Re: Minnesota Gophers Season Thread: 2014-2015
Weird. A couple of weeks ago Harley said, in giant bold letters, Minnesota no longer controls it's own destiny. He scoffed at my suggestion that if they could get to 22-23 wins they should be in pretty good shape, since historically teams with that number of wins find a way to an at-large bid.
http://collegehockeyranked.com/2015/02/04/tournament-cutlines-revisited/
I see you still don't get it. Yeah weird.
I'm not going to spend anytime explaining the PWR system to you as you're obviously too lazy and stubborn to try to understand it for yourself. But I will comment on Dahl's statistical methodology and the limitations of his analysis. I don't like doing this on USCHO, as it bores everyone to tears, but since you're such a smart*** I'll hit you over the head with it.
I will begin by stating probability models do not have robust predictability unless they are based upon valid quantitative research designs. Therefore, the absence of key predictor variables at this stage makes Dahl's model dreamy at best.
Dahl is using maximum likelihood estimations to calculate an outcome variable in a "linear" PWR probability model. The word "linear" is important because it essentially encapsulates the statistical limitations of his analysis. For example, his probability distribution model by default is not based upon a 95% CI, his parameters are not robust enough to assure homogeneity of variance, thus inferences are jeopardized; and that's very significant. Secondly, Dahl is not considering some vitally important factorial levels in his probability distributions, and that's also very significant because it lowers internal validity and reliability which significantly increases the probability of a Type 1 error (for example: teams have
wildly different numbers of games remaining in their regular season, from 6 to 12, so their potentials to make big moves will also differ accordingly). Thirdly, the overlap of the distributions (shaded area) must be quantified in preferably Cohen's
d to determine effect size, statistical power and causal significance. Since the distribution is two tailed, if I choose to accept Dahl's analysis based upon face validity alone, the Gophers have an equal probability of not making the tournament. Lastly,
IF I wanted to spend the time performing a multiple regression procedure to determine statistical probability of the Gophers making the tournament I would base it upon several independent variables, slope and a beta coefficient intercept of
THIS (click on UMN) to determine R squared or the portion of variance in the dependent variable explained by the predictors. That computation would make Dahl's model irrelevant.
Even though Dahl's model is fraught with many limitations, he still agrees with my analysis
:
Who controls their own destiny?
Teams that should make it if they continue to do we’ll are from #7 Michigan Tech down to about #18 Merrimack, which approaches the bubble with a bit over .500 in its remaining games
I'm not trying to flame Jim Dahl here or his work, and in all fairness the statistical methodology of his model might make a bit more sense in several weeks. But right now....no.
This still makes me chuckle
:
As for my predictions, nowhere have I predicted MN will go 16-0, 12-4 or even make the tournament.
You stated, "As for my predictions, nowhere have I predicted MN will go 16-0, 12-4 or even make the tournament".
Please explain to us your recent post:
"Fine, ignore the playoffs. MN has 16 regular season games left. I guarantee you they go 16-0, and finish the regular season 27-7-2, they are on the right side of the pwr bubble. But of course, we all know they don't even have to do that. But I'll throw a prediction out there. They go 12-4, or maybe 11-3-2 through the end of the regular season, assuming the losses don't come against Wisconsin, or all against OSU, or something like that, they are on the inside looking out when the conference tournament starts."
Yep, that's my prediction. If Minnesota goes 16-0, or even 12-4 or 11-3-2, they make the tournament. Please keep up.
So you make a prediction "assuming" they don't lose to...blah, blah, blah?
Then you deny making the prediction you made only several minutes before. Then you contradict yourself and write some IF ****, when you "guaranteed" it? Then you tell me to "keep up"? Keep up with what you f'kn looney tune!?
Your mental masturbation?
You seem like a old and confused Whioux fan with a now apparent obsession for the Gophers.
If you throw out enough (crap) eventually something will stick.