Splitting teams at SCT

[merv1618]

There's been a lot of heated debate about whether splitting up a school's A team at SCT to qualify more for ICT should still be a thing, especially recently. As this year's SCT is now over, I feel that it would be good for this to be hashed out. (Whether or not people also want to talk about autobids is up to them.)

I would also prefer that this discussion be limited to players in (or who have already finished) college.

High school players should feel free to contribute to this thread. --Mgmt.

[AKKOLADE]

I could see there being an argument that it violates the spirit of what the qualification process is trying to determine, which is the 32 best, or at least 32-X (where X is autobid count), teams in the country.

I don't have a real problem with autobids in theory, but if we're going to be using more hosts that are not regular powers in quiz bowl, then perhaps expanding the field by 8 or so would be beneficial.

[setht]

I'll admit upfront that I have not been closely following what's happened this year, so maybe there have been unusually deplorable shenanigans or something, but my general opinion is that allowing clubs to shuffle players around between SCT and ICT is less harmful than forbidding it. Splitting up A teams means that the ICT DI field generally doesn't have the 32 (or 32-X) best teams in the country, but forbidding player shuffling seems like it would also result in not having the 32-X best teams.

I'd love to see the ICT field sizes expand, but I'm not sure we're actually at a point where we can count on getting 40 teams to attend DI ICT.

Also, I haven't looked at D values or anything, but given that almost twice as many DII teams played SCT as DI teams, I'd have thought that expanding the DII field would make more sense than expanding DI. Is it actually the case that the DI bubble teams have higher D values than the DII bubble teams? Or maybe we should be comparing bubble D values with average in-the-field D values, or D value ranks, or something.

-Seth

[setht]

maybe we should be comparing bubble D values with average in-the-field D values, or D value ranks, or something.

It occurs to me that comparing D value ranks is a complete waste of time--since there are no D values for the X autobids in a division, presumably the D value ranks just read "this first bubble team is ranked 33-X" and so on.

-Seth

[Sima Guang Hater]

Yeah I really don't like it when teams split at SCT, especially good teams, since the whole point of SCT is to seed teams against each other at ICT. Things like Michigan and Chicago splitting their teams really don't help in this process, because it may be artificially inflating or deflating their ranking and thus screwing a 3rd place in a bracket team from a spot in the top bracket. Of course there's actually no way of enforcing it (other than saying your SCT team has to be identical to your ICT team I suppose?), so it's all academic anyway.

[Important Bird Area]

Yeah I really don't like it when teams split at SCT, especially good teams, since the whole point of SCT is to seed teams against each other at ICT. Things like Michigan and Chicago splitting their teams really don't help in this process, because it may be artificially inflating or deflating their ranking and thus screwing a 3rd place in a bracket team from a spot in the top bracket.

For the record: ICT seeding takes into account roster change between SCT and ICT; we do not just copy the SCT stats directly.

[kdroge]

I don't think it would be impossible to say that anyone who plays on an A or B team at SCT has to play on that same team at ICT. In the cases of us and Chicago, it's a lot more reasonable for Chicago to be splitting their teams right now because it's more unclear who their best players are (and probably irrelevant since their B team should qualify anyways). Since there are some people who play only one of SCT or ICT, it's unwise to seed based on SCT performance straight up anyways, so I don't see that as as much of a problem as it is that Chicago or our B team might take a spot at ICT from a more deserving team because we split our teams.

[The Ununtiable Twine]

Yeah I really don't like it when teams split at SCT, especially good teams, since the whole point of SCT is to seed teams against each other at ICT. Things like Michigan and Chicago splitting their teams really don't help in this process, because it may be artificially inflating or deflating their ranking and thus screwing a 3rd place in a bracket team from a spot in the top bracket.

For the record: ICT seeding takes into account roster change between SCT and ICT; we do not just copy the SCT stats directly.

Out of curiosity, what is the seeding procedure?

[Cody]

I don't think it would be impossible to say that anyone who plays on an A or B team at SCT has to play on that same team at ICT.

