- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: To defer or not to defer? LSU Coin Toss Results (Kelly Era)
Posted on 10/16/24 at 7:46 pm to CubsFanBudMan
Posted on 10/16/24 at 7:46 pm to CubsFanBudMan
quote:
Would you rather your defense rest for 7 minutes prior to playing a single down, or rest after at best causing a 3 and out, at worst giving up a 7 minute drive to the other team? Why do you need to rest before playing a down. In contrast, if you receive the 2nd half kickoff, you possible extend your defense's rest from a 20 minute half time break to 20 minutes plus the 7 minute drive.
I really don't know and that is why I'm asking.
If all things are constant, I really think it is all a wash.
However there are not constants.
0
Posted on 10/16/24 at 7:53 pm to CBandits82
quote:
This is the one that Kelly needs to just adapt to. Just defer......every time.
Nick saban disagrees
Posted on 10/16/24 at 8:49 pm to Nutriaitch
quote:
Again, receive has a higher winning % than deferring.
It’s not statistically significant either way.
If you look at NFL data you’ll find the same thing. No statistically significant difference in win% choosing to receive or defer.
The reality is that coin toss decisions simply don’t matter much, if at all.
Posted on 10/16/24 at 8:57 pm to LSUStar
quote:
My thought was that we would need many more, thousands or at least hundreds, data points. One way to narrow it which of course limits data would be to look at all top 5 matchups the last 20ish years. I am guessing that the outcome would be random as we understand that to be. If a strong correlation e.g. >.7 were found that would be interesting.
Yes, you would need a shitload of data points to actually see the edge one way or another. Particularly in college football where you have pretty dramatic disparities (relative to the NFL for example) between teams in a given game. It makes the data noisy as hell.
One way to get around this would be to look at performance ATS instead of straight up W/L, but even that would be noisy and would take a lot of data.
Posted on 10/16/24 at 9:44 pm to lostinbr
quote:
The reality is that coin toss decisions simply don’t matter much, if at all.
This is where I am leaning at this point
Posted on 10/16/24 at 9:55 pm to TFS4E
Strong work. I’ve always wondered about this but never felt like looking into it.
Posted on 10/16/24 at 9:59 pm to CBandits82
Absolutely, just defer, Kelly is stubbornly dumb on this tactic. My gosh
Posted on 10/16/24 at 10:39 pm to eugene1928LSU
quote:
Absolutely, just defer, Kelly is stubbornly dumb on this tactic. My gosh
So, in other words, you just completely ignored all the evidence in this thread that shows that it’s pretty much a wash. Who’s the one being stubbornly dumb here?
Posted on 10/16/24 at 10:52 pm to TFS4E
Lot of food for thought but wouldn't you want to account for the level of competition? For example, would it really matter what we did vs say Grambling?
Posted on 10/16/24 at 10:57 pm to TFS4E
Is this called Analytics ? It’s a coin toss ! Isn’t that 50/50 ?
Posted on 10/17/24 at 12:31 am to TFS4E
The question posed: What is the expected value of the first drive of each half for LSU vs for its opponents?
The advantage of receiving to start the game is getting the first chance to score. The disadvantage is that your opponent gets the first chance to score in the second half. Comparing the expected value of the first drive of each half should show whether it’s wise to receive or defer.
(Note that the results at the end of the first half don’t matter here—regardless of whether you receive the first half or second half kickoff, you want to be the last one to score in the first half.
Also, note that who decides to receive/defer doesn’t matter, either. All we care about is the EV of receiving first vs receiving second. That doesn’t change based on who decided the order.)
From your data (and treating all TDs as 7, FGs as 3, and other results as 0), LSU had a point differential of -6 over 22 games when it received the first half kickoff and a differential of -4 over 11 games when it received the second half kickoff.
This actually suggests LSU is better off receiving the ball. LSU’s point differential on the opening drives of each half is -0.27 when it receives the first half kickoff and -0.36 when it receives the second half kickoff.
The bigger issue to me is that the point differential is negative either way
The advantage of receiving to start the game is getting the first chance to score. The disadvantage is that your opponent gets the first chance to score in the second half. Comparing the expected value of the first drive of each half should show whether it’s wise to receive or defer.
(Note that the results at the end of the first half don’t matter here—regardless of whether you receive the first half or second half kickoff, you want to be the last one to score in the first half.
Also, note that who decides to receive/defer doesn’t matter, either. All we care about is the EV of receiving first vs receiving second. That doesn’t change based on who decided the order.)
