HSNCT-only predictions

Dormant threads from the high school sections are preserved here.
Locked
vathreya
Lulu
Posts: 35
Joined: Tue Jun 20, 2017 1:26 am

HSNCT-only predictions

Post by vathreya »

Hey everyone!

One day I was wondering why Morlan's rankings and the Pre-Nationals polls aren't quite as accurate in predicting results for nationals, and I noticed that while both seemed to be similar in their predictions, they weren't able to distinguish between teams that were good at housewrites, teams that were good at NAQT sets, and teams that were good at both. Additionally, HSNCT and NSC require different playing styles (HSNCT being very heavy on tossups and aggression and NSC being more focused on bonuses and relatively relaxed play) and have very different distributions (HSNCT favoring science, geography, and current events while NSC favors the arts and literature/philosophy), leading to a discrepancy in the results. Since the rankings visible to most teams don't usually account for this discrepancy, I present to you, my HSNCT-only predictions, that take into account only NAQT sets in order to more accurately predict HSNCT results.

Since I'm unable to find statistics for every team there is, I'd appreciate it if you could enter your team's statistics into the spreadsheet here: https://docs.google.com/spreadsheets/d/ ... sp=sharing

This is being updated live (as of now), so be sure to check back as more and more teams are added! I still haven't decided what I'll rank teams based on, and since this is my first time doing anything of this sort, advice is much appreciated.

Thanks in advance for your advice/contribution.
Vikshar Athreya
UW'22
Subotai the Valiant, Final Dog of War
Wakka
Posts: 236
Joined: Tue Aug 08, 2017 6:12 pm

Re: HSNCT-only predictions

Post by Subotai the Valiant, Final Dog of War »

I love this idea! I personally identify myself as an HSNCT player, and I've had thoughts about this for a while. Very small point: Different tournaments play on different rules; some play timed, while others (like pretty much all tournaments near NYC, for example) play 20 tossups all the time, which has a significant impact on buzzes per game. (I also mentioned this on the doc).
Daniel, Hunter College High School '19, Yale '23
User avatar
34 + P.J. Dozier
Wakka
Posts: 197
Joined: Sun Oct 15, 2017 10:01 pm
Location: Chicago, IL

Re: HSNCT-only predictions

Post by 34 + P.J. Dozier »

I think this is a great idea as well! However, I do think that averaging IS PPBs over the season, while probably the best methodology for the purposes of this poll, does fail to account for teams that have significantly improved or teams that have failed to play tournaments with their full rosters (both would apply to my team, which is admittedly not a nationally-contending team, but could also apply to many other teams that are nationally-contending).
Wonyoung Jang
Belmont '18 // UChicago '22
ACF; NAQT; PACE
vathreya
Lulu
Posts: 35
Joined: Tue Jun 20, 2017 1:26 am

Re: HSNCT-only predictions

Post by vathreya »

I've attempted to fix these discrepancies by using Buzzes/20TUH instead of buzzes/game.
Vikshar Athreya
UW'22
browen
Lulu
Posts: 39
Joined: Tue Aug 02, 2016 1:33 pm

Re: HSNCT-only predictions

Post by browen »

Thanks for making the spreadsheet Vikshar. I'm very interested in seeing how the combined average sorts out all of the contenders for Top 25 spots at the national championships.

I've got a question though.
vathreya wrote: Additionally, HSNCT and NSC require different playing styles (HSNCT being very heavy on tossups and aggression and NSC being more focused on bonuses and relatively relaxed play).
I've been coaching for quite a while now, and played both national tournaments my last three years in high school. I was fairly confident the consensus was the NSC rewards aggression more, and HSNCT rewards more conservative play. There's very little risk in going for the twenty-pointers at the NSC since you can limit your losses on the bounce-backs, compared to losing points at HSNCT and being locked out of the bonus.

The following might be purely anecdotal, but...
from recent times (last year):
(1) Winchester Thurston's finished tied for ninth last year while having less powers in each of the four games they won in the playoffs.
(2) Barrington's much more aggressive style of play cost them a deep HSNCT run, but rewarded them at the NSC.

from a long time ago:
(3) In my senior year of high school, St. John's only racked up twenty-eight negs through seventeen matches at HSNCT, but despite having the highest PPB in the Saturday afternoon playoffs they finished fourth in their group (I actually remember Jason Flowers discussing why St. John's had underperformed with another coach).