This is not true because SCT is really the only tournament to try out line-up changes before ICT (since NAQT is very different from [m]ACF tournaments). In addition, such a blanket rule doesn't take into account travel mishaps or sickness (h/t Evan Nagler here), which might force a school's A team to play short-handed were such a rule in place. (and before you start in on exceptions: what's to stop me from splitting teams at SCT and then not having the people that played up on the A team [say, they were normally C or D team people] not even come to ICT and then claim such an exception?)

While I personally dislike splitting teams, I'm not sure there's anything you can really do about it.

[ryanrosenberg]

I think what Kurtis was getting at is that people who couldn't go to SCT would be free to play on whatever team at ICT. However, if you played on your school's B team at SCT, you would have to play on the B team at ICT.

[Cody]

I think what Kurtis was getting at is that people who couldn't go to SCT would be free to play on whatever team at ICT. However, if you played on your school's B team at SCT, you would have to play on the B team at ICT.

That doesn't affect the above reasons against such a rule at all.

[ryanrosenberg]

I think what Kurtis was getting at is that people who couldn't go to SCT would be free to play on whatever team at ICT. However, if you played on your school's B team at SCT, you would have to play on the B team at ICT.

That doesn't affect the above reasons against such a rule at all.

The travel mishaps/sickness reason wouldn't be an issue; if the affected player is forced to miss SCT, he/she would be free to play on whatever letter team at ICT. I agree about not being able to switch around lineups on NAQT sets, though.

[Cody]

The travel mishaps/sickness reason wouldn't be an issue; if the affected player is forced to miss SCT, he/she would be free to play on whatever letter team at ICT. I agree about not being able to switch around lineups on NAQT sets, though.

Except the part where it can happen for ICT, as well.

[setht]

As I said before, I think a policy like "no moving players between teams after SCT" is a bad idea. There are lots of good reasons for a club to shuffle players around. I'm guessing that that kind of shuffling happens less often than shuffling because a club split its strongest players to try to grab more bids at SCT, but I don't think team-splitting is really a sin. What does feel like a sin to me is telling a team with a shot at a DI title, "No, you may not fill a sickness/availability-related roster hole with a strong player from your B team; you must play short-handed or grab some kid who's barely ready for DII, let alone DI (and burn his/her eligibility in the process)."

-Seth

[Cheynem]

I agree with Seth. I guess I'd like to see altogether less reliance on pure D-value for bidding and more of a general plea for teams not to split for purely bid grabbing reasons (I say this with the utmost hypocrisy as that is exactly what Minnesota did this year); trying to codify something is probably impossible.

[theMoMA]

Perhaps any players who score more than some cutoff percentage (50% or 40%, perhaps) of tossup points for different (qualifying) SCT teams could be required to play for separate teams at ICT. Teams could then shuffle their lineups (except the top scorers, who are pretty predictable) due to illness or improvement. So the rule would prevent splitting up A teams in the way we've usually seen.

(Note I have nothing against this splitting, and my team has benefited from it several times; this is just a proposal that would effectively end it without messing with the flexibility of bigger programs too much.)

[Cody]

Perhaps any players who score more than some cutoff percentage (50% or 40%, perhaps) of points for different (qualifying) SCT teams could be required to play for separate teams at ICT. Teams could then shuffle their lineups (except the top scorers, who are pretty predictable) due to illness or improvement. So the rule would prevent splitting up A teams in the way we've usually seen.

(Note I have nothing against this splitting, and my team has benefited from it several times; this is just a proposal that would effectively end it without messing with the flexibility of bigger programs too much.)

I thought about this, but consider UVA, this year: by trying Dennis out on the A team, they had to move either Evan or Tommy down (after all, Matt B + the Seal are definitely going to be on the A team at ICT) or else play a 3 person B team (which is just dumb when you have 8 people). Either of those two would almost certainly score 40-50% of their teams points and would guarantee an ICT qualification. Such a rule would basically force that player to tank their games in order to either not score enough points or not qualify so that that UVA could keep its options open. That's dumb and such a rule just encourages gaming the system.

[Important Bird Area]

Out of curiosity, what is the seeding procedure?

