Discuss 2016 NHBB Nationals Here

Dormant threads from the high school sections are preserved here.
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

Thanks to everyone who has commented so far specifically on Nationals, or on the specific changes we implemented for this year. I don’t want to go over all the specifics of why the changes were made, since much of that already exists in a prior thread from last year’s Nationals. But I do want to address a few things people have said, and expound a bit more upon my thinking / what’s likely to happen for next year.

Regarding comments above…

1. Personally, with very, very few exceptions (i.e. there was no way this year’s Nationals final match was going to have a question on Jugurtha), I don’t at all remember, nor pay attention to what answerlines have come up in past years. I’m sure the same is true for Brad, and all of our writers as well. To a certain extent, if there was a very notable answerline in a regionals set, I’ll try and ensure this doesn’t come up at Nationals, but this obviously isn’t watertight by any means.
Do note as well that the official NHBB repeat policy is that we don’t automatically control for repeats per se across the USHB set vis a vis the rest of the weekend (though we do try to limit repeated clues where possible, especially if it’s something memorable and specific).

2. Very briefly, the draft was instituted last year after two specific instances showed me that simply having the 1 seed play the 8 seed and 2nd seed play the 7th seed, etc. off of evening pool records then pts was suboptimal. First, I remember Rohan from High Tech coming up to me and being delighted that they were the 6 seed rather than the 5 seed last year, since that meant they were in the half of the bracket away from Saratoga (as it was, High Tech lost in the 1/4s last year to Bethlehem). Secondly, it seemed odd that, as we were going to a 6 pool format on Saturday, the second place teams (who normally would be the 7 or 8 seed) could easily be of the sort like DCC or IMSA last year, which would be unfair to the 1 or 2 seed, if the 5 or 6 seeds were weaker, as per a general consensus. Hence the draft. Thus as some people have said, there’s nothing inherently wrong with the draft, and it’s existence is there not for some gimmicky add on, but rather to ensure maximal fairness (in, admittedly, a novel way). The draft is here to stay – it works.

3. On the matter of three way games. Note that this can occur in two instances – either when the evening group has 3 teams at 2-1 (intragroup), or when across the pools (this was only the case in Varsity this year), there are either 6, 5, or 3 teams with 2-1 records (i.e. intergroup). On the intragroup, that’s there because there was a brouhaha last year when 3 teams went 2-1 in one group and the winner advanced solely on points. Now, with having the pool winners advance automatically, but allowing second place 2-1 teams two slots on Sunday from the intergroup, one could argue that we’d be better just having a 2 way match between the 2nd and 3rd on pts. in a group with 3 teams at 2-1, and letting the winner through. But that seems to skirt the whole reason as to why this was controversial last year. And what if the 3 teams shake out with pt totals of 900, 890, and 600 (which is possible)?

This is what I’m getting at when I’m arguing that this tournament was very, very close to Pareto optimization. We could tinker with things further, like with this way of resolving 3 teams at 2-1 (or, say, by having the 2 and 3 seeds meet, and then play the winner, though that still largely negates the value of wins, and adds further layers to what is already happening late on Saturday), but honestly, I think that a 3 way match is the fairest way to resolve this. Win your matches – and it won’t be an issue…

Same thing in the intergroup. If all 6 second place teams went 2-1, then how do we get just two teams? A two tiered playoff in the intergroup? That again seems to value points very highly at the expense of wins (plus it would make things go on that much longer). And with 5 teams, as was the case this year, we could have the 1st team on pts advance automatically, while playing off the 2nd spot in a 2 tiered format, but again, that seems to put an inordinately high value on the point distinction between 1st and 2nd. As is, the first place team has it easier by only competing against one team. Maybe that should be the 4 or even the 3 seed; on that, I’m a bit uncertain. But there’s nothing that I see inherently wrong in having 3 teams play at once – at most, it’s 12 players on the buzzer, which isn’t that different from 10 players on the buzzer which is common enough in various Bee finals.

4. On the Bee scoring system. First, I apologize that we didn’t have monitors as I had hoped. We will work on that for next year, as that would make it much easier for players and audience members alike to follow. Two changes that I’d like to make are as follows. First, we’ll be instituting this system in regionals playoffs for 2016-17 (as well as at the Olympiad this summer). That will give students a chance to get used to it before Nationals playoffs, which, admittedly, isn’t the ideal time to be trying something brand new.

Secondly, Greg Gauthier pointed out to me that it’s still possible to game the system. Say, if a player was on question 34 in a round limited to 35 questions, and had an 8 point lead and the question was over. If two students had buzzed in already and were wrong, then the leader could kill the question, take the -1, and ensure victory. To avoid this, in the new scoring system, with -1 at the end, we’ll now allow an unlimited number of students a shot at answering (though after 3 seconds of no buzzing, the question will still be dead). Thanks to Greg for pointing that out.

Otherwise, I think the new system has the incentives down the way it should. It still certainly rewards an educated guess (If you’re even 26% sure at the end, then go for it. If you’re not, then you shouldn’t be throwing out a relatively wild guess when someone else may have more solid knowledge on the topic). I do agree that 5 or 6 points is a lot, but at this point in the tournament, the Bee needs to differentiate between the top players, and power marking helps accomplish this in a way that leaves the competition less sensitive to buzzer races. The -2 on the other hand, makes sure that you really need to know your stuff before buzzing. Remember that last year, with no point disincentive to ring in early and very stiff competition, we often saw the pattern of wild early guess, wild early guess, followed by either a buzzer race off an obvious clue or question killing at the end. Alex’s suggestion of -1s throughout, with a 4, 3, 2 system is worth considering, but I do like the idea of both the question being worth less after the giveaway (which in the Bee playoffs is often the only place an answer becomes known due to the difficulty of the answerlines), and still relatively rewarding an educated guess (on a 3:1 ratio which seems about right) at the end.

This all brings me to…

5. Changes for next year’s Bee playoffs.
I do agree that 32 students in the National History Bee playoffs seems low (though again, we’re pressed for time on Sunday afternoon). Keep in mind that scores were high this year, as many rooms only had 6 players and 35 questions. But still, an increase to 40 students seems in order. So next year, we’ll have 5 rooms of 8, the top 3 in each room will advance automatically to the semis, and the highest fourth place player will as well. Then, from 16, we’ll take the top 3 in two rooms of 8, and have six player finals. We’ll institute this in both Varsity and JV.
That said, we’ll increase the round length from 35 to 40 questions in the National History Bee quarters and semis (as well as the USHB and USGO rounds, with the exception of the USHB Varsity finals), and increase the National History Bee finals to 50 questions. I’ll need to go back and figure out the optimal totals for automatic advancement off of this year’s tallies, but it’s likely to stay at 40 for most rounds, and stay at 50 in the finals (which will be harder to get to with an added player, or 3, in the case of the JV).

One further comment on the matter of various rooms having more or fewer students than other rooms. This is usually a matter of students dropping out, either in advance, or even mid-Bee (which, frustrating though it may be, is hardly unprecedented or rare). In rare occurrences, it can be due to the need to resolve a conflict (i.e. a student being given a wrong room / the need to accommodate a student with limited mobility / a student missing the start of a round for a reason that wasn’t their fault). We should clarify that while NHBB strives to maintain optimal fairness by both seeding the draw and trying to ensure parity in the number of students in each room, that this alone is not grounds for protest with the sole exception being if a student walks into the wrong room (this happened in 2015 in one notable example discussed last year), and thus skews the results in a way that was not authorized. NHBB reserves the right to adjust the scores of students in such a situation in as fair a way as possible given the specific circumstances (the lack of more definitive instructions is purposeful, since various situations lend themselves to different solutions depending on exactly what happened - e.g. it’s a different story if the student going into a wrong room scores 13 or 0 in the round).

6. Taking top 2 out of morning rounds.

This is likely here to stay – it leads to more competitive games in the afternoon, and most teams that finished in 3rd place last year and went to upper bracket were subsequently defeated resoundingly to no one’s great edification. I would like to explore the possibility of in the event of having the draw # be a multiple of 6 but not of 18 (i.e. the case with our JV draw this year) having the 2 or 4 remaining Upper Bracket slots be open to all teams, and not just those within a particular supergroup, but I’ll need to discuss the logistics and feasibility of this with our stats team to see if they feel confident in implementing such a scheme (especially if they would need to do this in both Varsity and JV, with an expanded field next year likely to be at or over 300 teams, given a projected major nationwide outreach effort at the start of the 2016-17 season.

