Bracketology 2016-17 Style

[Beeeej]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

[Tom Lento]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

tl,dr; I'm inclined to agree that 85% is too high given what I've been reading about the team and seeing in box scores/metrics (sadly I don't get to watch games anymore), but based on record to date it is about what I'd expect out of KRACH or Elo or whatever reasonable predictive ranking model one might use.

I couldn't find their methodology but most likely it's either a Monte Carlo simulation based on win probabilities or a simpler probabilistic combination based on those same probabilities. I would not be surprised if their win probability for Cornell vs RPI substantially over-weighted Cornell's chances, but without a pointer to how they figured that I have no way of knowing. A common approach for College Hockey geeks is to use KRACH ratings, which (if I'm doing it right) currently put Cornell at roughly 90% to beat RPI, and a slight favorite to beat Union. I found a site with Elo ratings, which is also common for such sports predictions, and it put Cornell at an 80% favorite over RPI and a slight underdog to Union.

Worth noting, from a PWR perspective losing to RPI and beating Union might very well be better for Cornell than beating RPI and losing to Union. It all depends on whether or not RPI would count as a "bad win" at the end of the season. I don't think they would given the importance of opponents opponents' record for RPI (the ranking), but their record is pretty bad so I guess we'll see. Union would almost certainly provide a quality win bonus. The real problem with a loss against RPI is it makes a 2-3-0 stretch run reasonably likely even with a QF sweep, which I suspect means no NCAA, while a win makes anything below .500 down the stretch pretty unlikely even with a 3-game QF round.

[Trotsky]

I love it when Cornell wins. I don't give a rat's ass whether possible future opponents win or lose, until they play Cornell. Except for Harvard.

Uh huh.

[BearLover]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

[Dafatone]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

One thing to consider (and I say this as one of the more optimistic optimists here) is that we've overperformed a little in terms of our goal differential vs our record.  So we might not "feel" as good as our record, and therefore based on that feeling the odds of an RPI loss are greater than simply a function of our record and theirs.  Plus, RPI played us kinda-close-ish last we played them, if I remember, and we might be thinking of them as better than their record based on what we saw.

All in all, I don't think a loss sinks us as much as one might think.  RPI is the sum of three components: win%, opp win%, and opp opp win%.  The first component doesn't care about RPI's record, and the second and third components don't care about whether or not we win.  So, as mentioned above, a loss to RPI and a win against Union could be better than the reverse (since there's a quality win boost I didn't describe above).  Not saying a loss to RPI wouldn't be bad, but I don't think, uh, RPI-the-metric cares that much about who you lose to, just who you play and your win%.

[Beeeej]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

I seriously don't know where you're getting that.

Basically, you're saying "85% is too high because it doesn't take randomness into account," and I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account." I said nothing about you thinking it was giving Cornell a 100% chance of beating RPI or anything remotely "absolute" like that.

Anyway, as all bad poker players do, after the Cornell-RPI game is over, some people will take the results as "proof" their predictive analysis was correct, and none of them will be right.

[Trotsky]

I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account."

Exactly.  This is the age-old Ron Darling argument that when statistical calculation and "the eye test" conflict it must mean the math is wrong.

Guess Yale had a "need" for a RHP...

Next week in this space: no, a fastball does not rise.  Physics > your perception.

[KGR11]

BearLover, from the sounds of it, you're on board with the process used (ie, Monte Carlo), you're just not on board with them using KRACH, right?

It "feels" tough to say we have a 90% chance of beating RPI, given that we were only leading for 35 seconds in that game, while they lead for over half the game. That said, I think that people often think that 90% is a sure thing. It really isn't. It's about the odds of rolling a 5 with two dice: Not likely, but not out of the ordinary. Rolling a 12 on the other hand, now THAT feels special (only happens 3% of the time)

[BearLover]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

I seriously don't know where you're getting that.

Basically, you're saying "85% is too high because it doesn't take randomness into account," and I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account." I said nothing about you thinking it was giving Cornell a 100% chance of beating RPI or anything remotely "absolute" like that.