From your data (and treating all TDs as 7, FGs as 3, and other results as 0), LSU had a point differential of -6 over 22 games when it received the first half kickoff and a differential of -4 over 11 games when it received the second half kickoff.
This actually suggests LSU is better off receiving the ball. LSU’s point differential on the opening drives of each half is -0.27 when it receives the first half kickoff and -0.36 when it receives the second half kickoff.
The bigger issue to me is that the point differential is negative either way
This post was edited on 10/17/24 at 1:07 am
Posted on 10/17/24 at 12:59 am to Portcityblues
quote:
judging from that 15-2 record when losing the toss, maybe we just need to try and lose the toss every game?
When an a pattern or anomaly occurs, particularly in a data series with only two varibles, we cannot just infer that there is a cause and effect relationship. Either of the two variables can nave exogenous influences not recognized in the underlying assumptions of the statements being tested in the study. The influences on either of the two variables may be different, and may affect. each of the two variables (winning the toss, winning the game) in the same or opposing ways, or not at all. Exogenous factors such as weather, strength of competition, random events like injuries and penalties may affect the two varibles in the study differently or not at all.
The place to start, of course, by making a clear statement of the hypothesis underlying the study.
Posted on 10/17/24 at 3:31 am to TFS4E
The only time I would ever not defer would be a weather related situation such as a forecast for worsening weather or wanting to go in a certain direction due to the wind i.e having the wind at your back in the 4th quarter.
Posted on 10/17/24 at 5:07 am to TFS4E
Good analysis, but it may be missing a key factor, the relative strength of the team’s offense and defense.
My impression is that Kelly wants to put his best 11 guys on the field first to seize momentum. In the last few years, that has been the offense. That selection may change as the defense continues to improve.
My impression is that Kelly wants to put his best 11 guys on the field first to seize momentum. In the last few years, that has been the offense. That selection may change as the defense continues to improve.
Posted on 10/17/24 at 7:44 am to TFS4E
quote:
LSU is 15-2 (.882) when we lose the toss
Maybe we should just defer participating in the coin toss... based on the data.
Posted on 10/17/24 at 8:29 am to lostinbr
quote:
Particularly in college football where you have pretty dramatic disparities (relative to the NFL for example) between teams in a given game. It makes the data noisy as hell.
that's why when I put up LSU's winning %, I removed non-Power 5 games.
but even then, there is usually a big difference between a top level Power 5 and a lower level Power 5.
Which makes it tough to decipher.
You'd also have to remove games affected by weather.
If a game is played in an absolute monsoon, that (at least theoretically) has a much larger effect on how a game is payed than who got ball first.
quote:
ATS instead of straight up W/L, but even that would be noisy and would take a lot of data.
but you'd also have to watch (at least the endings of) games and use your judgement.
Does a team "call off the dogs" and win, but not cover?
Does a team take a knee on last drive in striking distance that keeps them from covering?
Does some weird fluke play type thing happen that swings the spread (a bad beat)?
Does a key player, like a QB, get injured that changes the swing of a game?
Does a normally reliable kicker just shank a make-able kick (not affected by weather)?
there are so many variables that go into who wins, loses, covers, etc. that the very slight possibility of getting back-to-back possessions is way down the list.
Which is why the data shows there is not a significant difference in winning/ losing based off what half you get the ball first.
Posted on 10/17/24 at 8:32 am to TFS4E
Nice number crunch and spreadsheet
Posted on 10/17/24 at 8:38 am to OKBoomerSooner
quote:
Comparing the expected value of the first drive of each half should show whether it’s wise to receive or defer.
the decision is based off what you believe gives your team the best results for the entire game, not just any one particular drive.
Posted on 10/17/24 at 8:46 am to Nutriaitch
I would take a hard look at how well the opposing defense has been playing in the 1st quarter. If they've been taking a possession or two to settle in give me the ball. Otherwise I'd defer. The stats really don't support a clear advantage either way so I'd take it game to game.
This post was edited on 10/17/24 at 8:49 am
Posted on 10/17/24 at 9:03 am to Portcityblues
quote:
judging from that 15-2 record when losing the toss, maybe we just need to try and lose the toss every game?
Ref: Call it heads or tails
LSU: "Ice Cream!"
Ref: It's HEADS. LSU has lost the toss.
Page 3 of 6
Popular
Back to top
Follow TigerDroppings for LSU Football News