7. To clarify one last thing – no rules were adjusted at any point during the tournament, and the new bee scoring system was announced early last fall, in fact. The only thing I was uncertain of at one point, was whether, with respect to the intergroup playoffs, the 1st seed played the 5th or the 4th. I knew I had posted this on our website, but with the website inaccessible, I was about 80% sure I had said 1st plays 5th, and then announced that since I couldn’t get online, that would just be the way it would work regardless of what the website said. Note that I announced that at the start of the evening pools before anyone knew who would be 1st, 5th, or even if we’d end up in this situation (though in the event, we did). Though as it turned out, the website did say 1st played 5th and 2nd, 3rd, and 4th played in the other, so that was as it should have been.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
User avatar
jonpin
Auron
Posts: 2266
Joined: Wed Feb 04, 2004 6:45 pm
Location: BCA NJ / WUSTL MO / Hackensack NJ

Re: Discuss 2016 NHBB Nationals Here

Post by jonpin »

Great Bustard wrote:6. Taking top 2 out of morning rounds.

This is likely here to stay – it leads to more competitive games in the afternoon, and most teams that finished in 3rd place last year and went to upper bracket were subsequently defeated resoundingly to no one’s great edification. I would like to explore the possibility of in the event of having the draw # be a multiple of 6 but not of 18 (i.e. the case with our JV draw this year) having the 2 or 4 remaining Upper Bracket slots be open to all teams, and not just those within a particular supergroup, but I’ll need to discuss the logistics and feasibility of this with our stats team to see if they feel confident in implementing such a scheme (especially if they would need to do this in both Varsity and JV, with an expanded field next year likely to be at or over 300 teams, given a projected major nationwide outreach effort at the start of the 2016-17 season.

7. To clarify one last thing – no rules were adjusted at any point during the tournament, and the new bee scoring system was announced early last fall, in fact. The only thing I was uncertain of at one point, was whether, with respect to the intergroup playoffs, the 1st seed played the 5th or the 4th. I knew I had posted this on our website, but with the website inaccessible, I was about 80% sure I had said 1st plays 5th, and then announced that since I couldn’t get online, that would just be the way it would work regardless of what the website said. Note that I announced that at the start of the evening pools before anyone knew who would be 1st, 5th, or even if we’d end up in this situation (though in the event, we did). Though as it turned out, the website did say 1st played 5th and 2nd, 3rd, and 4th played in the other, so that was as it should have been.
To #6: I support only taking two teams from the morning. At some point, a field has to be cut down. If the first cut had been 6->3, then there would've been 81 (and thus actually 84) teams in varsity competing for 24 slots in the evening, and some group runner-ups would not have advanced.
Reviewing last year's stat files: In varsity, there were 13 afternoon groups in the top tier. The morning third place teams finished 5th and 6th in nine of those 13 groups. In only three groups did a morning third place team finish third, and no morning third place team made the top 32. In JV, a few morning third place teams made third place in their afternoon groups (but only two from each afternoon group advanced to the evening stage). Cutting these teams at lunchtime gives more room for the top teams to compete against each other in the afternoon.
It is definitely possible for the morning -> afternoon wild cards to come from any supergroup, because that is how it was arranged last year. To do otherwise is to unambiguously benefit certain teams over others based on which morning group they are drawn into. I am wary of expanding the field by another 36+ teams, simply from a basis of competent staff and physical space.

To #7: I think the only discrepancy between your evening announcement and what was on the website was what would happen in the highly unlikely case of five groups going 3/1/1/1 (thus having five one-win runner-ups competing for one spot). But there is a significant issue with comparing records from wholly separate groups, and I think I would be in favor of all runner-ups having access to the evening -> Sunday wild cards.
Jon Pinyan
Coach, Bergen County Academies (NJ); former player for BCA (2000-03) and WUSTL (2003-07)
HSQB forum mod, PACE member
Stat director for: NSC '13-'15, '17; ACF '14, '17, '19; NHBB '13-'15; NASAT '11

"A [...] wizard who controls the weather" - Jerry Vinokurov
User avatar
Bosa of York
Rikku
Posts: 362
Joined: Sun Jun 15, 2014 9:16 am

Re: Discuss 2016 NHBB Nationals Here

Post by Bosa of York »

Great Bustard wrote:(as it was, High Tech lost in the 1/4s last year to Bethlehem)
High Tech beat us and then lost to LASA.
Eric Wolfsberg
Bethlehem Central High School 2016
University of Delaware 2020
Stanford 2025 or whatever
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

What the what, Dave Madden?
Great Bustard wrote:2. Very briefly, the draft was instituted last year after two specific instances showed me that simply having the 1 seed play the 8 seed and 2nd seed play the 7th seed, etc. off of evening pool records then pts was suboptimal. First, I remember Rohan from High Tech coming up to me and being delighted that they were the 6 seed rather than the 5 seed last year, since that meant they were in the half of the bracket away from Saratoga (as it was, High Tech lost in the 1/4s last year to Bethlehem). Secondly, it seemed odd that, as we were going to a 6 pool format on Saturday, the second place teams (who normally would be the 7 or 8 seed) could easily be of the sort like DCC or IMSA last year, which would be unfair to the 1 or 2 seed, if the 5 or 6 seeds were weaker, as per a general consensus. Hence the draft. Thus as some people have said, there’s nothing inherently wrong with the draft, and it’s existence is there not for some gimmicky add on, but rather to ensure maximal fairness (in, admittedly, a novel way). The draft is here to stay – it works.
See, these are all (potentially) legitimate problems and instead of seriously thinking about them, you went: “ah, I know! A draft!”. What!?

The well-accepted principle of “one loss should not eliminate you from contention” applies here, distinctly from the prelims. If your format is so bad that it creates these kind of incentives, the appropriate measure is not to float ridiculous solutions that don’t fix much (someone is still on the “Saratoga” side of the bracket!). The appropriate measure is to change the format to eliminate the problem! If NHBB weekend didn’t implement this weird Russian-nesting-doll of events, you could afford to do double elimination, or some other more fair format!
Great Bustard wrote:3. On the matter of three way games. Note that this can occur in two instances – either when the evening group has 3 teams at 2-1 (intragroup), or when across the pools (this was only the case in Varsity this year), there are either 6, 5, or 3 teams with 2-1 records (i.e. intergroup). On the intragroup, that’s there because there was a brouhaha last year when 3 teams went 2-1 in one group and the winner advanced solely on points. Now, with having the pool winners advance automatically, but allowing second place 2-1 teams two slots on Sunday from the intergroup, one could argue that we’d be better just having a 2 way match between the 2nd and 3rd on pts. in a group with 3 teams at 2-1, and letting the winner through. But that seems to skirt the whole reason as to why this was controversial last year. And what if the 3 teams shake out with pt totals of 900, 890, and 600 (which is possible)?

This is what I’m getting at when I’m arguing that this tournament was very, very close to Pareto optimization. We could tinker with things further, like with this way of resolving 3 teams at 2-1 (or, say, by having the 2 and 3 seeds meet, and then play the winner, though that still largely negates the value of wins, and adds further layers to what is already happening late on Saturday), but honestly, I think that a 3 way match is the fairest way to resolve this. Win your matches – and it won’t be an issue…

Same thing in the intergroup. If all 6 second place teams went 2-1, then how do we get just two teams? A two tiered playoff in the intergroup? That again seems to value points very highly at the expense of wins (plus it would make things go on that much longer). And with 5 teams, as was the case this year, we could have the 1st team on pts advance automatically, while playing off the 2nd spot in a 2 tiered format, but again, that seems to put an inordinately high value on the point distinction between 1st and 2nd. As is, the first place team has it easier by only competing against one team. Maybe that should be the 4 or even the 3 seed; on that, I’m a bit uncertain. But there’s nothing that I see inherently wrong in having 3 teams play at once – at most, it’s 12 players on the buzzer, which isn’t that different from 10 players on the buzzer which is common enough in various Bee finals.
Having maintained a wide berth from all history bowl events in the past 5 years, I don’t even know what half of this means – but that doesn’t matter because I do know (and you should know) that three-way matches are not fair. Quizbowl (and National History Bowl) is structured to answer the question: which of these two teams is best? This completely falls apart when you add a third team! It definitely falls apart when the third team is added and you play a match on ten questions (unbelievable). The problem with a three-way match has everything to do with the way teams’s knowledge matches up and nothing to do with how many people are on the buzzer. (there are various hypotheticals I could present, but they are all a bit degenerate so, c'mon, just think about it!) The idea of “win your matches and it won’t be an issue” is baffling: by the very fact that you had to do it this year, you know it will continue to be an issue! The correct thing to do in such a scenario is to make sure you are breaking ties fairly.

