Search found 37 matches

by msaifutaa
Wed Jan 06, 2010 5:11 am
Forum: High school area archives
Topic: Centennial It's Academic Tournament (12/13)
Replies: 104
Views: 36942

Re: Centennial It's Academic Tournament (12/13)

My older brother played Quiz Bowl at MIT (at least up until he graduated I think), and even though I didn't play in college, I've been around for a while between my own high school experience and my brothers'. For various reasons, in my case mostly related to time, we were negligibly involved with ...
by msaifutaa
Tue May 06, 2008 10:17 pm
Forum: College area archives
Topic: VCU summer practices open to all
Replies: 19
Views: 4497

Re: VCU summer practices open to all

I won't be around, but I think Harvard will have summer practices, likely in conjunction with MIT as was the case last summer. Anyone in Boston is welcome. Send me info; I think Brown's practices will halt as most of our team will be away. But I'll probably have a car and can take what few are left...
by msaifutaa
Thu May 01, 2008 9:16 pm
Forum: College area archives
Topic: ACF nationals discussion
Replies: 89
Views: 18454

Re: ACF nationals discussion

I think Reissner-Nordstrom is one of those things that comes up often enough as a clue (to black holes, mostly) I can attest to at least anecdotal evidence of this--after reading this post this morning, we read Illinois Open '06 at practice today, and Reissner-Nordstrom opened a black hole toss-up,...
by msaifutaa
Tue Apr 29, 2008 7:49 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

I would also expect that the bottom bracket would have a higher percentage of dead tossups, but the overall conversion percentage for rounds 1-7 was 82% while it was 83% for rounds 8-14. The stratification of teams by ability did not appear to have a net effect on tossup conversion. I don't see why...
by msaifutaa
Tue Apr 29, 2008 7:15 pm
Forum: College area archives
Topic: Problems with Packets
Replies: 32
Views: 5980

Re: Problems with Packets

Tsukuyomi has shown up at least once as a tossup answer in the past three years with the Uke Mochi clue as a lead-in. I don't know what if anything this proves, but I thought I'd point it out. 12. He angered his sister when he killed Uke Mochi, the goddess of food, after the former sent him to repr...
by msaifutaa
Tue Apr 29, 2008 7:08 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

I admit that I gave a mistaken impression of how the ACF quota works, but I don't feel that the essence of the point was wrong: People submit 1/1 mythology for the round, but, on average 83% of the 18 (for instance) rounds will have a mythology questions and 83% will have a mythology bonus. How is ...
by msaifutaa
Tue Apr 29, 2008 6:59 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

Well, I'd be careful about using conversion percentage; Maryland negged the Bragi tossup, so it went unanswered, but we would have answered it had we not negged; presumably, that happened in some other rooms as well. Additionally, although I am not very good at mythology (and it's in my self-intere...
by msaifutaa
Tue Apr 29, 2008 5:50 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

Can you e-mail NAQT the specifics of this at errata@naqt.com? If I have the year correct, this error has never been reported and fixed in our packets. Needless to say, I didn't feel particularly enthused to pay money for those packets (so this also means I didn't read it and the moderator could hav...
by msaifutaa
Tue Apr 29, 2008 5:39 pm
Forum: College area archives
Topic: Problems with Packets
Replies: 32
Views: 5980

Re: Problems with Packets

The main problem with this question was it forced me to parse out what exactly they were asking for in the first line. I'm fine with lateral thinking on middle clues, but the first clue should always be about the person or thing being asked directly rather than a second-order clue. There's scores o...
by msaifutaa
Tue Apr 29, 2008 1:40 pm
Forum: College area archives
Topic: Problems with Packets
Replies: 32
Views: 5980

Re: Problems with Packets

There were also some issues with the myth; again, given the much-less-than-ideal conditions of the editing process, I don't think this year's set necessarily reflects on Matt or anyone else that worked on the myth. It's a little hard to remember things without the set, but I know the tossups on Iza...
by msaifutaa
Tue Apr 29, 2008 12:56 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

I guess you can put me down for genuinely believing that categories that can vary across packets is a useful way to define distributions because it allows you to work in increments other than 18/18 (if you have an 18-packet set). NAQT thinks there should be questions about foreign language in its s...
by msaifutaa
Tue Apr 29, 2008 12:31 pm
Forum: College area archives
Topic: ACF nationals discussion
Replies: 89
Views: 18454

