Server upgrade complete!

Welcome to the upgraded server! If you notice anything wrong with the board, please let Jeffrey know!Free packets for study/practiceAnnounced tournament dates

Are NAQT's computational math questions too easy?

Discussion of quizbowl topics not related to specific tournaments
Post Reply
User avatar
Posts: 68
Joined: Mon Sep 20, 2010 3:22 pm
Location: San Francisco Bay Area

Are NAQT's computational math questions too easy?

Post by bt_green_warbler » Thu May 05, 2011 4:22 pm

Coach Brown wrote:My prediction for all the tournaments is that state records are going to fall because these questions, in my opinion, are way TOO EASY. I hope the state questions show a higher difficulty. In particular, the math questions that tell you how to solve them are ridiculously easy. I have complained to NAQT about this, but they have completely ignored me in this regard. Games in which the two teams total 840 points (out of 900) back up my complaint. Look at Ladue's district where every game totaled 840 points. Simply put, bad quiz bowl. It's just as bad as when scores are 130 to 100. So, say goodbye to many state records that were accomplished on questions that were of higher difficulty than these questions.
(Source in the "State predictions" thread)

My apologies on behalf of NAQT for not responding to your previous feedback.

Some quick notes about difficulty and conversion:

1) Yes, the questions for states will be more difficult than the questions used at districts and sectionals.

2) We have found essentially no statistical evidence that our computational questions are "ridiculously easy."

Here are the aggregate statistics for the three tournaments that used Invitational Series #101A in Ohio, Illinois, and Minnesota: (101A was one of the sets that shared questions with the districts and sectionals sets, so these are literally the same questions.)

Computational tossups: were answered correctly in 69.37% of rooms, and powered (earning 15 points for an early buzz in NAQT's nationwide format) in 22.34% of rooms that answered the tossup correctly.

All other tossups: were answered correctly in 78.42% of rooms, and powered in 21.1% of rooms that answered the tossup correctly.

In short: our computational math questions are among the most difficult questions we produce (this is a very consistent trend in our feedback going back to at least 2004). I see no convincing case that making computation questions even more difficult would improve the quality of our sets.

(Edit: Here's a link to my thread on the national board presenting overall conversion stats on NAQT questions from 2005 to 2009; note that over a sample of nearly 20,000 tossups, computational math finished dead last in answerability.)

User avatar
Posts: 70
Joined: Thu Jul 22, 2010 12:26 am

Re: Are NAQT's computational math questions too easy?

Post by WilliamofOrange » Thu May 05, 2011 5:04 pm

I feel Coach Brown may have been using some... "abnormal" evidence in his conclusion. In each of the games in that district, either Ladue or Clayton was present, both of whose teams have insanely strong computational math players.

User avatar
Posts: 218
Joined: Sun Mar 14, 2010 5:56 pm

Re: Are NAQT's computational math questions too easy?

Post by Mewto55555 » Thu May 05, 2011 5:49 pm

In the sectional round whose score hit 870 total, I don't think any math questions made it beyond half a line, if not less, with both Will and me in the room. "Difficulty" was irrelevant in basically any game we've played, because, frankly, Ladue has a bunch of people good at math.

That said, I'd like to see math disappear.

Post Reply