The Quizbowl Resource Center

This is an announcement for the Midwest Mirror of the Harvard Fall Tournament being held at Rockford Auburn High School, taking place on Saturday, December 11, 2010.

Unless we have a power outage due to a snow storm like we did two years ago, this tournament will be held at Auburn's Main Campus, on the first floor, at 5110 Auburn Street, Rockford, IL 61101-2498.

The check-in is from 8 AM to 8:30 AM, with a coaches' meeting at 8:30. We hope to start the first round by 9.

As previously mentioned, this tournament is being run on the Harvard Fall Tournament set, which you can find information about at viewtopic.php?f=1&t=9870#p185165. There are no powers, but negs, with nonrebounding bonuses.

As of now, the field is closed.

Updated Field (as of 12/7/10)

Auburn
Boylan
St. Ignatius
Wheaton North
Wheaton-Warrenville South - 2
Hinckley Big-Rock
Culver (IN) -2
Carbondale
Loyola Academy
Stillman Valley
IMSA - 2
New Trier
Seven Lakes (Katy, TX) - 2
Keith Country Day School - 2
Buffalo Grove
Stevenson - 2
Lisle
Guilford

Lisle - 1st on wait list

total- 24 w/ 1 wait list

2 Questions:

A) ACF Finals this year?

B) How are ties being broken?

Seven Lakes is going to be here?? This should be interesting.

Secretary of Bobcats wrote:2 Questions:

A) ACF Finals this year?

B) How are ties being broken?

We discussed this today, and unless something changes, I'm pretty sure this is how things will work:

Prelims will consist of 4 pools of 6 each, full RR within each pool.
Playoffs will consist of 4 pools of 6 each, full RR within each pool. The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.
As far as finals are concerned, if a team within the playoff pool clears the field by exactly 2 games (5-0 vs. 3-2 teams) they win outright. If a team goes 5-0 and two teams go 4-1, they will play a half-packet (?) in order to get the right to play the 5-0 team in an advantaged final. If two teams go 4-1, they will play a single game final. In the event of a 3-team circle of death, the team with the highest PPG within the playoff pool will play the winner of a half-packet match between the two other teams with lower PPG's in a single game final.

In short, yeah, ACF finals, but there's a good chance we'll use half-packets in the situations shown. I dunno how consistent that is with an ACF-style final.

EDIT: clarity

Your Cup of Robust Tea wrote:Prelims will consist of 4 pools of 6 each, full RR within each pool.
Playoffs will consist of 4 pools of 6 each, full RR within each pool. The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.

PPB, or PPCB? Are you using rebounds?

dtaylor4 wrote:

Your Cup of Robust Tea wrote:Prelims will consist of 4 pools of 6 each, full RR within each pool.
Playoffs will consist of 4 pools of 6 each, full RR within each pool. The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.

PPB, or PPCB? Are you using rebounds?

I do not think we are planning to use rebounds. So PPB=PPCB.

Your Cup of Robust Tea wrote:The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.

Do you mean straight up highest ppb or record then highest ppb?

Secretary of Bobcats wrote:

Your Cup of Robust Tea wrote:The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.

Do you mean straight up highest ppb or record then highest ppb?

I'm pretty sure he meant just PPB, since it's likely that those teams will come from different brackets and have faced different opponents.

I think Nolan is alluding to the posibility of a team with a low record and a low TU count. It's feasible a team could have such a small sample size that their PPB could eclipse a team with 3-4 more wins than them.

Secretary of Bobcats wrote:

Your Cup of Robust Tea wrote:The 6 teams will be decided by the 4 pool winners + 2 remaining teams with highest PPB.

Do you mean straight up highest ppb or record then highest ppb?

Straight up highest PPB. So, yeah, it's theoretically possible that the two teams would come from the same bracket (although this would probably indicate that pool seeding was done poorly). Although what Garb said is possible, it's quite unlikely, no?

Your Cup of Robust Tea wrote:
Straight up highest PPB. So, yeah, it's theoretically possible that the two teams would come from the same bracket (although this would probably indicate that pool seeding was done poorly). Although what Garb said is possible, it's quite unlikely, no?

It definitely happens http://www.ihssbca.org/statistics/2010_ ... ndings.php

BGSO wrote:

Your Cup of Robust Tea wrote:
Straight up highest PPB. So, yeah, it's theoretically possible that the two teams would come from the same bracket (although this would probably indicate that pool seeding was done poorly). Although what Garb said is possible, it's quite unlikely, no?

It definitely happens http://www.ihssbca.org/statistics/2010_ ... ndings.php

Yea, check out KCD B, putting up almost 18ppb despite not winning a single prelim game.

dtaylor4 wrote:

BGSO wrote:

Your Cup of Robust Tea wrote:
Straight up highest PPB. So, yeah, it's theoretically possible that the two teams would come from the same bracket (although this would probably indicate that pool seeding was done poorly). Although what Garb said is possible, it's quite unlikely, no?

It definitely happens http://www.ihssbca.org/statistics/2010_ ... ndings.php

Yea, check out KCD B, putting up almost 18ppb despite not winning a single prelim game.

