Started By
Message
VandyFans: Will you please explain "regression toward the mean" to USC fans?
Posted on 8/29/14 at 8:40 am
Posted on 8/29/14 at 8:40 am
Thanks in advance.
Posted on 8/29/14 at 8:42 am to Tennessee Jed
shite talking Vandy. That's the Tennessee we know and love.
Posted on 8/29/14 at 8:51 am to Patton
Actually, it's directed at two teams. Maybe, I should have tried something a little less sophisticated.
Posted on 8/29/14 at 8:56 am to Patton
quote:
shite talking Vandy. That's the Tennessee we know and love.
The lack of comprehension implies you didn't go to Vandy.
Posted on 8/29/14 at 8:56 am to Patton
Yeah, this is pretty good. Thumbs up from me.
Posted on 8/29/14 at 9:08 am to Crowknowsbest
quote:
Underrated thread
Should have probably started it in the afternoon after all the commotion has died down.
Posted on 8/29/14 at 9:10 am to Tennessee Jed
I upvoted a UT fan. This must be a good thread. 2 birds, one stone. Nicely done.
Posted on 8/29/14 at 10:16 am to TreyAnastasio
quote:
Should have probably started it in the afternoon after all the commotion has died down.
Just have to keep it alive until people get a chance to appreciate it
Posted on 8/29/14 at 10:29 am to Tennessee Jed
I was really hoping that you would hook a Vandy fan to start in with the definition and a few sentences in go, "Hey, wait a minute!!!"
Posted on 8/29/14 at 11:19 am to Tennessee Jed
I've always felt that this phrase was misleading. Things don't truly tend to "regress", they simply are not likely to be an outlier the second time. Nothing has changed, and there is no causal relationship between scoring an extreme measurement and the likelihood of scoring near the mean on the second measure. The fact that you are more likely to score toward the mean remains constant, regardless of your first measurement.
The roulette wheel has no memory. If things truly regressed toward the mean, then a smart bet would be to wait until the ball hit a green (0 or double 0) and then bet heavier on red and black. But in reality, the wheel has no memory - the odds of hitting green remain identical to what those odds were before the last spin, regardless of the outcome.
The roulette wheel has no memory. If things truly regressed toward the mean, then a smart bet would be to wait until the ball hit a green (0 or double 0) and then bet heavier on red and black. But in reality, the wheel has no memory - the odds of hitting green remain identical to what those odds were before the last spin, regardless of the outcome.
Posted on 8/29/14 at 11:22 am to Crimson Legend
quote:
I've always felt that this phrase was misleading. Things don't truly tend to "regress", they simply are not likely to be an outlier the second time. Nothing has changed, and there is no causal relationship between scoring an extreme measurement and the likelihood of scoring near the mean on the second measure. The fact that you are more likely to score toward the mean remains constant, regardless of your first measurement.
The roulette wheel has no memory. If things truly regressed toward the mean, then a smart bet would be to wait until the ball hit a green (0 or double 0) and then bet heavier on red and black. But in reality, the wheel has no memory - the odds of hitting green remain identical to what those odds were before the last spin, regardless of the outcome.
You're assuming that football operates on simple probability like blackjack or roulette.
It doesn't work like that.
This post was edited on 8/29/14 at 11:23 am
Posted on 8/29/14 at 11:23 am to Tennessee Jed
Is it your opinion that Tennessee has a new normal?
Your long-run moving average is most definitely sloping downward.
A decade of mediocrity is statistically significant, I'm pretty sure.
Your long-run moving average is most definitely sloping downward.
A decade of mediocrity is statistically significant, I'm pretty sure.
Posted on 8/29/14 at 11:25 am to Crimson Legend
quote:
The roulette wheel has no memory. If things truly regressed toward the mean, then a smart bet would be to wait until the ball hit a green (0 or double 0) and then bet heavier on red and black. But in reality, the wheel has no memory - the odds of hitting green remain identical to what those odds were before the last spin, regardless of the outcome.
Except college football isn't a roulette wheel. Programs are different, thus they have different means.
Posted on 8/29/14 at 11:29 am to Volatile
quote:
You're assuming that football operates on simple probability like blackjack or roulette.
It doesn't work like that.
Not assuming that at all, but you do make a great point. There is nothing simple about the probability of a football outcome. IF we could identify all the factors that impact the outcome and create an identical set of circumstances, then the probability of the second result would be identical to the probability of the first result.
But this is the nature of the problem with determining the relative strength of a team in football. It is highly, highly complex, and most of us are guilty of oversimplifying it and giving far too much weight to one or two factors. "This team can't win on the road" or "Of course XYZ State was going to lose, their coach always loses big games" are usually ridiculously far from accurate.
And then, amazingly, we all are shocked when the team everyone thought would win it all loses in an upset. It happens over and over and over...
Back to top
Follow SECRant for SEC Football News