NAQT results

Old college threads.
Locked
vsirin
Lulu
Posts: 19
Joined: Mon Mar 07, 2005 2:39 pm

NAQT results

Post by vsirin » Wed Apr 13, 2005 3:50 pm

Looking at the stats I noticed that Michigan had both less ppg and a lower bonus conversion than Chicago, and still managed to win. A brief scan of the NAQT archives suggests that this is the first time a team led the tournament in both those categories without winning. Last year, for instance, Michigan A had a slightly better bonus conversion than Berkeley, but Berkeley had much higher ppg. I don't know enough about statistics to tell if this was a major upset, but it did seem odd.

User avatar
MLafer
Tidus
Posts: 520
Joined: Mon Apr 28, 2003 7:00 pm

Post by MLafer » Wed Apr 13, 2005 4:05 pm

Correction:

I noticed an unusually low bonus conversion for us in the game against Illinois. I then realized that the final score to this game had been recorded incorrectly: it was actually 485-145, not 385-145. This puts bonus conversion for that game as 16.67 and changes our total bonus conversion to 19.158, compared to Chicago's 19.162, certainly not a statistically notable difference.

Additionally, it changes our total points from 5675 to 5775, for a points per 20 tossups heard of 363.78, compared to Chicago's 359.26.

Rothlover
Yuna
Posts: 816
Joined: Wed Feb 25, 2004 8:41 pm
Contact:

Post by Rothlover » Wed Apr 13, 2005 4:36 pm

Yes, I noticed the erroneous Mich-Ill result in addition to many other minor NAQT stat recording errors for us personally. Having not looked in depth into this in years past, is this par for the course for them?

User avatar
Steve Kaplan
Wakka
Posts: 115
Joined: Fri Aug 15, 2003 1:26 am
Location: Los Angeles

Post by Steve Kaplan » Wed Apr 13, 2005 6:42 pm

Rothlover wrote: is this par for the course for them?
I don't know if its par for the course, but they definitely messed up my stats for this year. More amusingly, they completely omitted me from the stats for the 2000 ICT (Berkeley Div. 2). This resulted in Berkeley's division 2 team, playing on the same set of questions as division 1, to seem to have the third highest bonus conversion at the tournament behind only Subash led Illinois and Andrew Yaphe led Chicago. This problem, though brought to the attention of the statkeeper five years ago was never corrected.

User avatar
matt979
Lulu
Posts: 80
Joined: Wed Apr 13, 2005 10:22 pm
Location: Alameda, CA
Contact:

Post by matt979 » Thu Apr 14, 2005 1:10 pm

Steve Kaplan wrote: I don't know if its par for the course, but they definitely messed up my stats for this year. More amusingly, they completely omitted me from the stats for the 2000 ICT (Berkeley Div. 2). This resulted in Berkeley's division 2 team, playing on the same set of questions as division 1, to seem to have the third highest bonus conversion at the tournament behind only Subash led Illinois and Andrew Yaphe led Chicago. This problem, though brought to the attention of the statkeeper five years ago was never corrected.
At this point the problem from 2000 might not be correctable. This is the first I've heard of it personally, though I didn't staff an ICT control room until 2002. The 2000 stats might have been done on my stats program, though what we'd really need is to go through the scoresheets assuming we still have 'em, and fix web pages by hand.

(By the way, if you're still using that, please update to Chris Sewell's SQBS program now. My spreadsheet is painfully obsolete compared to SQBS. http://www.stanford.edu/~csewell/sqbs/)

As for 2005, I believe the same person possesses both D1 scoresheets and the D1 SQBS file. If you send round-by-round stat corrections to matt-at-naqt then I can get someone to look at the scoresheets, though it wouldn't happen soon.

Locked