The correct way to do this, which you would know if you asked Andrew Feist (one of the foremost logistical minds in quizbowl and probably the only reason NHBB doesn’t completely fall apart every year), is to seed the teams by PPG (or a team-independent measure, as needed) and have the requisite 3-way tiebreak for two spots, which is: 1 v 2 (winner is in); loser of [1 v 2] v 3 (winner is in). Since the tiebreak is on 10 questions, this should take, like, 17 minutes at most.

re: bolding (emphasis mine). As someone who runs many good tournaments and also attends many well-run and fair tournaments, your continued statements that NHBB runs in some sort of Pareto optimal manner is pretty offensive. You are nowhere close, and will likely never be very close, to "Pareto optimal".
Cody Voight, VCU ’14.
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

Considering that we're never going to run Bowl matches on Friday, since teams need that for a travel day (the same reason we're not going to run the most popular of the individual events, the NH Bee on a school day too), the only way to have more time for the Bowl playoffs would be to scrap the Bee entirely on Sunday. That's never going to happen for all sorts of obvious reasons, so that argument is a red herring. And the way you've characterized my thought process on the draft is comically absurd - I've obviously showed that a lot of thought went into it, not that I just threw up my hands and said "oh golly gee, doesn't a draft sound like fun!"
The notion of a ten question tiebreak isn't unbelievable; it's basically fine, though we may add two questions to it next year. Does this inject more variability than a half packet would? Sure, but again, keep in mind that NHBB has far less variability than standard 20 / 20 quiz bowl because it's single subject, and because things don't swing as wildly with 30 point bonuses. And the ultimate point of NHBB is to find the best history team in the country, not the best of two teams - if a three team match accomplishes that goal in a better way, then so be it.
Quite frankly - the opinions of those who have always kept a wide berth of NHBB (except to critique it, and usually unfairly at that) are not nearly as important to me as those of the top teams and coaches who play NHBB. Ultimately, this is their tournament and I will touch base with them personally before considering the intergroup playoff format for next year, and take their views into account.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
User avatar
gimmedatguudsuccrose
Wakka
Posts: 163
Joined: Thu Jun 13, 2013 7:17 pm
Location: Boston

Re: Discuss 2016 NHBB Nationals Here

Post by gimmedatguudsuccrose »

Basil II wrote:
• Make the question worth 4 points for super-power, 3 points for power, and 2 points for the end of the question and after the question has finished. My idea with this is that, for most questions being played at the playoff level, getting the question right before "for the point" requires about the same knowledge as getting the question after it has finished, so someone who buzzes in a word before the question finishes really shouldn’t get an advantage over someone who buzzes in just after the last word. This also helps the transition into this format, because it’s the same as the 4th quarter of bowl matches.
• Reduce all penalties to one point, regardless of whether the neg came during or after the question. The idea that someone attempting to make an educated guess during the question gets penalized more than someone attempting to do that after the question seems a bit weird to me. Whoever buzzes in during the question must have at least some knowledge relative to the question in order to make the guess, so they shouldn't be more penalized then someone who waits till the end of the question and gets it wrong.
Nate714 wrote:
I personally agree with this. Many of the times during USHB and NHB Playoffs, I took educated guesses based on the question. I felt that since I had relative knowledge on the subject being asked, I shouldn't be negged more than the person who waits until the final "obvious" clue to buzz in. However, I feel like since this is the first year that NHBB did this, they will only improve on scoring from here on out and will work with the NHBB and Quiz Bowl Community in order to develop the scoring system into something that is fair for everyone.
thesouthindian wrote:
Although I did not participate in the Nationals (I was LITERALLY as far away as one could be), here are my two-cents based on discussing it with myself and my friends:

1. I personally feel that it is not a good system. One of the ways that you win the competition is by guessing, and this really doesn't encourage that. Hence, the very nature of the Bee changes from rewarding logical deduction and educated guesses to rewarding brute 'mugging-up' and regurgitation of facts, which was not as emphasised as before.
bluejay123 wrote:I really enjoyed the playoffs point system-- I felt that it rewarded knowledge in the form of the whole sequential point system thing. Although, I felt that the neg 1 at the end of the question (when the moderator had finished) was a bit harsh and discouraged educated guessing...

...

Perhaps I'm missing something, but I was under the impression that quizbowl was about buzzing when you know the answer, not buzzing when you have an 'educated guess.'
Last edited by gimmedatguudsuccrose on Mon May 02, 2016 12:13 pm, edited 1 time in total.
χ Smith
FHS '15
Chicago '19
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

Great Bustard wrote:Considering that we're never going to run Bowl matches on Friday, since teams need that for a travel day (the same reason we're not going to run the most popular of the individual events, the NH Bee on a school day too), the only way to have more time for the Bowl playoffs would be to scrap the Bee entirely on Sunday. That's never going to happen for all sorts of obvious reasons, so that argument is a red herring. And the way you've characterized my thought process on the draft is comically absurd - I've obviously showed that a lot of thought went into it, not that I just threw up my hands and said "oh golly gee, doesn't a draft sound like fun!"
No, I've characterized your thought process as completely lacking in any sort of serious consideration for what the actual problem is with the format (which is that these problems are intrinsic to single-elim).
Great Bustard wrote:The notion of a ten question tiebreak isn't unbelievable; it's basically fine, though we may add two questions to it next year. Does this inject more variability than a half packet would? Sure, but again, keep in mind that NHBB has far less variability than standard 20 / 20 quiz bowl because it's single subject, and because things don't swing as wildly with 30 point bonuses. And the ultimate point of NHBB is to find the best history team in the country, not the best of two teams - if a three team match accomplishes that goal in a better way, then so be it.
A 10 question tiebreak between two teams can be fine, sure. Three teams? Not so much. (also: the whole point of a 30-point bonus is to reward knowledge -- that's not a "wild swing", that's getting points for what you know!)
Great Bustard wrote:Quite frankly - the opinions of those who have always kept a wide berth of NHBB (except to critique it, and usually unfairly at that) are not nearly as important to me as those of the top teams and coaches who play NHBB. Ultimately, this is their tournament and I will touch base with them personally before considering the intergroup playoff format for next year, and take their views into account.
Most of my criticisms (especially these) are perfectly on point, but because you live in this weird fantasy universe (post still correct to this day), you will never concede that what you are doing is definitely not the best way to do things. As someone who has expended a lot of effort to "do things the right way" at the tournaments I've been involved in, I will always maintain a wide berth from NHBB, an organization that semi-randomly decides to eschew fair practices and continues to display a complete disregard for the concept of "good logistics". I recognize bad practices when I see them, and I don't wish to be involved in any way, shape, or form. (note: I have staffed an NHBB event and used to help produce NHBB sets. after the first NHBB Nats I no longer wished to staff anything and thankfully am completely out of the game of producing NHBB sets.)
Last edited by Cody on Mon May 02, 2016 12:15 pm, edited 2 times in total.
Cody Voight, VCU ’14.
User avatar
Kouign Amann
Forums Staff: Moderator
Posts: 1188
Joined: Sat Dec 06, 2008 12:44 am
Location: Jersey City, NJ

Re: Discuss 2016 NHBB Nationals Here

Post by Kouign Amann »

Great Bustard wrote:Quite frankly - the opinions of those who have always kept a wide berth of NHBB (except to critique it, and usually unfairly at that) are not nearly as important to me as those of the top teams and coaches who play NHBB.
I mean, do you ever think that maybe you could get more people into that latter group by putting any effort at all into engaging with the former? Cody and I and others realize that NHBB is big and has a lot of money and momentum behind it and is here to stay, so our critiques are aimed towards trying to make it better. But that's so ridiculously hard when your response is literally "well, my newly invented, demonstrably unfair, It's Crackademic-style tiebreaker format isn't really a problem if there are no ties." Like, do you even hear yourself here? It's so hard, in fact, that I've mostly given up and just check in on the post-nats thread once a year to shake my head ruefully. The reason it's so hard is, as it always has been, that you live in a fantasy universe.