Right. In the end, though, in this situation, Lisle (21.71 PPB) and Carbondale (22.81 PPB) would still have been the two other teams to have gone through, no? Maybe I'm misunderstanding your point here, but among the 20 non pool-winning teams, the top 2 teams in terms of PPB are pretty clear.

EDIT: Never mind, Jonah just explained to me what I'm missing: high bonus conversion on small sample size. Would it make sense to make it so that only teams that are in 2nd or 3rd place within the pool are in contention for getting through, then?

That was a tournament that allowed bonus rebounds but did not differentiate controlled bonuses vs rebounded bonuses in the stats. You probably wouldn't get stats like that if you kept stats normally.

EDIT: I'm an idiot.

Westwon wrote:That was a tournament that allowed bonus rebounds but did not differentiate controlled bonuses vs rebounded bonuses in the stats.

No, it wasn't. There were no bouncebacks at Loyburn. C'mon, you know Riley better than that.

I don't think breaking ties purely on points per bonus is a good idea when you can potentially be excluding teams that won more games. For example, if you have a situation where one bracket has three teams go 4-1, and all the other brackets have the third place team going 3-2, I would say a tiebreaker allowing in the 4-1 3rd place team over the other teams is fairer regardless of their ppb, because even if their schedules aren't equal, one has won more games than the rest, which is ultimately the metric you use to determine the results of a tournament. They are also being compared on a more equal footing that way, because if your tournament is seeded fairly then the brackets should not be ridiculously out of whack, and unlike with PPB, which is prone to more randomness due to the fact that teams are being compared on a metric that is reliant on how many tossups a team got and which is ultra unequal since there are undoubtedly going to be many different bonuses heard by each team in a schedule, by weighting win-loss first you are comparing teams by how they did against similarly ranked opponents over the course of hearing the same 5, 20 tossup, 800 point games, which is a lot more uniform in how you can compare teams. I also think that bonus conversion is far too prone to statistical error to be a reliable statistic to use to compare teams in the middle of a tournament. Bonus conversion is useful in comparing teams after the fact, but ultimately there is far too much randomness inherent in it for me to feel comfortable with tournaments using it in as the primary component of a tiebreaker mechanism.

What if a team converts one tossup total in the prelims, but 30s the bonus? They'd be 0-5, but wouldn't straight-up PPB put them in the highest bracket?

whitesoxfan wrote:What if a team converts one tossup total in the prelims, but 30s the bonus? They'd be 0-5, but wouldn't straight-up PPB put them in the highest bracket?

STOP PROPOSING wildly outlandish SCENARIOS.

A system is flawed if it would allow a mistake that extreme. It's unlikely that it could happen, but if it does, it would be the worst bracketing mistake in all of history. Other people have brought up more likely scenarios that would also result in mistakes.

What if every team ever had the exact same stats? What would happen then under any tournament format? WHAT THEN

SirT wrote:

whitesoxfan wrote:What if a team converts one tossup total in the prelims, but 30s the bonus? They'd be 0-5, but wouldn't straight-up PPB put them in the highest bracket?

STOP PROPOSING wildly outlandish SCENARIOS.

1) This is outlandish, as Cody indicated.

2) It's by no means the finalized structure, no questions. Be patient.

Sorry for not reading Abid's post carefully enough where he said that things might change. My example was outlandish, but it was only intended to be a simple but extreme example of the flaw that teams with a small number of tossups could have a high PPB.

However, I don't understand why some people felt the need to use caps lock.

whitesoxfan wrote: However, I don't understand why some people felt the need to use caps lock.

Welcome to the Boards.

Some of the information posted previously was premature - the team and I had discussed a couple of options for the afternoon rounds, but not decided on one of them yet. Our main point of discussioin was whether to have 6 or 8 in the top flight, going for the championship. Given the strength of the field, we are going with 8.
Here is what we plan to do: All teams are guaranteed 10 rounds of play, 5 in the AM and 5 in the PM. In the AM there will be four pools of six, with round-robin play for five rounds. Pools will be seeded using available comparable tournament results (HSAPQ 15 and whatever else I have access to) so that each of the pools will contain two of the top teams in the field. Based on the morning results, we will re-bracket for the afternoon (round-robin within each bracket) into four groups - a championship flight consisting of 8 teams (the top two from each AM pool), a consolation flight consisting of the next six teams, and two other flights consisting of 5 teams each (those two flights will play a cross-over to get in their fifth PM match).
Since the championship flight has 8 (rather than 6) teams, and we want round-robin play among those teams to determine the final places, we will count the one match played against the other team in the same pool in the morning and teams will play the six other teams in the afternoon; those teams will have played a total of 11 rounds, 7 of which count for the championship. We will use ACF finals format, so unless one team has cleared the field by two, a couple of teams may play a 12th round (or more in the case of ties).
The AM pool winners will be determined by record, then PPG, since they will have played a round-robin format on the same packets of questions (there are no byes). Placement of the other teams in the other three flights will be determined the same way, except that it may be possible for, say, 3 teams from one pool and just 1 from another to go into a particular flight, depending on their records and PPG. There is no need to decide whether to use PPG or PPB to seed the afternoon matches within each bracket because it is round-robin play.
I hope this clarifies things for those who are interested.
Also, here's a minor field update: I am giving Auburn B's spot to Lisle. And, we have two more teams on the wait list: Rockford Guilford, and Byron.

