[OT] Lax - it's a whole new ballgame...

[ugarte]

Rich Hovorka '96 wrote:

Princeton 11 Brown 3

1 in 3 chance of the AQ.  I don't have much hope for an at-large selection.

Let's go RANDOM CHANCE!!!

So do you think Princeton wants the AQ, or do you think that they want an Ivy colleague to join them?

[kingpin248]

Dartmouth wins the men's draw and gets the automatic bid.

http://www.ivyleaguesports.com/article.asp?intID=2334



Post Edited (05-03-03 19:40)

[Jim Hyla]

The worst possible outcome for CU's chances.:-/

[jtwcornell91]

Ah well.  I guess we should be glad it wasn't wasted on Princeton, although Dartmouth probably had a shot at getting an at-large bid in the wacky world of NCAA lacrosse, where their "quality win" over Princeton would have mattered more than their #11 RPI ranking.

[getred]

Oh, well...not unexpected, but still disappointing:

 http://www.ncaasports.com/lacrosse/mens/story/6353179

[DeltaOne81]

< rant >Alright, I've been trying not to do this, but I just can't. Is it me, or does the lax selection criteria make NO SENSE. As with hockey, it's objective, so it's fair in as far as everyone knows what they're dealing with from the start of the year, but it's really messed up.

Making quality wins pretty much the sole selection factor just doesn't make sense to me. First of all, it totally screws over teams who aren't able to schedule the upper teams on a given year. Secondly, it just doesn't make any sense. We were #10 in RPI. We lost to everyone we played who was higher than 10, and we beat everyone we played who was less than 10. That seems pretty good proof to me that we *are* a #10 team, and, as such, would deserve to be in a 16 team tournament (unless there are more than 6 autobids that were below the 16 team mark). However, since we never "won up", we don't get picked. Well, in theory, a true #10 team never should win up.

Imagine if we had beaten Princeton, Syracuse, Hofstra, and Georgetown, but lost to Yale, Brown, Hobart, and Harvard. We'd have the same exact record, which plenty questionable losses, but we'd be a LOCK for the tournament.

It also makes no sense whatsoever to use RPI as a way to pick teams 1-5, 6-10, 11-15, yet completely ignore RPI as far as selection which teams get in. If RPI is good enough to show who the top 10 teams are as far as "quality wins" go, then why isn't it good enough to pick the top 16ish teams for the tournament?

Man these guys are screwed up. < / rant >

[jtwcornell91]

Kinda makes you thankful you're a hockey fan, doesn't it? ;-)

[kingpin248]

They certainly are screwed up.  The selection show was scheduled for 8:00 - it didn't go off until 8:45.  Not only that, a bracket was posted on the NCAA website at 7:30 - and it was the wrong bracket.

As for the ratings. For those interested, the RPI is available at http://www.vaporia.com/sports/collegelacrosserpi03.html .  CU finished #12 in the final RPI, and in fact beat one team ranked above it (Penn State at #11), while not losing to anyone ranked below. Using the RPI alone to pick the nine at-large teams, Cornell is the last team out (the nine teams selected are the ones that did in fact get in).

I've also been calculating KRACH-style ratings for this season. I call them the Z-Ratings (after Zermelo), because "KRACH" specifically implies hockey. They are at http://www.people.cornell.edu/pages/mfc22/ZRatings/Z-MLX.HTM . Cornell finished #12 in these ratings as well. If the nine at-large teams had been picked solely on those rankings, Cornell is the last team in (instead of Penn State - the other eight selections match the actual picks). In fact, the top four teams in the Z's match the four seeds.

And were the criteria followed? http://forums.laxpower.com/read.php?f=1&i=61742&t=61742 gives a table of the quality wins of teams in the at-large pool (scroll down to see it). "jhu72" put it well:

[q]This table presents one glaring problem for the committee, if they consider a top 5 win, better than any number of top 10 wins, then Princeton or UMASS won't go to the tournament. This is not likely. 0-3-1 UMASS and Princeton will likely be placed ahead of all teams that are 1-0-0.[/q]

That was exactly what happened.

The big problem with the lacrosse criteria is the valuation of specific games over others as a primary criterion, as Fred alluded to. The hockey PWR does contain a sort of "quality win" criterion (the record against TUCs, as well as the new bonus points), but they are not the overriding factor in the consideration of teams for selection.  Even the people in charge of hoops selection allude to evaluating the "body of work" of a team.

Hockey may not have a perfect system, or even a really great one, but the one we do have isn't that bad.

[DeltaOne81]

[Q]As for the ratings. For those interested, the RPI is available at http://www.vaporia.com/sports/collegelacrosse... . CU finished #12 in the final RPI, and in fact beat one team ranked above it (Penn State at #11), while not losing to anyone ranked below. Using the RPI alone to pick the nine at-large teams, Cornell is the last team out (the nine teams selected are the ones that did in fact get in).[/Q]
There are 16 teams in the tournament, there are 7 auto-bids. That's implying that *every* auto-bid went to a team below #9, is that accurate?
Edit: Actually, it's implying that 7 of the 9 did.

[Q]The big problem with the lacrosse criteria is the valuation of specific games over others as a primary criterion, as Fred alluded to. The hockey PWR does contain a sort of "quality win" criterion (the record against TUCs, as well as the new bonus points), but they are not the overriding factor in the consideration of teams for selection. Even the people in charge of hoops selection allude to evaluating the "body of work" of a team.

Hockey may not have a perfect system, or even a really great one, but the one we do have isn't that bad.[/Q]
Exactly. I think hockey's is very good, but lacrosse makes it look awesome. Hell, it even makes bouncy ball's look good. Lax looks at one thing and one thing only. I would say their system is okay if the top three "primary criteria" were given equal weight. Instead, by giving a "priority order", and having criteria which have enough variation to be high unlikely to be tied, you never get past the first one.

This is a scenario I realized last night, and was very close to actually happening this year. Say we have 0-0-2 quality wins (0 against 1-5, 0 against 6-10, 2 against 11-15) and we're competeing for a bid against a team that's 0-0-3 . Now, we're ranked 10 (which we were through Friday), and we beat #11 (Dartmouth through Friday). This means, had we LOST one more non-quality win game, we'd probably flip places with 11, and we'd have a better record of 0-1-1 , which will likely beat 0-0-3. So you can actually improve your chances by losing! I know any objective system with hard cutoffs is going to have some oddities, but by solely relying on one criteria, and such a imperfect one as this, leads to some seriously messed up possibilities.



Post Edited (05-05-03 11:40)

[kingpin248]

The participants, with RPI rankings:

Automatic qualifiers:
6 - Georgetown (ECAC)
8 - Towson (Colonial)
13 - Dartmouth (Ivy League)
21 - Ohio State (Great Western)
22 - Albany (America East)
26 - Army (Patriot)
37 - Mount St. Mary's (Metro Atlantic)

At-large berths:
1 - Virginia
2 - Johns Hopkins
3 - Maryland
4 - Princeton
5 - Massachusetts
7 - Syracuse
9 - Hofstra
10 - Rutgers
11 - Penn State



Post Edited (05-05-03 12:14)

[ugarte]

The field is 1 through 11 plus autobids, right?  If this is true, then there is no cause for complaint (other than the generic opposition to RPI-as-oracle).

[DeltaOne81]

[Q]The field is 1 through 11 plus autobids, right? If this is true, then there is no cause for complaint (other than the generic opposition to RPI-as-oracle).[/Q]
There is no cause for saying we got screwed. Note I've never said we got screwed, merely that the system sucks and has plenty of potential for screwing people over. I think this year is an example of "even a broken clock is right twice a day."