You may claim that your loyal customers would never critique you in this way, but you have to keep in mind that your continued inhabitation of this fantasy universe has in fact lost you customers over the years. I was one. If you think back to the 2010-11 year, you may remember that I had multiple hours-long meetings with you to discuss your ideas, I ended up agreeing to host one of your regional tournaments, and I did everything I could to spread the word and get other people excited about this format. I thought, "well, this guy has some wacky and bad ideas, but he's new on the scene and fresh and enthusiastic, and I'm sure he'll get things straightened out over time." But as a year went by, and then another, and as it became increasingly apparent that you are 100% dedicated to doing things exclusively in your own way despite what anyone else suggests, you lost me. St. Anselm's didn't attend 2012 NHB. You lose customers over this stuff.
Aidan Mehigan
St. Anselm's Abbey School '12
Columbia University '16 | University of Oxford '17 | UPenn GSE '19
User avatar
Fado Alexandrino
Yuna
Posts: 834
Joined: Sat Jun 12, 2010 8:46 pm
Location: Farhaven, Ontario

Re: Discuss 2016 NHBB Nationals Here

Post by Fado Alexandrino »

The NAQT card system + playoff format seems to work for very large fields where repeated rebracketing would take too long.
Joe Su, OCT
Lisgar 2012, McGill 2015, McGill 2019, Queen's 2020
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

Honestly, I would prefer to run double elimination, but my understanding is that that requires more rounds than we have time for. If we started playoff rounds at 7:30pm on Saturday, we could run 5 matches on Saturday evening, and 4 more on Sunday. That's the absolute max. So, given that we have space for 9 rounds, is it possible to run a tournament where we have the 1-16 seeds playing double elimination, and the 17-32 seeds playing single elimination? If it is, I'll seriously consider that for next year.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
Angry Babies in Love
Yuna
Posts: 819
Joined: Wed Dec 19, 2007 10:09 pm
Location: MD//DC

Re: Discuss 2016 NHBB Nationals Here

Post by Angry Babies in Love »

Kouign Amann wrote:demonstrably unfair
I have not seen this idea fleshed out. How is putting three teams against each other any different than putting eight individuals against each other? There's no way to demonstrate whether or not this is more fair or not; if you know more and are better at History Bowl, there's no inherent reason to believe you will not win this tiebreaker. That's the definition of fairness.
Raynell Cooper
Arcadia ES '04
Richard Montgomery HS '11
George Washington University '15
University of Maryland, College Park '17
Hella things, National History Bee and Bowl
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

Great Bustard wrote:Honestly, I would prefer to run double elimination, but my understanding is that that requires more rounds than we have time for. If we started playoff rounds at 7:30pm on Saturday, we could run 5 matches on Saturday evening, and 4 more on Sunday. That's the absolute max. So, given that we have space for 9 rounds, is it possible to run a tournament where we have the 1-16 seeds playing double elimination, and the 17-32 seeds playing single elimination? If it is, I'll seriously consider that for next year.
It depends on how you do the double elim -- the way NAQT does it is that the day starts with some teams seeded into the loser's bracket (one loss and you're out) and some teams in the winner's bracket (two losses and you're out); this eliminates one round. If you start your double-elim bracket this way with 32 teams, it takes at most 10 rounds (incl. the possibility of a loser beating the undefeated team in the first game of an advantaged final). If you start with 16 teams this way, it takes at most 8 rounds; 8 teams, 6 rounds. (assuming my recollection isn't wrong).

A seeded double-elim with no one starting in the loser's bracket takes one more round than the above.
Cody Voight, VCU ’14.
User avatar
Fado Alexandrino
Yuna
Posts: 834
Joined: Sat Jun 12, 2010 8:46 pm
Location: Farhaven, Ontario

Re: Discuss 2016 NHBB Nationals Here

Post by Fado Alexandrino »

Great Bustard wrote:Honestly, I would prefer to run double elimination, but my understanding is that that requires more rounds than we have time for. If we started playoff rounds at 7:30pm on Saturday, we could run 5 matches on Saturday evening, and 4 more on Sunday. That's the absolute max. So, given that we have space for 9 rounds, is it possible to run a tournament where we have the 1-16 seeds playing double elimination, and the 17-32 seeds playing single elimination? If it is, I'll seriously consider that for next year.
You can do 1-32 winners bracket 33-64 losers bracket in a 9 round, NAQT style playoff (in round 6, 1 has a bye, 2-7 3-6 4-5 play to determine the losers in the final 4; r7 is final 4, r8-9 is either the advantaged final or the two stage final)
Joe Su, OCT
Lisgar 2012, McGill 2015, McGill 2019, Queen's 2020
User avatar
John Ketzkorn
Wakka
Posts: 197
Joined: Tue Oct 08, 2013 2:54 pm

Re: Discuss 2016 NHBB Nationals Here

Post by John Ketzkorn »

Angry Babies in Love wrote:
Kouign Amann wrote:demonstrably unfair
I have not seen this idea fleshed out. How is putting three teams against each other any different than putting eight individuals against each other? There's no way to demonstrate whether or not this is more fair or not; if you know more and are better at History Bowl, there's no inherent reason to believe you will not win this tiebreaker. That's the definition of fairness.
There is always a probability of a team "catching a packet". If you put one team up against two, then the chances of the best team winning go down because each of the other two teams can catch packets they know just as well, if not better, than the best team. The probability of this goes up even more when there are only 10 questions in said packet.

We should be awarding the team with the most knowledge by limiting this variability.
Last edited by John Ketzkorn on Mon May 02, 2016 1:22 pm, edited 1 time in total.
Michael Etzkorn
Illinois Mathematics and Science Academy '16
UIUC '21
User avatar
Skepticism and Animal Feed
Auron
Posts: 3238
Joined: Sat Oct 30, 2004 11:47 pm
Location: Arlington, VA

Re: Discuss 2016 NHBB Nationals Here

Post by Skepticism and Animal Feed »

Another argument against three-team games is that it penalizes teams of generalists by potentially exposing them to two sets of specialists at the same time.
Bruce
Harvard '10 / UChicago '07 / Roycemore School '04
ACF Member emeritus
My guide to using Wikipedia as a question source
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

Angry Babies in Love wrote:I have not seen this idea fleshed out. How is putting three teams against each other any different than putting eight individuals against each other? There's no way to demonstrate whether or not this is more fair or not; if you know more and are better at History Bowl, there's no inherent reason to believe you will not win this tiebreaker. That's the definition of fairness.
Didn't you play It's Academic many times? The problem with a 3-way match is inherent in the format: it's about how knowledge matches up. A normal quizbowl (and National History Bowl) match is about: which team has more knowledge? In a 3-way match, it's about: which team has the most knowledge not overlapping with the other two teams? This is not the same thing that's decided by a normal match and is why it's demonstrably unfair (especially when you've reduced your sample to 10 questions!).* Mike also makes a good point.

*there are many degenerate hypotheticals one could consider, which is why I have gone with a mere description above. But consider a team that always gets the European history on the first three line plus a smattering of other questions in the last three lines, vs a team that always gets American history on the first three lines plus a smattering of other questions in the last three lines, vs a team that gets every question within the last three lines. Who advances is very different from what would happen in a normal tiebreak.
Cody Voight, VCU ’14.
User avatar
Habitat_Against_Humanity
Rikku
Posts: 476
Joined: Sun Jan 21, 2007 8:51 pm
Location: Syracuse, NY

Re: Discuss 2016 NHBB Nationals Here

Post by Habitat_Against_Humanity »

I don't really have a dog in this fight, but I can't think of any sort of non-Calvinball competition in which the number of teams would vary between games. In what universe do three team matches make any sense when literally every other single match participants will ever play in the history of forever will be between two teams? This isn't deciding who's best at the competition; this is a whole different game. It's gimmicky at best and terribly detrimental at worst.
Rachel
UChicago 09
tabstop
Wakka
Posts: 122
Joined: Tue Mar 20, 2007 10:48 am
Location: NNVA

Re: Discuss 2016 NHBB Nationals Here

Post by tabstop »

This year the final stages of the playoffs was six rounds + the tiebreaker thing; 16 winners + 32 losers would take eight rounds, which is going to be an extra packet = an extra 40 minutes at the end of Saturday night, plus another pile of questions. (It might be possible to do 24 straight-up double-elim in eight rounds, but I'd have to draw it out more fully.) Fifteen rounds on a Saturday would be quite a haul.