Teams are ranked from 1-32 using a variety of inputs. (SCT results are for obvious reasons the primary input. However, we also take into account non-NAQT results and previous years' results, especially in cases where team rosters have changed between SCT and ICT, and in evaluating the strength of the teams from the host schools that did not play SCT. We do our best to balance the ICT brackets both by team strength and by geographic region.)

[merv1618]

How does that logic apply to autobids then? I say this because my team is extremely affected by them this year.

[theMoMA]

I thought about this, but consider UVA, this year: by trying Dennis out on the A team, they had to move either Evan or Tommy down (after all, Matt B + the Seal are definitely going to be on the A team at ICT) or else play a 3 person B team (which is just dumb when you have 8 people). Either of those two would almost certainly score 40-50% of their teams points and would guarantee an ICT qualification. Such a rule would basically force that player to tank their games in order to either not score enough points or not qualify so that that UVA could keep its options open. That's dumb and such a rule just encourages gaming the system.

They could just have Evan or Tommy play for a C team, and renounce the third bid. The rule could say that renouncing a bid would result in all players on that team being free to play for any other qualifying team from the same school.

[Cheynem]

What if they didn't want to pay for a C team?

[Cody]

They could just have Evan or Tommy play for a C team, and renounce the third bid. The rule could say that renouncing a bid would result in all players on that team being free to play for any other qualifying team from the same school.

That's kind of absurd. 1) They only had 8 people, so this doesn't solve the problem of having a short-handed B team. 2) Another team costs $$$. 3) This means they're forced to play solo? 4) What if they play on the B team and want to take the bid? There's nothing to say that a normal full UVA B wouldn't have still qualified--why would they have to sacrifice their bid?

[Matt Weiner]

That seems like a ridiculously unworkable solution to a problem that doesn't exist. No one is going to pay more for a C team just to comply with some ridiculous rule about roster changes not being allowed. This year we saw at most one team take an undeserved bid by splitting--Minnesota B got their bid by special arrangement involving editing, UVa B and Chicago B surely would have qualified anyway if they had played their ICT A and B lineups, and only Michigan B got in by splitting their A team (and it's entirely plausible that real Michigan B would have qualified on its own).

What we did see was Buffalo, Central Florida, Delaware, McMaster, Texas A&M, Truman State, and Washington get autobids from hosting. Washington turned theirs down and Delaware had some shot of qualifying anyway, so we'll call that four host bids that definitely displaced qualifying teams. WUSTL was 49th in D-value and would not have come close to qualifying had they not won a depleted miniature Sectional, and Cornell seems to have benefited from the generously small correction given to teams who play DII opponents on DII questions. Let's assume Cornell would have squeaked in anyway (I'm also very sure that Rice would have qualified in an actual DI tournament) and say we have a definite minimum of five DI bids that were given to teams that could not have qualified under a system where everyone had an equal burden to qualify by 1) actually playing the tournament 2) against actual DI opponents 3) on actual DI questions 4) without claiming an auto-bid for putting up 2 powers per game and 14 points per bonus while beating Truman State and Harding nine times.

The autobid situation is way more pressing than the B team situation by any measure and is also way more plausible to deal with. I don't know why the focus has fallen on the less important, less practical-to-fix aspect of qualification.

[setht]

Perhaps any players who score more than some cutoff percentage (50% or 40%, perhaps) of tossup points for different (qualifying) SCT teams could be required to play for separate teams at ICT. Teams could then shuffle their lineups (except the top scorers, who are pretty predictable) due to illness or improvement. So the rule would prevent splitting up A teams in the way we've usually seen.

(Note I have nothing against this splitting, and my team has benefited from it several times; this is just a proposal that would effectively end it without messing with the flexibility of bigger programs too much.)

