play


With all of this said, the question “to defer or not defer” must first be answered with the assumption that “both teams are equal and play perfect”, so based on the game theory what is best.

Then the answer to that question should be weighed in terms of the fact that teams aren’t equal (in ability or style), and may not play perfect or to their capability.

quote:

---

The more I think on this question, the more I realize that trying to use historical data to prove or disprove the hypothesis regarding deferring the option being the strategy that gives you the best chance of winning isn’t the proper way to approach the question.

---

well until you can find a way to accurately predict the future, you have no choice but to view historical data and apply it going forward.

and historically, there is no significant difference that separates one from the other.

quote:

---

2- the “value” of each possession in the game is fluid, and based on a number of factors.

3- the game is timed, and the amount of time is finite

4- the amount of time remaining in both the half and the total game affects the relative value of each possession

5- teams have more influence over the time remaining in the game when possessing the ball.

6- whether you are leading or trailing, and the score differential, also effect the relative value of each possession.

7- the value of a possession starts off as the same for both teams, but changes throughout the course of the game based on score, time remaining, and who is winning or losing.


8- the relative value/need for possession affect the game play

---

and literally ever one of these values change on a game by game basis.
which is why there is no evidence that one outweighs the other as a blanket answer.

every game is different.
different matchup, health of your team, health of their team, paying site, weather conditions, etc.

and that's why Nick Saban has been quoted more than once saying their decision is made the morning of the game.
based on every factor you can think of and probably even more that he's also thought of but none of us have.

quote:

---

How does the other part of the coin toss - field direction - affect the outcomes? Do those factors change indoors vs. outdoors? Home vs. away? College vs. pro?

---

Field direction is one of the things Saban mentions when talking about it.

if the wind is above a certain MPH, he wants the wind at his back in the 4th quarter. That's more important to him in a close game than which order you get the ball.
And he gave examples of games that he either won or lost that were affected by the 4th quarter wind.

that's hardcore research