ORCS Bids 2020

I know that these bids are all still very much live, but I wanted to start a threat to talk about the ORCS makeup now. Right off the bat, I know there has been a lot of discussion of Lancaster being tough, but Santa Monica looks hellish - UCSB, UCLA, UC Davis Stanford, Cal, UC Irvine, Arizona, Fresno State, and UCSD all have nationals abilities, and most of those programs are bringing both A and B as it stands now. Thoughts on the rest of the ORCS development?

Here's a link to the ORCS tier lists, I'll update it as AMTA gives out the rest of the open bids and shifts teams around: https://docs.google.com/spreadsheets/d/19bdU2JW53E_aARQdQ9qCnT7ZvrUjhqgww1nIEXRtDC8/edit?usp=sharing

 Santa Monica looks ridiculous... It seems very likely they'll shift a top team to Geneva, which currently looks the weakest.

I agree about Geneva but the new system (if I understand it right) creates an interesting situation: Moving a top team to Geneva doesn't necessarily seem to solve the problem. It would drive down the average TPR, yes. But right now through tiers B, C, and D in Geneva, there is only one team with a TPR of 97. Everybody else is over 100. Moving in another Tier A team (say, swap Howard for UIC) won't change the fact that every tier A team is pretty much guaranteed three rounds against teams with a TPR above 100. So the ORC remains "easy".

The fact that you have a stronger team in Tier 3 in Greenville in Howard B, than in Tier 1 in Geneva in Illinois-Chicago; its just craziness. The logical solution that jumps out to me is HowardA/B goes to Geneva in exchange for Gonzaga A/B or Oregon A/B to Greenville; William & Mary to Geneva; Washington to Cincinnati. Top end of Santa Monica is also nuts, but I'm not sure what to do on that. The obvious move to me would have been to shift UCSD to Memphis, but it looks like they're done with Memphis already.

Ghost of Perjuries Past wrote:The fact that you have a stronger team in Tier 3 in Greenville in Howard B, than in Tier 1 in Geneva in Illinois-Chicago; its just craziness.  The logical solution that jumps out to me is HowardA/B goes to Geneva in exchange for Gonzaga A/B or Oregon A/B to Greenville; William & Mary to Geneva; Washington to Cincinnati.  Top end of Santa Monica is also nuts, but I'm not sure what to do on that.  The obvious move to me would have been to shift UCSD to Memphis, but it looks like they're done with Memphis already.

Well that may be a logical solution from a purely competitive standpoint, but we have to remember these are also students traveling on university budgets most likely - cross country flights for Oregon/Gonzaga to South Carolina seems hard to pull off. Agree with the top end of Santa Monica - its brutal.

No doubt on the flights issue. My logic was in already being slated for Geneva, they're already schedule to fly cross country. So regardless of which ORCS they go to, it'll be a flight. Gonzaga/Oregon are both in a weird spot in that even if its Santa Monica for them, its still a 13/18 hour drive, so they're flying no matter what

AMTA also just added Rice B to Santa Monica. While they are technically unranked, they went 7-1 at Regionals, splitting with Oaklahoma A and beating UT Austin. They will be added to the bottom tier making the tournament that much harder. Curious to see what AMTA will do about this.

The Tier 1 in Santa Monica is indeed packed, but why does that matter? The Tier 1 teams won't hit another Tier 1 team more than once anyway right?. It's impossible in this new system for, say, UCLA A to be forced to hit Irvine and Stanford - so what is the issue there? The problem to me does not seem to be that any Tier 1 is super-stacked, but rather that the middle tier in Geneva is so weak. Miami A (TPR: 3) will not be forced to hit any team with a TPR better than 97 in three rounds. Whereas every other Tier 2 has teams roughly ranging from 50-100.

Calimocker123 wrote:AMTA also just added Rice B to Santa Monica. While they are technically unranked, they went 7-1 at Regionals, splitting with Oaklahoma A and beating UT Austin. They will be added to the bottom tier making the tournament that much harder. Curious to see what AMTA will do about this.

Rice got recoded because their "A" team didn't make it out. The Rice team at Norman will be Rice A at Memphis.

Pacificus wrote:The Tier 1 in Santa Monica is indeed packed, but why does that matter? The Tier 1 teams won't hit another Tier 1 team more than once anyway right?. It's impossible in this new system for, say, UCLA A to be forced to hit Irvine and Stanford - so what is the issue there? The problem to me does not seem to be that any Tier 1 is super-stacked, but rather that the middle tier in Geneva is so weak. Miami A (TPR: 3) will not be forced to hit any team with a TPR better than 97 in three rounds. Whereas every other Tier 2 has teams roughly ranging from 50-100.

I agree, this is the problem - the way I see it now, Miami will have no trouble whatsoever with any team at Geneva. Maybe I will be proven wrong, but I doubt it.

happygolucky wrote:

Pacificus wrote:The Tier 1 in Santa Monica is indeed packed, but why does that matter? The Tier 1 teams won't hit another Tier 1 team more than once anyway right?. It's impossible in this new system for, say, UCLA A to be forced to hit Irvine and Stanford - so what is the issue there? The problem to me does not seem to be that any Tier 1 is super-stacked, but rather that the middle tier in Geneva is so weak. Miami A (TPR: 3) will not be forced to hit any team with a TPR better than 97 in three rounds. Whereas every other Tier 2 has teams roughly ranging from 50-100.