Re: ACF nationals discussion

The tournament did go kind of long, but I was actually quite surprised by the fact that it didn’t become a drag (at least for me and many of my teammates) by the end. We all had quite a lot of fun, and I’m not sure we were expecting that. Having not actually met almost any of the teams here before (...
by msaifutaa
Fri Apr 18, 2008 4:12 pm
Forum: College area archives
Topic: CCs and the ICT
Replies: 37
Views: 6820

Re: CCs and the ICT

I am not really all too worried about the 25th best Division II team missing the ICT. Having space for CCs is a perfectly legitimate goal, and axing the 25th through 32nd best DII teams to meet that goal seems like a reasonable tradeoff. The 25th best Division II team is always going to be a fringe...
by msaifutaa
Thu Apr 17, 2008 6:09 pm
Forum: College area archives
Topic: Sub-distribution suggestions
Replies: 33
Views: 5251

Re: Sub-distribution suggestions

I hope I'm not posting this twice. Norse myth offers a lot of tossup-worthy answers relative to its size in this proposed distribution, I think. I guess I agree with Seth on treating Norse, Egyptian, Hindu, and Celtic/European mythology equally (though if there's a least among equals here, it's pro...
by msaifutaa
Thu Apr 17, 2008 5:38 pm
Forum: College area archives
Topic: Sub-distribution suggestions
Replies: 33
Views: 5251

Re: Sub-distribution suggestions

Anyway, here's a counter-proposal with a total of 15/15: 6/6 Classical 1/1 Norse 1/1 Egyptian 1/1 Indian 1/1 European (including Celtic/Arthurian, Eastern, etc.) 1/0 or 0/1 Mesopotamian 1/0 or 0/1 East Asian 1/0 or 0/1 American 3/4 or 4/3 whatever, including mixed-pantheon stuff If space is made fo...
by msaifutaa
Thu Apr 17, 2008 3:50 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

My personal distribution (keeping in mind that I know NAQT does want to bias more towards the most widely-known mythologies than I would normally) for 18 packets would probably be more like: 4/4 Classical 2/2 Norse 2/2 Mesopotamian 2/2 Egyptian 2/1 Celtic (the 2/1 and 1/2 are interchangable here an...
by msaifutaa
Wed Apr 16, 2008 5:01 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

Hmm, maybe a lot of the SCT myth this year was in the last two rounds for finals that I didn't read then. That's the other issue with the no in-packet distribution. And I know at SCT the rest of my team who were playing didn't even hear all of the 4/3 I got to read because their readers only got th...
by msaifutaa
Wed Apr 16, 2008 4:05 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

Yes, there's a fixed quantity of mythology. The ICT contained 9/9 myth, distributed like this: 4/5 classical 1/1 Norse 1/1 Egyptian 1/1 Mesopotamian 1/1 American 1/0 East Asian That's 1/1 myth in every two packets, with a roughly even split between classical and everything else. "Should there ...
by msaifutaa
Wed Apr 16, 2008 3:28 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

I would be more likely to play if RMP received its own distribution I don't understand this. Why do we care how things are categorized? I'm pretty sure religion, myth and philosophy all get fixed numbers of questions in SCT/ICT sets. Are you saying you want to see more of each? I agree with you on ...
by msaifutaa
Tue Apr 15, 2008 9:01 pm
Forum: College area archives
Topic: NAQT's Post-ICT Survey
Replies: 84
Views: 20188

Re: NAQT's Post-ICT Survey

-The conflation of mythology with literature. Who advocates this? Why not have a true RMP distribution? And why not have a more extensive non-western myth distribution, as most other tournaments do? Quoted for Truth. Every year on these surveys, I've selected the "Drastically Increase" (o...
by msaifutaa
Thu Mar 06, 2008 2:52 am
Forum: College area archives
Topic: Pronunciation question
Replies: 24
Views: 6255

Re: Pronunciation question

Within the Altaic language family, Kazakhstan's name is pronounced Qozoqistan (or other spelling variants) in the Oghuz (Turkmen, Azeri, Turkish) and Karluk (Uzbek, etc.) groups. