Stupid Selection Committee Tricks?

The possibility of a "good road win, bad home loss" bonus point system being tacked onto the current PWR selection process was broached on a USCHO thread, in the wake of the Committee chair's public statements.

Source: [board.uscho.com]

Now, this couldn't hurt Cornell as of today, and indeed could help if say the OSU and Harvard road wins were considered worthy of a bonus. Still, it's an awful idea unless precisely defined and published in advance of the selection (and preferably with a little more time than just a few weeks before selection).

USCHO is reporting on the stupid committee tricks:

[uscho.com]

Sounds like a poorly thought out idea to me. Adding a very crude modifier like "good road wins" and "bad home losses" is dumb. This criterion may be "objective" in the sense that the rules for applying it are cut and dried, but they're not telling anyone how it works so it might as well be back room subjectivity. Plus, eliminating the transparency of the process is a major step backwards, all in the name of "adding a little mystery". Don't they realize that hockey fans don't want mystery and excitement on selection day - they want a fair process that's clear and open.

As Greg says, this shouldn't hurt us as of the moment, unless other teams gain while we sit still. Still a very bad idea...

Awful idea. What's next, making margin of victory count too?

It's easy to see why this hurts teams in weaker conferences, but this also discourages any of the major teams from wanting to schedule CHA or MAAC teams. So much for the advancement of the game. This essentially deepens the emerging class system that is IMO the biggest threat to the college game. Auto-bids and expansion of the NCAA tournament were steps in the right direction, and now this happens. Why should teams like Maine and Michigan ever want to use an OOC game on a Quinnipiac or Wayne State if there's much more for them to lose than gain?

Is Colgate low enough for this to hurt us? Is Dartmouth?

big red apple wrote:

Is Colgate low enough for this to hurt us? Is Dartmouth?

Who knows?

I suppose the good news is we've finished the regular season with no home losses--to any opponent. Now we have to sweep the ECAC quarterfinals.

It appears that the article has been edited. [uscho.com]

While good wins (OOC wins over a top 15 RPI team) are rewarded with bonus RPI points, bad losses are not penalized with deductions. The bonus is based on the end of season (post-conference tournament) RPI.

As it currently stands, we would get bonus points for beating BU (x2), but OSU (16) and WMU (19) would not give us bonus points. Maine and OSU get credit for beating us.

OK...that's a little better. Still, in strong conferences where there are 3-5 top 15 teams, there are more opportunities for that "good win" bonus. One could argue that the "good win" bonus is to compensate for the worse record a bubble team would likely have by being in a strong conference. Why not instead use a system like KRACH which inherently accounts for the strength of opponents instead of this "good win" voodoo component? (Yes, count me as another convert)

Wasn't this "good win" logic pretty much how CU got into the NCAA Lax tournament last year, on the basis of our win vs. Syracuse?

Rich, being in a strong conference will make no difference, since only out of conference games are being counted in the "good wins" criterion.

Looks like the home losses against weak teams factor has already disappeared. Wonder what changes tomorrow will bring.

So far this new wrinkle looks like a way to keep Harvard out of the tournament as an at-large team, since they have no "good wins" this year. [And let's hope that doesn't change in Albany ]

I was about to comment on Al's post, but then I wondered - do conference tournament games count as non-conference games???

What about non-conference games (eg holiday tournament games) that happen to be against conference opponents?

Right, Jordan. It seems like they are providing a strong incentive to schedule the top teams for OOC matchups.

Now I will look into my crystal ball: If basketball is any indication, you will see plenty of Cornell v. North Dakota type matchups (sort of like annual Indiana - Kentucky games) but very little like Cornell - Niagara (Kentucky very rarely schedules EKU or Murray State). Since the big players will be willing to schedule us (when we are strong) the SOS advantages for RPI and the potential bonus points will probably outweigh any risk of losing to the big guns or altruism towards the MAAC and CHA. The bottoms of the leagues will end up scheduling eachother.

And the best laid plans of mice and men will backfire when the strength of the teams differ in the year in which the games are played from the year the games are arranged.

OOC isn't the same as NC. OOC is a game against a team from a different conference. NC is any game other than a conference RS game. (God knows what the HE-WCHA games in the 80's were -- interconference meetings that counted in conference standiongs ).

If conference tourny games can count, then the strong conferences do get a boost as there are lots of possible big wins to be handed out during their tourny. If they do not, then theoretically strong conferences get a tiny bit hurt, since the probabilty that a random NC opponent will be top 15 is lower than for a member of a weaker conference. Either way, the Ivy restriction below the NCAA limit is an even worse handicap than before.

I would look for the traditional powers to lock up cozy home-and-home series agreements. Since there's no loss associated with dumping a "big win," might as well max out the number of potential big wins on your schedule. There is no incentive for a top 15 team to schedule a second tier team, lower its strength of schedule, and risk handing out a big win with no chance of getting one themselves, when instead they can schedule another top 15 giant, get the s.o.s. bump, and have a 50-50 shot at picking up their own big win.

