2021 ACF Nationals Qualifier Discussion Thread

Old college threads.
Locked
User avatar
Krik? Krik?! KRIIIIK!!!
Rikku
Posts: 325
Joined: Wed Aug 02, 2017 9:17 pm

2021 ACF Nationals Qualifier Discussion Thread

Post by Krik? Krik?! KRIIIIK!!! »

This thread will be to discuss questions and content related to the 2021 ACF Nationals Qualifier Packet. The packet can be found here. This thread will also be used to discuss advanced statistics produced for this tournament.
Ganon Evans
Misconduct Representative
ACF President, PACE VP of Editing, MOQBA
Francis Howell High School 2018, University of Iowa 2021
User avatar
naan/steak-holding toll
Auron
Posts: 2514
Joined: Mon Feb 28, 2011 11:53 pm
Location: New York, NY

Re: 2021 ACF Nationals Qualifier Discussion Thread

Post by naan/steak-holding toll »

Thanks to ACF for putting this together and settling on a reasonable qualifier format. I think this tournament ultimately did its job and had a lot of interesting content that made it an enjoyable listening experience; it was also very well run and the moderators were superb.

Most of this post is about the most noticeable aspect of this set - its variance. Bonus variance is something that happens in every tournament, but in a set consisting exclusively of bonuses, it certainly sticks out a bit more.

I initially reacted negatively to this tournament's rather pronounced difficulty variance between many bonuses. However, upon further thought, I do think that whether or not this was intentional, the impact of this variance is not only minimized by the fact that each team heard every bonus, but potentially could have some value. I could see an argument, ala one presented by Stephen Liu in the 2017 Regionals thread, for having a few tougher hard parts to really cut off the top teams from the ones that are more standard "good teams are very likely to 30 this" since you don't have the random factor of "did I get the tossup with the hard bonus."

On the other hand, I'm not sure this sort of variance is necessary to produce the granularity of selecting who gets to play at Nationals. Without seeing stats I have no idea, but I'd guess there was some pretty reasonable variance between categories on overall conversion. I think it also probably makes the set more frustrating to middle teams, who may only be able to consistently 30 bonuses in some categories, but are getting the same 20 as others and not able to distinguish themselves on tossup skill, simply because the hard parts in their area are super-hard.

EDIT: I doubt this qualifier format will happen again, but if it does, maybe four tiers of bonus part difficulty could be worth it / four-part bonuses? Or perhaps two-part bonuses to cover a wider spectrum of topics? Not sure how people feel about this, but could be worth discussing.
Will Alston
Dartmouth College '16
Columbia Business School '21
Votre Kickstarter Est Nul
Rikku
Posts: 363
Joined: Wed May 06, 2015 2:09 pm

Re: 2021 ACF Nationals Qualifier Discussion Thread

Post by Votre Kickstarter Est Nul »

naan/steak-holding toll wrote: Thu Apr 08, 2021 9:07 pm EDIT: I doubt this qualifier format will happen again, but if it does, maybe four tiers of bonus part difficulty could be worth it / four-part bonuses? Or perhaps two-part bonuses to cover a wider spectrum of topics? Not sure how people feel about this, but could be worth discussing.
I also enjoyed and am thankful to ACF for putting this together. Like Will, I will caveat my comment by saying it appears (hopefully) unlikely that this format will be necessary again. But, I think a format (either 2 part of 3 part) bonuses which don't have an easy part could make some sense for a qualifier like this. Given that, on some level, the purpose is to determine which teams would put up the best fight on Nats packets, it's somewhat irrelevant whether one on-the-edge team can 10 a chemistry bonus and the other can't; neither of those teams is likely to get a tossup in that category at Nats against any better team, or even at all, depending on that specific Q's difficulty. If team A 10s half of the chem bonuses, and otherwise puts up 14 PPB on everything else, and team B 0s every chem bonus but has a 15 PPB, and the math (which I'm too lazy to do) would see A be placed higher, I'm not sure that makes sense for the purpose of Nats qualification. At Nationals both teams will win or lose games in spite of, not because of, their respective chemistry knowledge. I know this is true if receiving a random bonus in a regular game, but those games have tossup-skill as an additional differentiator between the teams, which I think makes it a different enough situation.