This seems problematic for a club with, say, 5 or 6 strong players. Look at Minnesota's (DI) situation in 2010: here's a club that split 7 people over 3 teams, and took the top 3 spots at its SCT. I assume you wound up getting 3 bids by the time all was said and done. Obviously if a "no player shuffling"-type rule is in play you wouldn't have split things up so much; let's assume for the sake of argument that at SCT you would have fielded the same A and B teams that you wound up fielding at ICT. Does Mike Cheyne score 40+ or 50+% of the B team's points at SCT in this scenario? He certainly did at ICT, so I'll go with "yes." Now imagine that Rob found out, after SCT, that he couldn't make it to ICT. Would you be okay with not being allowed to bring Mike onto your A team, which is clearly in the mix for the overall title? I guess you should ignore considerations of the DI undergrad title for this thought experiment.

-Seth

[theMoMA]

Sure, there are still downsides to any rules scheme to avoid team splitting. There are ways to minimize them, of course, and that's all I'm trying to demonstrate. It's still a question of whether the overall benefit to the field outweighs those downsides. So far, it seems like the general response is that the marginal benefit doesn't outweigh the potential cost (in both dollars and flexibility), and that's fine.

[Important Bird Area]

I started a new thread for autobid discussion.

[The Kid Who Collects Spider-Man]

What would be wrong with leaving some discretion to NAQT officials to determine whether or not teams can shuffle players between SCT and ICT? The default could be no shuffling, and teams could make appeals if they have a legitimate reason to alter their rosters.

[evilmonkey]

EDIT: As mentioned in my below post, I retract the conclusions of this post.

The autobid situation is way more pressing than the B team situation by any measure and is also way more plausible to deal with. I don't know why the focus has fallen on the less important, less practical-to-fix aspect of qualification.

Last year, we saw teams from Ohio State B, Illinois B, and Chicago C all take bids by splitting their teams, and all finished bottom bracket; SCT Chicago B contained two members of the ICT Chicago A team, and ICT Chicago B finished bottom bracket. I would also count as a bit shady Michigan B (at SCT, was Kurtis solo), who only managed to make the third bracket.

Meanwhile, the following earned host autobids: (top bracket) Harvard; (2nd bracket) Michigan A, UCSD; (3rd bracket) Carleton, Arizona State; (4th bracket) Toronto, Texas A&M. To the best of my knowledge, all of the bottom three earned a bid through winning their sectional during the 2010-11 season.

7 teams, spread fairly evenly throughout the bracket; or 5 teams, none of which finished higher than 3rd bracket. Which is the bigger issue?

Disclaimer: I do feel a bit disingenuous making this point, when obviously my team benefits from the host bid. That fact alone has kept me from mentioning this to this point. However, since people are railing against the evil that is the host autobid, I thought it would be useful to increase our sample size of prior years to 1.

[Unicolored Jay]

The autobid situation is way more pressing than the B team situation by any measure and is also way more plausible to deal with. I don't know why the focus has fallen on the less important, less practical-to-fix aspect of qualification.

Last year, we saw teams from Ohio State B, Illinois B, and Chicago C all take bids by splitting their teams, and all finished bottom bracket; SCT Chicago B contained two members of the ICT Chicago A team, and ICT Chicago B finished bottom bracket. I would also count as a bit shady Michigan B (at SCT, was Kurtis solo), who only managed to make the third bracket.

We didn't split our teams last year.

[evilmonkey]

It appears that I should verify my research more before posting, or give it a second glance when it is not 3 AM. Ohio State B and Chicago C's weak showings appear to be more based on the absence of strong players, than any attempt to split up the A team to win bids. I recant my support for my above conclusions that splitting teams is a "bigger issue". However, I do think it is instructive to look at where the bottom bracket came from:

Chicago B - At Large (5th in D-Value) - Contained 3/4 scorers from SCT Chicago A, 1/2 scorers from SCT Chicago C
VCU - At Large (16) - Full SCT team
Toronto - At Large (30) - Best player at ICT did not play SCT
Illinois B - At Large (24) - Split A team in attempt to earn second bid
Ohio State B - At Large (26) - Contained 2/4 scorers from SCT Ohio State B; Lead Scorer was called up to A team to replace absent member
Texas A&M - Host Bid - Qualified for 2011 ICT through SCT performance
Guelph - Site Winner (29) - 70% of SCT scoring present at ICT
Chicago C - At Large (31) - Best player at ICT was 4th scorer at SCT; no other ICT player played SCT