I agree, this is the problem - the way I see it now, Miami will have no trouble whatsoever with any team at Geneva. Maybe I will be proven wrong, but I doubt it.

I get where you're coming from, but I think it's important to recognize that teams are still ultimately competing for bids to nationals, not just individual round wins. Even if you avoid hitting more than one 'top team' at ORCS, you're still competing with everyone at the tournament to move on. The A tier alone in Santa Monica isn't egregiously hard (it has less TPR than Tier A in Memphis and not significantly more than Tier A in Cincy or Princeton), but the 3 top ranked teams across ALL the B tiers are currently ALL in Santa Monica. Having 9 teams in Santa Monica who would be projected to bid based on national rank competing for only 6 bids is still worth addressing even if they'll avoid direct match-ups for 2 or 3 rounds. 

I also don't think Miami is going to have a ton of trouble no matter what the field is, based on their recent excellent nationals performances, so I'm not sure that encapsulates the problem. The bottom half of Geneva is actually pretty strong (relative to other C & D tiers), with the most teams ranked between 100-200 of any ORCS by far. I'm not as concerned with overall TPR/rank or whether any team will be expected to dominate, but the strength near the 6 bid cutoff point. Multiple teams in Geneva's A tier wouldn't be in the A tier if they were placed at any other tournament, and same for almost their entire B tier. 

I don't think it would be nearly as bad if there was a comparable A tier full of Top 50 teams and then a large drop off, even if that meant teams would still only hit 1 top 100 team. Not having to beat out teams who would be reasonably expected to bid at any ORCS they went to is the bigger problem to me, rather than easier head-to-head match-ups.

I imagine the B and C buckets in Geneva will be deceptively strong. Taking a look at B/C bucket v top team matchups at at ORCS, NIU won and tied a ballot against Illinois-Chicago, ND B had two very close ballots against Michigan A (and ND's A team in the B bucket may also pose a threat), Oregon A and C each took a ballot from Berkley A, and Denver took a ballot off Arizona. They will be dealing with a lot of low-ranked teams in Geneva who have shown they can rise to the occasion and make things interesting against top teams. I wouldn't be surprised if we saw some upsets. Miami A will likely have a pretty free run, but I don't think any of the other A bucket teams are necessarily guaranteed easy rounds.

University of Denver took a ballot of of Arizona’s B team. Southern Methodist University took a ballot off of Arizona’s A team. Southern Methodist swept University of Denver. Look for SMU to make a run

Now that we see exactly how this tier system is going to work out at specific ORCs, I think it's more apparent how this new ORCs pairing system may not be the best. The goal of this new system is to have more balanced schedules for teams, but when I look at some of these tiers I see major differences in the strength of teams within certain tiers. At this point in the season, TPR can't tell you everything about a team: some teams will have improved significantly, and others not be performing as well as they have in previous years. When you look at regionals tab summaries, some of the teams performing the best at regionals are the ones in the bottom tier: Minnesota B, Macalester B, Colorado College, Oklahoma, Hamilton A&B, etc.
Some of these performances may be flukes, but it's clear that TPR isn't a perfect indicator of how a team will perform in any given season. In my opinion, there should some way to account in performance at regionals into these tier assignments so team performance this season is somehow taken into account. If this tier system accounted for both TPR and regional performance, this would probably allow these tiers to more accurately reflect the strength of teams at this point in the season.

No longer up in the air. Only one Open Bid got declined, and now all ORCS bids are resolved. Looks like this year was brought to us by the number 27.

The flaw with using TPR is that, even if TPR were generally predictive, ORCS- by design- is going to wind up with disproportionately many teams that overachieve relative to their TPR (since some teams that did not make ORCS last year will make ORCS this year). AMTA could recompute TPR based on Regionals results (i.e., generate preliminary TPR points for the current year), before setting up these tiers, to capture teams that outdid expectations at Regionals without ignoring all the history. Some of the teams scoring a bunch of ballots at Regionals might be a little lucky, but- at the same time- all TPR is is just last year's Regionals/ORCS/Nats results. The chaos is everywhere.

Pairing is also the wrong place to solve this problem IMO. There's no good way to guarantee fair pairings for all rounds for all teams, since you need to know in advance how good the teams are (and data for that is v. sparse since we just have 1-3 official tournaments per team per year). A more robust (albeit more controversial) solution would be to adjust scoring and use a system (like Bayeselo*) that gives teams scores not just based on their W/L record but also on the strength of their opponents. Pairing is still important, given the small number of rounds per tournament, but there is no way to solve everything by only changing pairing.

* Bayeselo is controversial, though, because if, say, Miami were to face some weak team, they could "lose" points even for winning. At the same time, a normal team wouldn't get hit as hard just for losing to Miami. Oh, and there is the ouroboros thing where your pairings determine your ultimate score: e.g., a good team hitting good-to-great opponents all 4 rounds could end up 0-8 and therefore also make their opponents look bad (like they'd just beat an easy team). So we'll always need some effort on the pairing side since the ground truth of team strength is unknowable.