A 3 part easyish medium / hardish medium / hard bonus may be more conducive to finding out the teams best equipped to hang at Nats, since the category of "teams who won't be getting this at Nats" will be folded into one group (presumably 0ing the bonus) without the seemingly irrelevant differentiator of "do they know a Regs level easy part or not."

Anyway, this is all academic, but I'm bored writing a paper and this thought is swirling through my head. My logic may also be entirely off. I swear this thought doesn't originate from the fact that I'd lose a chemistry packet to empty chairs.
Emmett Laurie
East Brunswick '16
Rutgers University '21
The Sawing-Off of Manhattan Island
Rikku
Posts: 433
Joined: Fri Feb 24, 2017 4:41 pm

Re: 2021 ACF Nationals Qualifier Discussion Thread

Post by The Sawing-Off of Manhattan Island »

To Will's point, I thought the variance at this tournament probably actually had some effects on the results just because of how pronounced the cross category variance was, even if everyone played the same set. Looking at the bonus where supersonic flow is the easy with no reasonably easy clues followed by a bonus where "In the Penal Colony" was the middle part with full information was super jarring, for instance. There's a few bonuses I'd cite like that (library of babel medium with full info, beams as an easy part eg) - as a result I think this set disproportionately rewarded elite science knowledge and punished "medium" literature knowledge.

I was not a fan of the Hoare / null pointer / deadlock bonus. Deadlock feels like a tough easy already, and hoare a tough hard without full info (or even saying something like "his namesake logic.") The null pointers part was confusing to me - besides the Hoare quote, I don't know what distinguishes this part from a part on dangling or wild pointers. I thought it could have really used a more specific definition of a null pointer, considering how many things in c++ lead to undefined behavior.

I thought it'd have been a nicety to make clear that the Virgil in painting part wasnt looking for the title character, but I understand that that's part of the knowledge being tested and it was certainly fair not to do so. Overall, I liked the painting in this set, as well as the myth which I thought was nicely distributed as well.
Hot Soup
Lulu
Posts: 14
Joined: Sat Oct 19, 2019 7:00 pm

Re: 2021 ACF Nationals Qualifier Discussion Thread

Post by Hot Soup »

I just wanted to drop by and mention that the bonus on soloists who performed with Bernstein was my favorite bonus of the year; thank you for making my day. Really can't get any better than having my favorite symphony orchestra + two of my favorite pianists show up as the three answerlines.
Paul Lee
Dunlap '15
Penn '19
WUSTL '2X
User avatar
rahulkeyal
Wakka
Posts: 150
Joined: Tue Jul 15, 2014 2:26 am

Re: 2021 ACF Nationals Qualifier Discussion Thread

Post by rahulkeyal »

John Quincy Adams's Alligator wrote: Fri Apr 09, 2021 1:10 pm To Will's point, I thought the variance at this tournament probably actually had some effects on the results just because of how pronounced the cross category variance was, even if everyone played the same set. Looking at the bonus where supersonic flow is the easy with no reasonably easy clues followed by a bonus where "In the Penal Colony" was the middle part with full information was super jarring, for instance. There's a few bonuses I'd cite like that (library of babel medium with full info, beams as an easy part eg) - as a result I think this set disproportionately rewarded elite science knowledge and punished "medium" literature knowledge.

I was not a fan of the Hoare / null pointer / deadlock bonus. Deadlock feels like a tough easy already, and hoare a tough hard without full info (or even saying something like "his namesake logic.") The null pointers part was confusing to me - besides the Hoare quote, I don't know what distinguishes this part from a part on dangling or wild pointers. I thought it could have really used a more specific definition of a null pointer, considering how many things in c++ lead to undefined behavior.
I agree with both of Vishwa's points here. My impression was that the medium parts of the literature bonuses were likely too easy for the field across the board, and would expect as such to bear out in the stats - the European and world literature bonuses in particular felt easier, whereas the Anglo-American content seemed a bit more in line with other categories. Similarly, I found null pointer confusing and a bit imprecise (at least approaching it from a classroom context) and think the bonus may have worked better where that part functioned as an easy.

Issues aside, I'm still very grateful to ACF and the Qualifier team for putting together a fun set of bonuses and a superbly-run event! There were many, many questions that I and my team enjoyed listening to, and this event seemed like a great alternative (hopefully just this once) given the circumstances of this competition season.
Rahul Keyal
Berkeley '21
Member, ACF
Locked