Anyway, as all bad poker players do, after the Cornell-RPI game is over, some people will take the results as "proof" their predictive analysis was correct, and none of them will be right.

Dude, literally all I'm saying is that an 85% chance of making the tournament presupposes an incredibly high chance of beating RPI.  I don't know how it came to that probability, because I do not know precisely how sophisticated the model is.  But I bet that if it looked at stats like, say, goal differential, shot differential, etc., rather than just at how often the 9th-best team beats the 57th-best team, the model would not be giving Cornell a 90% chance of winning.  Maybe it does look at that stuff and I'm wrong.  But I doubt it.  

KGR11, a 90% chance of beating RPI is extremely high--is that indeed the chance the model us?  

I read fivethirtyeight and similar blogs religiously.  I don't need to be lectured on probability/results-oriented thinking.  At the same time, I'm genuinely in the dark about this predictor and could definitely be wrong in this particular case.

[Dafatone]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

I seriously don't know where you're getting that.

Basically, you're saying "85% is too high because it doesn't take randomness into account," and I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account." I said nothing about you thinking it was giving Cornell a 100% chance of beating RPI or anything remotely "absolute" like that.

Anyway, as all bad poker players do, after the Cornell-RPI game is over, some people will take the results as "proof" their predictive analysis was correct, and none of them will be right.

Dude, literally all I'm saying is that an 85% chance of making the tournament presupposes an incredibly high chance of beating RPI.  I don't know how it came to that probability, because I do not know precisely how sophisticated the model is.  But I bet that if it looked at stats like, say, goal differential, shot differential, etc., rather than just at how often the 9th-best team beats the 57th-best team, the model would not be giving Cornell a 90% chance of winning.  Maybe it does look at that stuff and I'm wrong.  But I doubt it.  

KGR11, a 90% chance of beating RPI is extremely high--is that indeed the chance the model us?  

I read fivethirtyeight and similar blogs religiously.  I don't need to be lectured on probabilities/results-oriented thinking.  At the same time, I'm genuinely in the dark about this predictor and could definitely be wrong in this particular case.

There's also a pretty good chance we still get in despite a loss in that game, so that adds to the math.

[Beeeej]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

I seriously don't know where you're getting that.

Basically, you're saying "85% is too high because it doesn't take randomness into account," and I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account." I said nothing about you thinking it was giving Cornell a 100% chance of beating RPI or anything remotely "absolute" like that.

Anyway, as all bad poker players do, after the Cornell-RPI game is over, some people will take the results as "proof" their predictive analysis was correct, and none of them will be right.

Dude, literally all I'm saying is that an 85% chance of making the tournament presupposes an incredibly high chance of beating RPI.  I don't know how it came to that probability, because I do not know precisely how sophisticated the model is.  But I bet that if it looked at stats like, say, goal differential, shot differential, etc., rather than just at how often the 9th-best team beats the 57th-best team, the model would not be giving Cornell a 90% chance of winning.  Maybe it does look at that stuff and I'm wrong.  But I doubt it.  

KGR11, a 90% chance of beating RPI is extremely high--is that indeed the chance the model us?  

I read fivethirtyeight and similar blogs religiously.  I don't need to be lectured on probability/results-oriented thinking.  At the same time, I'm genuinely in the dark about this predictor and could definitely be wrong in this particular case.

You interpreted what I said as suggesting that you think the model gives Cornell a 100% chance of beating RPI, then you addressed your arguments to that position instead of my actual one. So I clarified my actual position with smaller words. I'm not lecturing you on your understanding of probability, I'm lecturing you on your reading comprehension.

[Jeff Hopkins '82]

Finally, according to this, our probabilities for tourney advance are currently:

.85 Tourney
.42 QF
.19 SF
.09 F
.04 That Which Shall Be Nameless

Is there a way to check what a loss to RPI would do to us in the PWR?  Because unless such a loss would knock us down just a few spots, that 85% number feels way too high.

Without taking into account anything else that might happen on Friday night, a loss drops Cornell to 15th.