Icky idea.

You ask too many questions, Neil. If the committee were to read 'em, they'd change the rules again tomorrow.

Don't like this at all. It's unfair to teams that play fewer nonconference games, and it adds unnecessary fog to the selection process.

It sounded like McCaw liked being able to keep the public in the dark. Jerk.

Sure, it adds to the "mystery" if we don't know the exact procedure, but it also adds to the perception that they are a bunch of biased and corrupt wheeler-dealers.

This is just stupid - and since I don't have to convince anyone on here of this, I'll just get to the point - any idea where I can send off an email / letter to bitch? If a buncha people do it, it might get some notice, but the NCAA website isn't exactly feedback-friendly.

Hey, btw, just to point out yet another reason why this is a horrible idea...

What if Team A is 16th team which beats Team B, the 15th team, and in doing so earns enough RPI points to pass Team B, knocking them down to 16, taking away their point, and flipping it back, which restores the points... it's not mathematically consistent!

This is ridiculous. They already have a criteria for "Big Wins" in the PWR. Clearly, the committee has realized they can be replaced with John Whelan's script, and they decided that reflected on their manhood or something.

Clearly this hurts ECAC teams which play fewer nc games, Ivies especially. And if the goal was to influence nc game scheduling - why announce at the END of the season? Its so mind-boggling I don't know where to begin my tirade. (Well, I do, I finally went and got myself a login to this forum so I could comment.

Welcome, Larry. I enjoyed your tirade.

Btw, I managed to track down, using perfectly public means, the email addresses of all the selection committee members. If anyone is interested in sending a note, I'll be glad to share to save you the work.

Also, if anyone wants to see my note for ideas, I'll be glad to post it here or send it to you personally.

Don't hold back, post the addresses, and your note. Or you can just sent it to me, I definitely want to see it.

Why are the bonus points going to be awarded for beating a top 15 team in an ooc game? It seems like 15 is an arbitrary choice (it's not a power of 2), and that they might be able to pick a number to suit their needs and wishes.

gee what happened to the days when a win was a win? this shouldn't be so complicated to determine the teams that make it to the tournament. the more they try to make it objectively fair, the more they reveal the inherent biases in the whole ncaa organization. this is why i just watch the games and leave all this hullabaloo to the committees. I figure as long as they win, it'll work itself out.

I wouldn't post anyone's addresses w/o their permission - there's always spam reasons, as well as just personal privacy - the Denver and Air Force addresses were the hardest to find, having to wait through the *.ocsn.com pages, the others were available in a standard directory search. I'll gladly share if people want, but I don't wanna post them - IM or email me.

As for my letter... To: Mr. McCaw, CC'ed to everyone else:

-----------------------------------------------

Dear Mr. McCaw:

I wanted to let you know that I think it's pretty close to despicable the "big win" change that's are being put into place for this years NCAA Men's Division I Ice Hockey Selection Committee, and it's even worse the way it has been handled.

To not go public about this until a few weeks before the end of the regular season and conference tournaments makes it seem as if the committee feels it has something to hide. It was a vital piece of information that was withheld from public knowledge for no good reason. That encourages the opinion of selection committee as smoky-room bought-for-a-price hucksters.

Of course, perhaps the committee was hiding it because it honestly is something that needed to be hid. Perhaps the committee knows what an outrage this should cause and knows it has to keep it secret. In your interview you spoke as if putting 'mystery' into selection day is a positive effect... it is not!! College hockey fans want to know that when their teams are selected for the tournament, they're selected on the basis of merit, and moreover, they're selected on a basis of understood, thought-out, clearly defined merit. If the selection committee refuses to release the details of the process, then there is no way to know if the 15th team beat out the 14th for legitimate reasons, or because two of the selection committee members are friends of the AD of the 15th team.

There's simply no way to know that the process is honest and just unless all details are disclosed ahead of time, so that decisions on what counts how much can't be made *after* the results occur, and the results can be spun to someone's favor.

Additionally, I think this criteria will have unintended, negative side effects . These include encouraging the top team to only schedule other top teams. Since losing doesn't hurt you in this criteria, you may as well play as many top teams as you can schedule, making the already insular 'upper crust' that much more insular. In a time of conferences trying to move up, I think that's a horrible way to go. Additionally, it hurts teams that don't play as many games, since they have less chances to get that big NC win - Ivies particularly, as if they're not already at enough of a disadvantage, as well as, I believe, the military academies.

There is also one reason, perhaps above all, which shows why this process has not been fully thought through. That reason involves the fact that the mathematics simply may not work out. What if Team A is the 15th team in RPI and Team B is 16th. What if Team B beat Team A, such that it raises their RPI enough to pass the 15th team. But, now Team B no longer has a win against the top 15 teams with the new rankings, so they drop back down. This moves Team A back into 15th, so they now have a win against them, moving them back up... this cycle is infinite and will never end. How would the committee deal with this? Would they just go "Top 15" from the initial rankings? Would they decide the win shouldn't count? That it counts half as much? Unless we know all this in advance, there's no way to make sure the committee made a sound judgement and not one based on who they would be helping/hurting.