When I saw the joy on David's face when he realized he could do a three-team game, I knew we were going to have to do this tiebreaker at least once. I was hoping this year would get it out of his system. But yes, it's easy enough for a Condorcet winner to lose a three-team playoff, and I think we'd rather have the Condorcet winner advance (even if I did spend a long time this year trying to explain to a parent why head-to-head isn't a good tiebreaker). I don't know that how often this would happen in practice, but setting up a system that is built to give occasional wrong results isn't a great start.

In general, I would prefer that wild cards should come from everywhere and not just one division (this year there was a symmetry with one division having two fewer qualifiers and then having the two wild cards come from that division, but that's not a symmetry I particularly care about).
Andrew Feist
User avatar
heterodyne
Rikku
Posts: 427
Joined: Tue Jun 26, 2012 9:47 am

Re: Discuss 2016 NHBB Nationals Here

Post by heterodyne »

tabstop wrote: When I saw the joy on David's face when he realized he could do a three-team game, I knew we were going to have to do this tiebreaker at least once. I was hoping this year would get it out of his system. But yes, it's easy enough for a Condorcet winner to lose a three-team playoff, and I think we'd rather have the Condorcet winner advance (even if I did spend a long time this year trying to explain to a parent why head-to-head isn't a good tiebreaker). I don't know that how often this would happen in practice, but setting up a system that is built to give occasional wrong results isn't a great start.
I'd hate to be the team that was screwed over because Dave Madden was too enraptured with the idea of a three-team game for anyone to convince him that it was a horrible idea.
Alston [Montgomery] Boyd
Bloomington High School '15
UChicago '19
UChicago Divinity '21
they
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

A few things here - first, is that I think the scope of this problem is being largely overstated. I'm not saying I'm not interested in a fair way of resolving the tournament (if I weren't I wouldn't have changed the tournament format twice in the past two years to get away from single elim, and then address the issue of a pool winner advancing solely on points with no chance for a second place team to advance). NHBB, with a single subject focus and a playoff format that accounts for these things is going to routinely have the cream rise to the top. Don't forget that in 6 years, we've had five 1 seeds and 1 two seed (who easily could have been a one seed had I had more regional results from LASA to go off this year) win, plus, we had 7 of the 8 top seeds make it to Sunday this year, with the 8th being another hard to seed team (who in retrospect should have been seeded in the top 8).

If people can realize those things and not claim that I (still) live in some fantasy universe and blithely ignore the advice of the community, then I'll be a lot more interested in continuing this discussion with them. More to the point, here are some other things to consider:

-One other reason we moved to pool play in the evening was to give the top teams more games against top teams. While getting a champion is the ultimate goal of the tournament, keep in mind that nearly all teams come to Nationals with other prerogatives - one of which is competing in as many games as they can (esp. for the best teams in the country). Realistically, even in a year where the competition was relatively even like this one, only the top 8-10 teams or so had any shot of winning Varsity. The pool play, though, gives all of the top 24 teams at least 3 additional games.

-I'm still waiting on more players and coaches to comment, whether here, or once I get a chance to ask them later this spring / before next fall. Ultimately, if the top teams are deadset opposed to 3-way matches, then, as I've mentioned, we'll find another way. But I still am not really convinced that any of the alternatives here that have been presented so far are fairer, with the possible exception of double elim.

-Though having said that, I'm a bit confused as to how double elim with 24 (or ideally, 32 teams) is possible within 8 rounds. Wouldn't there need to be some sort of workaround? Even just with 16 teams:

Round 1 - All 16 play
Round 2 - Top 8 play in winners' bracket, bottom 8 in losers' (after which 12 remain)
Round 3 - Losers of Top 8 (4 teams) from round 2 play winners of bottom 8 (after which 4 remain in upper, and four remain in lower).
Round 4 - 4 teams in winners' bracket play, 4 teams in losers' bracket play (after which 6 remain)
Round 5 - Losers of Top 4 in winners' bracket play winners' of 4 teams from losers' bracket (after which 4 remain)
Round 6 - 2 teams in winners' bracket play, 2 teams in losers' bracket play (after which 3 remain)
Round 7 - Loser of winners' bracket plays winner of losers' bracket (after which 2 remain)
Round 8 - Adv final game 1
Round 9 - Adv final game 2 (if necessary)

Am I missing something here?

-I understand the argument against having 3 teams in the sense that a team of specialists is then going to have more of an advantage then a team of generalists. A 3 way match is also going to reward more aggressive buzzing, and in a sense, a different style of play. I do concede the point that this isn't ideal in the sense that the rest of the tournament is not set up to reward that style of competition, so it's a bit unfair to expect teams to have to play under these rules just in the tiebreak. It's certainly a valid point, and if this is what teams and coaches think is the more salient point, then I'll look to adjust.
Having said that, there's a counter argument here, which is that if wins are really the deciding factor, then isn't it a bit artificial to have 2 teams who were 2-1 in a group play for the right to play a 3rd team who was 2-1, especially, say, if the 1 seed has only 10 more points than the 2 seed, but the 2 seed has 200 more points than the 3 seed? And if we did this, we'd need to have two more levels of tiebreaks (thus pushing things even later on Saturday, not to mention, it's another 24 tossups that need to get written - and that's keeping the tiebreaks at 10 questions, which, admittedly is a low number). Another option would be to extend the tiebreak length to 15 questions in the 3 way matches - more questions there would help ensure the top teams win.

At some level, this whole situation parallels the debate about whether a tournament's questions should get harder as the tournament continues. I fully understand the rationale behind why in normal quizbowl circumstances this is not considered a desirable practice. But NHBB in general, and in particular at Nationals has the widest variability of teams competing on the same set, and at the same time the most restrictive set of material to ask about with a huge number of questions needed across the various events. Sure, we could axe the Bee / have a much more stringent qualifying threshold / have fewer rounds in which all teams compete on Saturday, but there are very valid reasons for not doing any of these things. Meanwhile, harder questions towards the end of the tournament do a better job at differentiating who the best team really is. That's why we don't have finals level difficulty during the day on Saturday (which would lead to a bad tournament experience for the lower third of the field, at least), and why we don't have Bowl prelims difficulty in the finals on Sunday (which would lead to buzzer races galore). And at the same time, the degree to which the tournament would be made fairer by altering along these lines is incremental - in a sense, the tournament is already less upset prone than HSNCT, for example, because of its focus on one subject, and the lack of 30 point bonuses on a topic that one team might find to be in its wheelhouse (or not).

That's basically where I'm at with this at the moment. I'm certainly comfortable with Westview, Naperville North, and Auburn having advanced out of the tiebreakers this year. Of the 3 teams that contested the intragroup tiebreaker this year (Stevenson with Westview and Auburn), Stevenson had 20 points more than Auburn Auburn in the evening (Westview then had 110 more than Stevenson), but in the afternoon, Westview was 5-0, 1730 and Auburn was 4-1, 1680 while Stevenson was 3-2, 1490. Then, among the 5 teams that played in the intergroup tiebreaker, Team Kronos, High Tech, and Naperville North played in the 2-3-4 seeded game with Naperville North winning while Westview played as the 1 seed and defeated the 5th best 2-1 team in 2nd place in evening group play, which was Centennial. In each case, the top seed won - Naperville North had 890 pts. in the evening to Kronos's 780 and High Tech's 750. Westview had 980 points to Centennial's 690. Or looked at from the afternoon, Westview was 5-0, 1730 to Centennial 3-2, 1320. And Naperville North went 5-0, 1770 compared to Kronos at 3-2, 1290 and High Tech at 4-1, 1590.