But keep in mind that the 85% number is going to inherently account for (a) what are the odds that Cornell beats RPI; (b) what are the odds that Cornell makes up for any potential loss to RPI by winning in the quarterfinals; (c) what are the odds Cornell can beat Union

Exactly. 85% isn't too high because the fact that RPI is 57th in the Pairwise, and therefore "likely" to lose to us, is already "baked in." The fact that we'd drop precipitously if we lost is also baked in, since it's supposed to be "unlikely."

I understand that, but I do not think the predictor accounts for how much randomness exists in a single game of hockey.  If we lose to RPI, we are probably less than 50% to make the NCAAs.  We've lost/nearly lost to many bad teams this year.  Any prediction predicated on what seems to an extremely high likelihood of beating a team we had to score twice in the final seven minutes to beat a few weeks ago seems faulty to me.

I don't think you understand how this works, then. If the predictor didn't take into account that there's randomness in a single game of hockey, and weaker teams can beat stronger ones, our chance of making the tournament would be much higher than 85%.

It really is baked in.

I know it is taking innate hockey randomness into account.  I'm saying it isn't adequately taking innate hockey randomness into account, especially w/r/t this particular Cornell team.  More specifically, I'm saying the predictor is giving Cornell too high a chance of beating RPI, not that it's giving Cornell a 100% chance of beating RPI.  Which is why I used language like "how much randomness" and "extremely high likelihood" rather the absolute wording you read into my post.

I seriously don't know where you're getting that.

Basically, you're saying "85% is too high because it doesn't take randomness into account," and I'm saying, "85% isn't too high because it does take randomness into account, this just happens to be where the numbers come out when you take randomness into account." I said nothing about you thinking it was giving Cornell a 100% chance of beating RPI or anything remotely "absolute" like that.

Anyway, as all bad poker players do, after the Cornell-RPI game is over, some people will take the results as "proof" their predictive analysis was correct, and none of them will be right.

Dude, literally all I'm saying is that an 85% chance of making the tournament presupposes an incredibly high chance of beating RPI.  I don't know how it came to that probability, because I do not know precisely how sophisticated the model is.  But I bet that if it looked at stats like, say, goal differential, shot differential, etc., rather than just at how often the 9th-best team beats the 57th-best team, the model would not be giving Cornell a 90% chance of winning.  Maybe it does look at that stuff and I'm wrong.  But I doubt it.  

KGR11, a 90% chance of beating RPI is extremely high--is that indeed the chance the model us?  

I read fivethirtyeight and similar blogs religiously.  I don't need to be lectured on probability/results-oriented thinking.  At the same time, I'm genuinely in the dark about this predictor and could definitely be wrong in this particular case.

You interpreted what I said as suggesting that you think the model gives Cornell a 100% chance of beating RPI, then you addressed your arguments to that position instead of my actual one. So I clarified my actual position with smaller words. I'm not lecturing you on your understanding of probability, I'm lecturing you on your reading comprehension.

Guys, there's math and there's emotion.  Based on the methodology, the math is what it is.  But our emotion says the method is weak.  This team doesn't blow people out and  has it's moments of let down that allow teams that we should beat to stay in games or even win.  So while the methodology does take into account some degree of probability, it just "feels" optimistic.

OTOH...DROP THE PUCK!

[KenP]

I understand both points.  KRACH handles inter-game mathematics perfectly, but treats all wins equally, whether it is eeked out or a 10-0 blowout.

What about team / intra-game statistics?  A review of both standard and advanced stats at http://www.collegehockeynews.com/stats/ indicates we are not leaders in any major categories.  (Except for ShA: Shots On Goal Against... but that stat favors teams that play fewer games.)  So to BearLover's point, how to manage expectations?  Is this team a true NCAA top-10 team?

Like all of us I'm really enjoying this season and it's exciting to still be in the thick of things.  Regardless of final outcomes, thank you Team and Coach!

[upprdeck]

you can use all the math you want, but dont the stats show that over time almost all these super cool tools predict the winners just about the same as a pet rooster?

[Jeff Hopkins '82]

you can use all the math you want, but dont the stats show that over time almost all these super cool tools predict the winners just about the same as a pet rooster?

There's the problem.  We need the Octopus that picks the World Cup winners!  ::dribble::