Finally, the committee fails to realize that there is already a category which gives you more credit for a win against bigger teams. It's called RPI! You get credit for a win, you get more credit if it's against a team with a higher RPI! Additionally, your record against top teams is covered in the TUC category. This is a *third* way in which wins against top teams matter - it's simply overkill. It's a redundant patch to fix something that's not really broken. And if you think RPI isn't doing it's job on it's own, then adjust or fix RPI, or replace it with something like KRACH or an altogether different system.

If the committee feels they need to keep details secret, to justify their existence, and to add 'mystery' to selection day, then they need to realize they are undermining the trust that the college hockey fan base and athletic teams hold in them, and destroying the 100% objective process that has made the college hockey selection process a model of how it should be done. The committee already has a major, irreplaceable purpose, and it's called seeding. No matter how much you know, or how the PWR comes out, I have yet to see people, including USCHO and other experts, exactly predict how the seeding will be chosen. The committee's place is in picking the best teams in an open and defined manner, and in seeding those teams in a skillful way.

I highly encourage you to take this all into consideration and use want means are available to you to rectify the situation, by doing no less than releasing the full numbers of this 'big win' plan, and by eliminating this horrible idea in the future.

Yours truly,

Fred Trinkoff
College Hockey Fan
Cornell University Undergrad

DeltaOne81 '03 wrote:

...but the NCAA website isn't exactly feedback-friendly.

Sh'yeah, like the NCAA gives a fuck what anyone thinks.

No need to take something stupid and make it seem even stupider.

(1) The bonus points are awarded based on the RPI at the end of the conference tournaments. If a win by Team A (16) over Team B (15) moves A up to 15, that is where they stay. End of story, no infinite loop.

(2) I think they chose 15 because with the CHA and MAAC (at least for the time being) only the top 14 teams in the RPI can conceivably make the NCAA tournament without needing to win its conference tournament. If 16 jumps 15, it doesn't matter.

(Does (2) sound like a post-hoc rationalization? Yeah, it does to me also. I don't really believe it. But it is true that having 16 jump 15 doesn't matter for tournament selection. "But Apple," you say (because we are close), "can't it effect seeding if #5 jumps over #4?" Yes, that is correct also. Hey I never said I thought that this was a good idea.)

(1) If that's the way it works, fine, but I was just pointing out that it seems like they'd hadn't thought it through.

(2) It may not matter for 15 and 16, but it might matter for others - i.e. if 15 and 16 switched right now we'd be in better shape, picking up an away win against OSU . And while that may only matter for seeding, it's perfectly concieveable that switching 15 and 16 could effect the 13 and 14 teams, and with one surprise tourney winner, that could decide who's in and who's out.

Re (2): No, it can't. Seriously. RPI is set from 1-15 at the end of the season. Bonus points are awarded based on the old RPI. They bonus points are added to the old RPI. Voila, new RPI. No multiple switching, no second order effects. One switch has no affect on other switches. It also means that it doesn't matter who is 15 or 16 now, only who is in that position at the end of the year.

Yes it would help us if OSU was #15 at the end of the year, but the bonus points will not have any influence on whether OSU gives us bonus points or not.

Given (1), (2) is moot. I understood that, but I shoulda made it more clear that I did. I'm ready to consider the infinite-flipping thing a closed issue.

But until they release the numbers, I'll never trust this category. There's no way to know that they're not doing it to put Providence in Providence to boost ticket says, and leave Denver out - now they can say "oh, yeah, it's the formula, just the way it worked out." Simply put, I don't trust them enough. I probably trust them a fair deal, I think it's pretty unlikely they'd do it, but I don't *know*. Give me transparency or give me, uh, well, just give me transparency damnit!

Even that aside, I still think it's a bad idea because it'll hurt (not helps, as McCaw said) the small schools. Rather than give the small schools bonuses for having to play on the road, it'll make it that much harder for them to get scheduled at all, if the big, power schools know they'd be much better off by trying to get a win versus another team that's likely to be high. It's also blatantly unfair to Ivies and Air Force (and Army?) who play less games. It's not like we're # of good wins / games played, or # of good wins - # of bad loses or something (though I hate bad loses too, but at least that wouldn't have this problem). Instead, it's simply a hard number, mediated by nothing. It'd be like ranking schools purely by # of wins (a list we'd currently be tied for 5th on, probably worse if you half count ties).

But Larry what we really want to know is what Lake Placid finally offered you to replace the package you won.

Next they'll start counting "moral victories" as wins.

We "meant" to beat Dartmouth earlier....really we did. That's gotta be at least .5 points.