Sure, there was some variability here - it wasn't inevitable, especially on just 10 questions, that the top seeds would advance (one of the reasons, I'm leaning towards upping this to 15 questions next year, if we stick with this). But the fact that they did isn't a coincidence either, and had they won their pool at 3-0, they wouldn't have found themselves in the tiebreaker position in the first place. Again, the point of a quiz bowl tournament is not to run on a format where upsets never happen, and the best team will always, 100% of the time find themselves in the winning position. Compared with almost every other competitive activity I can think of (basically every Olympic sport, Scrabble, baseball, football, etc.), the odds that the top team advances farther are better in quiz bowl and especially so in NHBB given the single subject focus, and even more so at NHBB Nationals with the format we had this year. And at the same time, I have other prerogatives to balance, including the time available, the interests of teams who are not going to be in contention for the national title, and the amount of questions that need to get written, to name just 3. I hope that everyone who thinks I'm whimsical for the sake of whimsy, ignorant of good practices, uninterested in improving the format, and am blithely screwing teams over can take a step back and consider those points.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
User avatar
jonpin
Auron
Posts: 2266
Joined: Wed Feb 04, 2004 6:45 pm
Location: BCA NJ / WUSTL MO / Hackensack NJ

Re: Discuss 2016 NHBB Nationals Here

Post by jonpin »

tabstop wrote:This year the final stages of the playoffs was six rounds + the tiebreaker thing; 16 winners + 32 losers would take eight rounds, which is going to be an extra packet = an extra 40 minutes at the end of Saturday night, plus another pile of questions. (It might be possible to do 24 straight-up double-elim in eight rounds, but I'd have to draw it out more fully.) Fifteen rounds on a Saturday would be quite a haul.
An NAQT-style double elimination system that starts with 16 + 16 will usually take 7 rounds (but has the possibility of 8). What I mean by that is that all teams play all rounds (until the very end), rather than a traditional DE where the winners' bracket skips rounds throughout.

I think a reasonable system would be a 16+16 DE playoff for varsity and a 16-team 4x4 -> 4-team SE playoff for JV. Varsity plays 4 rounds Saturday night, JV plays its group stage and tiebreakers to win the group. Sunday, the first two rounds of bowl playoff are the Round of 6 and the semifinal(s) for varsity, simultaneous with the semifinal and final for JV; then the final (which might be a 1-game or might be an advantage final).
I don't think that will take longer than it did this year on Saturday, because there was a mighty long time of standing around waiting for everyone to figure out what was going on before the tiebreakers. And in terms of Sunday, yes there would be the possibility of 4 rounds instead of 3, but that delay is shortened by the fact that you can run the second final IMMEDIATELY after the first final (as in, two minutes for coaches to talk or kids to run to the bathroom). You could say 12:45 and 1:20 for the first two start times, and 2:05 for the varsity final start--even with a 2-game final, it would end by about 3. The biggest impediment to scheduling on Sunday is what it has been every year: KEEP. THINGS. MOVING.
I don't know when exactly the first knockout round of the bowl started, but I reported to the staff room at 1pm for a second knockout round match scheduled to start at 1:15, and by 1:30, I was wondering whether I was in fact at the right location because nothing was ready to go yet. I am not certain, but I think someone said around this time that the packets were being printed--I really hope I misheard that! When I left at the end of the Bee semifinals, I think the tournament was down to only about 20 minutes behind schedule, but that's because there are always such lengthy times booked in the schedule (35 minutes for a round of the bee playoffs which is 40 tossups, many of which will get powered).
When I saw the joy on David's face when he realized he could do a three-team game, I knew we were going to have to do this tiebreaker at least once. I was hoping this year would get it out of his system.
I'm a little terrified that the joke my team and I had of "Madden went out and got a 3-team buzzer system and he'll be damned if it's gonna go unused!" was not too far away from fact.
In general, I would prefer that wild cards should come from everywhere and not just one division (this year there was a symmetry with one division having two fewer qualifiers and then having the two wild cards come from that division, but that's not a symmetry I particularly care about).
Didn't that division have two fewer qualifiers because it had one fewer group in the first place? Disregarding wild cards, both supergroups were qualifying 1/3 of all teams.
Jon Pinyan
Coach, Bergen County Academies (NJ); former player for BCA (2000-03) and WUSTL (2003-07)
HSQB forum mod, PACE member
Stat director for: NSC '13-'15, '17; ACF '14, '17, '19; NHBB '13-'15; NASAT '11

"A [...] wizard who controls the weather" - Jerry Vinokurov
User avatar
Stained Diviner
Auron
Posts: 5085
Joined: Sun Jun 13, 2004 6:08 am
Location: Chicagoland
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Stained Diviner »

Round 1 - All 16 play
Round 2 - Top 8 play in winners' bracket, bottom 8 in losers' (after which 12 remain)
Round 3 - Losers of Top 8 (4 teams) from round 2 play winners of bottom 8, and 4 undefeated teams play each other (after which 2 remain in upper, and 6 remain in lower)
Round 4 - 2 teams in winners' bracket play, 6 teams in losers' bracket play (after which 1 remains in upper and 4 in lower)
Round 5 - Team in winner bracket gets bye, 4 in lower play
Round 6 - Team in winner bracket gets another bye, 2 in lower play
Round 7 - Adv final game 1
Round 8 - Adv final game 2 (if necessary)

You only add one round if you go to 24 or 32 teams.
David Reinstein
Head Writer and Editor for Scobol Solo, Masonics, and IESA; TD for Scobol Solo and Reinstein Varsity; IHSSBCA Board Member; IHSSBCA Chair (2004-2014); PACE President (2016-2018)
User avatar
Fado Alexandrino
Yuna
Posts: 834
Joined: Sat Jun 12, 2010 8:46 pm
Location: Farhaven, Ontario

Re: Discuss 2016 NHBB Nationals Here

Post by Fado Alexandrino »

Here is a 32 winner 32 loser 9 round bracket, basically the same way NAQT does it, but I didn't include any round 1 or round 2 byes, or a "round 0" like you see at HSNCT.

ROUND: #Winners | #Losers
Winners remain winners/Winners become losers | Losers stay alive/Losers eliminated

1: 32 | 32
16/16 | 16/16

2: 16 | 32
8/8 | 16/16

3: 8 | 24
4/4 | 12/12

4: 4 | 16
2/2 | 8/8

5: 2 | 10
1/1 | 5 /5

6: 1 | 6
1/0 | 3/3
The 1 seed, the remaining undefeated team, has a bye in round 6. Six one-loss teams play each other to determine who will join #1 in the final four.

7: Four teams remain. 1v4, 2v3 play in seminfinals.

8-9:
If 1 wins, then 1 plays W(2v3) with the advantage in a 2 game final and 4 plays L(2v3) in a one-game match for third place. The 2012 HSNCT finished this way.
If 1 loses, then 4 plays W(2v3) to see who plays 1 in a one game, winner takes all final. The loser of this play-in game is third, while L(2v3) takes fourth. The 2013-2015 HSNCTs finished this way.

Bookkeeping notes: The 2007 HSNCT had 61 teams at 6-4 or better out of 160, the 2008 HSNCT had 68/176. Currently, HSNCT playoffs take around 100 teams and require 10 rounds, though the first round is a play-in game to determine the 64 teams, 9 round affair I described. (Well not quite because some winner-bracket HSNCT teams have two Sunday morning byes so numbers are a bit off).

You can easily fit 15 rounds on Saturday, and play the final 4 rounds Sunday. Last year, despite HSNCT hitting TWO snafus Saturday afternoon, it still finished 16 Saturday rounds at a reasonable hour and with ample time for side events. Now how to incorporate byes I don't know, since you have literally no time due to the bee. Better organization and better readers will also help.
Joe Su, OCT
Lisgar 2012, McGill 2015, McGill 2019, Queen's 2020
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

jonpin wrote: I think a reasonable system would be a 16+16 DE playoff for varsity and a 16-team 4x4 -> 4-team SE playoff for JV. Varsity plays 4 rounds Saturday night, JV plays its group stage and tiebreakers to win the group. Sunday, the first two rounds of bowl playoff are the Round of 6 and the semifinal(s) for varsity, simultaneous with the semifinal and final for JV; then the final (which might be a 1-game or might be an advantage final).
When you say 16 + 16, is that implying that 16 would start in a position where they need to lose twice to be eliminated, and 16 are in single elim position? If that's possible within 8 rounds, then I'll strongly consider that. As mentioned, teams that finish second in their afternoon group who go 4-1 have to make evening rounds, so that means I'm looking at a minimum of 24, and ideally, 32 teams that would be able to compete on Saturday evening. 4 rounds on Saturday evening, and 4 rounds on Sunday (esp. if the 4th is just immediate as you mention) is certainly workable.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
User avatar
jonpin
Auron
Posts: 2266
Joined: Wed Feb 04, 2004 6:45 pm
Location: BCA NJ / WUSTL MO / Hackensack NJ

Re: Discuss 2016 NHBB Nationals Here

Post by jonpin »