SirT wrote:

whitesoxfan wrote:What if a team converts one tossup total in the prelims, but 30s the bonus? They'd be 0-5, but wouldn't straight-up PPB put them in the highest bracket?

STOP PROPOSING wildly outlandish SCENARIOS.

This really isn't all that outlandish. Extremely unlikely, yes, but not outlandish. There are certainly teams out there whose players have been waiting their entire careers to power that one Brett Favre question only to have it followed by that one Lonely Island bonus. They would lose that game 45-x after putting up a 1/0/0 statline. However, they would have AWESOME PPB. If this happened in their first match, they could refuse to buzz at all for the next four rounds and earn a sure wildcard bid. In fact, if a team 30d their first bonus in their first match, they could just not buzz the rest of the morning and make the playoffs.

The point is that while we might not live in daily fear of this sort of thing happening, it is easy to construct systems that prevent it. In fact, such systems, like breaking ties by record then PPB, are better for other situations as well (not just the outlandish), so EVERYBODY WINS!!!

Lloyd, Abid, and the rest of the Auburn team will come up with a very reasonable system. HFT@Auburn will be enjoyed by all.

Dominator wrote:The point is that while we might not live in daily fear of this sort of thing happening, it is easy to construct systems that prevent it. In fact, such systems, like breaking ties by record then PPB, are better for other situations as well (not just the outlandish), so EVERYBODY WINS!!!

Well, there's actual reasons to argue against using head-to-head (one I've seen is that it effectively assigns twice the importance to a single game).

You could also handle the ppb issue by stating that teams must answer, you know, multiple tossups to qualify for a wild card.

Coach G wrote:...minor field update: I am giving Auburn B's spot to Lisle. And, we have two more teams on the wait list: Rockford Guilford, and Byron.

That's a very courteous gesture. Thanks.

I have been told that this is 5-on-5.

It is 5 on 5, but teams may play with just four if they choose to, or if they find themselves short-handed that day due to illness, an emergency, etc.

Seven Lakes 620, Stillman Valley 10.

Live stats? Any chance, people who are in the stats room?

mlaird wrote:Live stats? Any chance, people who are in the stats room?

Given who's doing it, I doubt it.

mlaird wrote:Live stats? Any chance, people who are in the stats room?

I don't think any of those people read this board, but Seven Lakes defeated Wheaton Warrenville South B 555-(-5).

St. Ignatius (Deveau playing solo): beat Keith Country Day A, Stevenson B, Wheaton Warrenville South B 400-95, and Stillman Valley 430-30. One of the first two scores was 465-65, but I don't know which and I don't know the other.

edit: added another match

About to read Seven Lakes A vs. St. Ignatius. Will post updates.

175-165 Seven Lakes at half.

I'm interested in a New Trier update if I can get one.

Westwon wrote:I'm interested in a New Trier update if I can get one.

They're in a bracket with Stevenson A, Lisle, IMSA B, Guilford, and Hinckley-Big Rock. I'll try to get you results at lunch.

350-305 Deveau final. A lot of ridiculous powers from Andrew in the second half.

jonah wrote:

Westwon wrote:I'm interested in a New Trier update if I can get one.

They're in a bracket with Stevenson A, Lisle, IMSA B, Guilford, and Hinckley-Big Rock. I'll try to get you results at lunch.

3-2, having lost to Stevenson A and Lisle.

Top bracket: Stevenson A, Auburn, St. Ignatius, IMSA A, Lisle, Carbondale, Seven Lakes A, Loyola. The playoffs are a full round robin with the already-played game carried over, so the first four are 1-0, having defeated the corresponding team in the second four (Stevenson over Lisle, etc.)

Top individual scorers for prelims: 1 Deveau (St. Ignatius), 2 Kevin Malis (Stevenson), 3 Ben Chametzky (Carbondale), 4 Andrew Van Duyn, 5 Jiawei (Seven Lakes A).

Championship bracket: Stevenson A, Auburn, St. Ignatius, IMSA A, Lisle, Carbondale, Seven Lakes A, Loyola.

Top consolation bracket: New Trier, Culver A, Buffalo Grove, Keith Country Day A, Seven Lakes B, IMSA B

Second consolation bracket: Stevenson B, Wheaton North, Wheaton Warrenville South A, Boylan, Wheaton Warrenville South B

Third consolation bracket: Guilford, Hinckley-Big Rock, Culver B, Stillman Valley, Keith Country Day B

Seven Lakes A defeats IMSA A, 420-240.

Carbondale 255, IMSA A 215.

Seven Lakes A 330, Stevenson 325 (!)

Loyola 365, Auburn 355

Auburn 425, Seven Lakes A 290

Carbondale 305, Lisle 240

Stevenson 410, St. Ignatius 205

Auburn 515, IMSA 130