There might be a thread in the theory section on this, but I haven't found one. What does everyone else think?
Brian Owen
Dorman High School Assistant Coach (2014-present)
Dorman '13
5/31/2009: Never Forget
User avatar
Carlos Be
Wakka
Posts: 217
Joined: Sun Jun 25, 2017 11:34 pm

Re: HSNCT-only predictions

Post by Carlos Be »

Does buzzes per twenty tossups heard include negs? It seems like it does, but I can't think of what the use that stat would be for comparing teams if it can't distinguish between teams that score a lot and teams that neg a lot. Similarly, what's the advantage of using powers/buzz instead of powers/20TUH?
Justine French
she/her
vathreya
Lulu
Posts: 35
Joined: Tue Jun 20, 2017 1:26 am

Re: HSNCT-only predictions

Post by vathreya »

I was of the opinion that since HNSCT is timed and has more tossups and no bouncebacks, it places a greater emphasis on tossup-getting than bonus conversion, because you don't have a chance at bouncebacks and have to rely almost solely on tossup-getting.
Vikshar Athreya
UW'22
User avatar
TheExclusionary
Kimahri
Posts: 3
Joined: Fri Mar 09, 2018 11:35 am

Re: HSNCT-only predictions

Post by TheExclusionary »

browen wrote:(2) Barrington's much more aggressive style of play cost them a deep HSNCT run, but rewarded them at the NSC.
Hmmmm as a so-so to decent player from Barrington this statement is true for everyone involved last year but I tend to be a very conservative player especially if my teammates know the topic better than I do (thus a very low PPG from HSNCT.) On a Sunday morning when everyone is tired and being overly confident from a good run the prior day is what I think personally costed us a very good shot at making it to the finals. I have my deepest respect to my former and current teammates who buzz aggressively on their specialities which indeed can get anyone far or outright knock them down. PACE I think rewards people with deeper knowledge of topics while NAQT allows for more generalization.

An interesting and thoughtful spreadsheet but anything can happen at any tournament regardless of who is favored to win.
Kyle Asta
Barrington '18
DePaul '22?
vathreya
Lulu
Posts: 35
Joined: Tue Jun 20, 2017 1:26 am

Re: HSNCT-only predictions

Post by vathreya »

justinfrench1728 wrote:Does buzzes per twenty tossups heard include negs? It seems like it does, but I can't think of what the use that stat would be for comparing teams if it can't distinguish between teams that score a lot and teams that neg a lot. Similarly, what's the advantage of using powers/buzz instead of powers/20TUH?
Yes, the buzzes per game does include negs. The buzzes per game isn't a factor in the rankings itself - rather I included it as a benchmark so that I could have something to compare the Power ratio and Neg ratio stats (my primary method of gauging tossup performance) to. Powers per buzz, in my opinion, is a better indicator of skill than powers/20TUH, because teams in relatively competitive regions could power a lot when they buzz, however, not buzz often due to good opponents, whereas teams in regions that aren't as competitive could buzz often but not power as often when they buzz (due to a lack of competitors that can get stuff even earlier and/or competitors that can win buzzer races).

For example, if Team A buzzes in X times over the course of a tournament, of which 70% are powers, that team is arguably better than Team B, which buzzed in Y times over the course of a tournament but of their buzzes only 50% were powers. Even if Team B got the same amount of powers over the course of a tournament, statistically, there's a greater likelihood that if Team A buzzed, it was a power, which in my opinion is an indication that Team A is a better team.
Vikshar Athreya
UW'22
Mad Lounge Czar
Lulu
Posts: 9
Joined: Thu Feb 14, 2013 8:28 pm

Re: HSNCT-only predictions

Post by Mad Lounge Czar »

browen wrote: from a long time ago:
(3) In my senior year of high school, St. John's only racked up twenty-eight negs through seventeen matches at HSNCT, but despite having the highest PPB in the Saturday afternoon playoffs they finished fourth in their group (I actually remember Jason Flowers discussing why St. John's had underperformed with another coach).
For clarification: St. John's finished 4th at HSNCT and 10th at NSC (not making it to the championship superplayoff) in 2013.

As a member of that St. John's team I'm inclined to agree with Brian's observations. Since I started my high school career as a 3rd/4th scorer I was trained to play conservatively, which carried over as I continued to improve. While this allowed me to maintain high G/N ratios, it also led to comparatively weaker power numbers which may have cost us higher finishes (if you look at 2014 St. John's you'll notice that our power numbers are relatively weak compared to our overall finishes). Up until my senior year NSC was always my greatest source of self-frustration because my team was expected to contend for top bracket, though we would always fall short in the playoffs. I always thought that I would've done better at NSC based on my specialty areas, though as it turns out my team's playing style may have been a better predictor of our overall finishes at the different nationals.

Another thing to consider is that NSC doesn't neg you for incorrect buzzes, further incentivizing aggression since an incorrect buzz doesn't immediately set you back.
Carlo D (formerly SLHS, SJS, TAMU)
Locked