Great Bustard wrote:
jonpin wrote: I think a reasonable system would be a 16+16 DE playoff for varsity and a 16-team 4x4 -> 4-team SE playoff for JV. Varsity plays 4 rounds Saturday night, JV plays its group stage and tiebreakers to win the group. Sunday, the first two rounds of bowl playoff are the Round of 6 and the semifinal(s) for varsity, simultaneous with the semifinal and final for JV; then the final (which might be a 1-game or might be an advantage final).
When you say 16 + 16, is that implying that 16 would start in a position where they need to lose twice to be eliminated, and 16 are in single elim position? If that's possible within 8 rounds, then I'll strongly consider that. As mentioned, teams that finish second in their afternoon group who go 4-1 have to make evening rounds, so that means I'm looking at a minimum of 24, and ideally, 32 teams that would be able to compete on Saturday evening. 4 rounds on Saturday evening, and 4 rounds on Sunday (esp. if the 4th is just immediate as you mention) is certainly workable.
Yes. Top 16 seeds (which would include all group champions presumably) start in winners' bracket, next 16 start in losers' bracket (which would include all runner-ups).
Round 1: 16 teams in winners' bracket + 16 in losers' bracket. | JV groups play round one.
Round 2: 8 + 16 | JV groups play round two.
Round 3: 4 + 12 | JV groups play round three.
Round 4: 2 + 8 | JV groups break any ties. A two-way tie could just play a full game, a three-way tie would require something else (three-way matches should be avoided)
These four rounds in Varsity would be on a pre-defined bracket, and Saturday would end here. The possibility exists of reseeding or otherwise planning the final rounds overnight. My idea would be "the unbeaten team plays the lowest seeded team they didn't already play Saturday night".

Round 5: 1 + 5. One cross-bracket game and two losers' bracket games. | JV semifinals
JV final in the big room in Round 6. Meanwhile, in Varsity:
If the unbeaten team wins their first game on Sunday, the two other winners play a semifinal in Round 6, and then there's an advantage final in Rounds 7 and 8.
If the unbeaten team loses, then there are four teams, all with a loss. Two semifinals in Round 6, and then a one-game final in Round 7.
Jon Pinyan
Coach, Bergen County Academies (NJ); former player for BCA (2000-03) and WUSTL (2003-07)
HSQB forum mod, PACE member
Stat director for: NSC '13-'15, '17; ACF '14, '17, '19; NHBB '13-'15; NASAT '11

"A [...] wizard who controls the weather" - Jerry Vinokurov
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

Thanks for this. I'll need to mull this over, but I like this version a lot - I think it accomplishes a lot of what I'm trying to get at, and if it's significantly less controversial than the format this year, then so much the better.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
Scottietodd
Wakka
Posts: 122
Joined: Sun Nov 13, 2011 1:55 pm

Re: Discuss 2016 NHBB Nationals Here

Post by Scottietodd »

From a coach who loves the actual format of NHBB, and coming from a coach who has not had a team make it to the highest levels (in NHBB) very often yet, here are my two cents.
I would not like to play in the 3-way tie-break. I cannot cite any of the statistics, theories, etc... that have been made on here or even claim to understand all of them. My reason is simply this.

We do not play 3-way ties during the year...ever. If I have a team that actually makes it deep into an event, they are already going to be pretty tuned up and nervous. It seems a bit much to have them
stress about a format that they are unfamiliar with before a win or die match. There is a lot of, for lack of a better term, "muscle memory", to doing well in these events. You are familiar with the format, you know what comes next, and you have instincts for what you will need to do to win and how that changes after every correct or incorrect buzz. Now you've not only got to play but you have to try to think about how playing
against 2 teams at the same time will be different.
Todd Garrison
Glasgow High School
Coach - 2012 to present
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

Okay, one last post (barring the need to refute something or other) on this. Having considered the time we have to work with, the likely number of teams in the draw next year, and various logistical matters (e.g. a desire to have the JV finals run one round before the Varsity), we're going to do a 32 team double elimination next year in the Varsity with teams ranked 17-32 (who by definition, will have at least one loss in the afternoon rounds) going into Saturday evening in a single elimination position. Thanks to Jon Pinyan and David Reinstein for pointing out this is an option. Four of those rounds will run on Saturday evening, with the remaining 3 or 4 as needed running on Sunday. I'm eager to see how this goes, and I think this combined with the Saturday midday rebracket gives the Varsity division a nice format that's a bit of a hybrid of NSC and HSNCT.

For the JV, with the smaller field size, for now, we'll stick with 3 rounds of pool play on Saturday, and 2 rounds on Sunday. For 2017, at least, only 16 teams will play on Saturday evening, but this will allow all teams at 4-1 and in second place in the JV upper brackets on Saturday afternoon to advance. We can cap JV field size at 144 - and that's a very high end estimate anyway - and still take 48 into upper bracket and have 16 2nd and 1st place teams play Saturday evening. In all likelihood, any 4-1 team, even in 3rd place would advance too.

With 16 teams, though, only group winners would advance (so there's no intergroup playoff here at all). In a 2-1, 2-1, 1-2, 1-2 split, the two top teams would play a power-marked tiebreaker, which I'd like to be 12 tossups, up from 10 this year. In the dreaded (2-1) x 3 scenario, which, with just 4 pools is less than 50% likely to even come up (unlike the intergroup tiebreak, where a 3 way match is extremely likely to be needed), my preference is to still maintain the 3 way tiebreak, either at 12 or 15 power-marked questions. I have considered the arguments for and against this, and basically, I'm still most convinced by the fact that if wins are really what matter most, then having all 3 teams on the same record have a chance to play each other simultaneously is a fairer way to resolve this than a two stage tiebreaker, where the top team on points gets a bye in the first stage. I recognize that there are legitimate arguments against this - Todd's point that it's unfamiliar, and Cody's argument that it can privilege a team of specialists are valid. Though, on the other hand, the whole notion of a two stage tiebreak gives the top team on points an advantage that in my mind is so great that it outweighs these counterarguments. This is particularly true in a situation where the first and second seeds are considerably closer on points than the third seed is to the second seed. A match between 3 teams is still not fundamentally different aside from calling for somewhat more aggressive buzzing, especially on the part of the 3rd place team, but this sort of variance happens all the time (and to a far greater extent) in Bee play - the effect here is less than that.

So, that's my preference, but, having said that, this isn't a hill I'm prepared to die on if there is a clear preference among the JV teams who are most likely to finish among the top 12 at Nationals next year. We can leave the matter unresolved for now - it's not going to fundamentally alter a team's entire annual strategy - and then, roughly a month out of nationals next year, I can poll the top 15-20 JV teams who are coming to Nationals, and see what they prefer. They are the ones who will ultimately be playing this, and I am happy to let them have the decisive say. If there's no clear consensus, then we'll stick with the 3-way match, otherwise, I'm ok with going to a 2 stage tiebreaker.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

One last point: I find the level of vitriol in this thread rather bewildering since:
a) this is dealing with a relatively minor improvement (cf. this year's playoff format accomplishing exactly what it was meant to do)
b) these are all ideas I mooted last year in the post-Nationals thread then (largely to see what the community thought of them and get feedback) and no one had any objections at the time. See: http://www.hsquizbowl.org/forums/viewto ... &start=100
and
c) For the third year in a row, I'll be changing the playoff format based on community input that comes from good arguments.
That's all.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
bluejay123
Lulu
Posts: 51
Joined: Thu Apr 30, 2015 10:35 pm

Re: Discuss 2016 NHBB Nationals Here

Post by bluejay123 »

On the USHB stats, I noticed that Rohan Hegde's name was incorrectly entered as Hedge.
Jaya Alagar
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

As many people have tried to explain to you (including Andrew Feist!), you are wrong and a three-way tie breaker is intrinsically unfair. To continue to ignore people because you think it's more fair (despite the fact that you very clearly don't even understand it) is...quite something, let's leave it at that.

Also, I thought the whole thing with the three-way tie break is that it advances two teams? In that case, as I posted above, after the teams are ordered by PPG, 1 plays 2 and the loser plays 3. The winners of each match advance. Your conception of fairness is extremely skewed -- if wins are all that matters (according to you), why is this not fair? (the same argument applies to the three-way tiebreak for one spot -- a bye doesn't mean you don't have to win your game! Though there were only a couple of 3-way ties, I can point you to an instance at ACF Nationals where a 2/3 from a 3-way tiebreaker advanced. I can also point you to an example at a recent VCU high school tournament, where the 3-way was to an advantaged final. Maybe don't make grandiose claims about the fairness of tiebreaker formats when people that are much smarter than you devised them and have been using them for over a decade? The idea that your tiebreaker format is better than that used by NAQT, ACF, and PACE is ludicrous.)
Cody Voight, VCU ’14.
User avatar
jupiter
Lulu
Posts: 8
Joined: Mon May 27, 2013 11:26 pm
Location: Austin, TX

Re: Discuss 2016 NHBB Nationals Here

Post by jupiter »

Great Bustard wrote:We're going to do a 32 team double elimination next year in the Varsity with teams ranked 17-32 (who by definition, will have at least one loss in the afternoon rounds) going into Saturday evening in a single elimination position.
Does this mean that during the afternoon on Saturday next year there will be 16 (or 8) upper groups? Because if there isn't then you will have a weird cutoff on who is 1-16 and who is 17-32. It would work best if you had 16 groups and had the winners be 1-16 and the runner ups be 17-32. This is what I believe you are proposing, but it might mean you have to take the 3rd place team in some morning groups (to get to 16 afternoon groups) which you don't want to do. It all depends on how many teams you have next year. 8 afternoon groups would also work, with the top 2 from the group going into the top bracket and the middle two going into the lower bracket, but that would only work if the field size shrunk, and I know you don't want to do that.

Now if what I said is your plan, I would recommend re-seeding all the teams after the morning. We (LASA) were 5th in points and Naperville North was T-2nd but we ended up in the same afternoon group (there were 9 groups) because everything was pre-seeded; the winner or our morning group and the winner of Naperville's morning group would end up in the same afternoon group regardless of how well the winners did. Now that wasn't an issue this year as we both advanced and no losses carried over, but in this new system if something like that happens there are severe playoff implications. Yes re-seeding would take more time and could be more complicated, but it would make it much harder for teams to complain that they had a hard bracket.
John DiCarlo

Kealing '12
LASA '16
MIT '20
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: Discuss 2016 NHBB Nationals Here

Post by Cody »

additionally, your stated reasoning for preferring a 3-way tiebreaker for 1 spot in the specific case of 1 and 2 being separated by a marginal difference in PPG doesn't hold water for another reason, as well: if 1 and 2 are so evenly matched, then what effect does adding a 3rd team to the match have? Why, it is pretty much guaranteed to have a disproportionate impact on one of the two teams (instead of an equal impact) in terms of what questions the 3rd team gets as compared to what questions the 1 and 2 would get in a solo match!
Cody Voight, VCU ’14.
User avatar
jonpin
Auron
Posts: 2266
Joined: Wed Feb 04, 2004 6:45 pm
Location: BCA NJ / WUSTL MO / Hackensack NJ

Re: Discuss 2016 NHBB Nationals Here

Post by jonpin »

jupiter wrote:Now if what I said is your plan, I would recommend re-seeding all the teams after the morning. We (LASA) were 5th in points and Naperville North was T-2nd but we ended up in the same afternoon group (there were 9 groups) because everything was pre-seeded; the winner or our morning group and the winner of Naperville's morning group would end up in the same afternoon group regardless of how well the winners did. Now that wasn't an issue this year as we both advanced and no losses carried over, but in this new system if something like that happens there are severe playoff implications. Yes re-seeding would take more time and could be more complicated, but it would make it much harder for teams to complain that they had a hard bracket.
Reseeding all the teams after the morning would be awesome. It's also ABSOLUTELY not practical, given several things about NHBB but primarily the size and the existence of off-site groups. The only even remotely possible method of making reseeding work would be if lunch were even longer, and we had a crew of a dozen people to text out afternoon group assignments. Even then, the online schedule program would need to be rewritten (I don't know whether that would be a 5-minute job or a long, frustrating one).

There are a number of things about the structure of NHBB Nationals that I would improve, but the pre-planned group structure isn't one of them. (That said, having that pre-planned group structure completed more than ~4 hours before the tournament *IS* one of them!)
Jon Pinyan
Coach, Bergen County Academies (NJ); former player for BCA (2000-03) and WUSTL (2003-07)
HSQB forum mod, PACE member
Stat director for: NSC '13-'15, '17; ACF '14, '17, '19; NHBB '13-'15; NASAT '11

"A [...] wizard who controls the weather" - Jerry Vinokurov
tabstop
Wakka
Posts: 122
Joined: Tue Mar 20, 2007 10:48 am
Location: NNVA

Re: Discuss 2016 NHBB Nationals Here

Post by tabstop »

jonpin wrote:
jupiter wrote:Now if what I said is your plan, I would recommend re-seeding all the teams after the morning. We (LASA) were 5th in points and Naperville North was T-2nd but we ended up in the same afternoon group (there were 9 groups) because everything was pre-seeded; the winner or our morning group and the winner of Naperville's morning group would end up in the same afternoon group regardless of how well the winners did. Now that wasn't an issue this year as we both advanced and no losses carried over, but in this new system if something like that happens there are severe playoff implications. Yes re-seeding would take more time and could be more complicated, but it would make it much harder for teams to complain that they had a hard bracket.
Reseeding all the teams after the morning would be awesome. It's also ABSOLUTELY not practical, given several things about NHBB but primarily the size and the existence of off-site groups. The only even remotely possible method of making reseeding work would be if lunch were even longer, and we had a crew of a dozen people to text out afternoon group assignments. Even then, the online schedule program would need to be rewritten (I don't know whether that would be a 5-minute job or a long, frustrating one).
I have something built-in (which was supposed to get the JV teams in the right places but didn't quite work, but I think I've fixed it this time), but it doesn't do some basic things like "check that two teams from the same morning pool aren't in the same afternoon pool" which I would think would be important; even if we did get it all automated I would still want a pair of eyes (or two) on it to double-check. It's also a bit of an open question how "up" the website will be to get everything done and displayed (I thought we had a fairly large chunk of pipe as these things go, but I might be misremembering).
Andrew Feist
Great Bustard
Auron
Posts: 1463
Joined: Sun Aug 29, 2010 9:23 pm
Location: DC, NJ, and everywhere else
Contact:

Re: Discuss 2016 NHBB Nationals Here

Post by Great Bustard »

jonpin wrote:(That said, having that pre-planned group structure completed more than ~4 hours before the tournament *IS* one of them!)
This was an anomaly this year - in 6 years of NHBB, this was the first and only time the draws ever came out late (and as far as I know, this didn't lead to any backups in getting the Bowl started). Next year, this should again happen sometime between Tuesday and Thursday of Nationals week for the Bowl.
David Madden
Ridgewood (NJ) '99, Princeton '03
Founder and Director: International History Bee and Bowl, National History Bee and Bowl (High School Division), International History Olympiad, United States Geography Olympiad, US History Bee, US Academic Bee and Bowl, National Humanities Bee, National Science Bee, International Academic Bowl.
Adviser and former head coach for Team USA at the International Geography Olympiad
User avatar
jupiter
Lulu
Posts: 8
Joined: Mon May 27, 2013 11:26 pm
Location: Austin, TX

Re: Discuss 2016 NHBB Nationals Here

Post by jupiter »

We played against Naperville North 3 times. I'm curious, how many times have two teams played each other more than once at a single national tournament?
John DiCarlo

Kealing '12
LASA '16
MIT '20
jonah
Auron
Posts: 2383
Joined: Thu Jul 20, 2006 5:51 pm
Location: Chicago

Re: Discuss 2016 NHBB Nationals Here

Post by jonah »

jupiter wrote:We played against Naperville North 3 times. I'm curious, how many times have two teams played each other more than once at a single national tournament?
I don't know about NHBB national tournaments, but in general 3 is far from unusual. A lower bound on the record number of times is 5 between Wilmington Charter A and Dorman A at the 2009 HSNCT (Dorman actually won the series 3-2 even though Charter won the tournament).

Edit: That's the record at any NAQT national championship, tied with Illinois vs. Chicago A at the 2000 ICT. The overall record in the NAQT database is a 2007 Sectional.
Jonah Greenthal
National Academic Quiz Tournaments
swimmerstar
Lulu
Posts: 31
Joined: Mon Oct 27, 2014 3:34 pm

Re: Discuss 2016 NHBB Nationals Here

Post by swimmerstar »

When will the USHB packets be posted?
Fred Zhang
Longfellow Middle School 2015
TJHSST 2019
Yale 2023
Locked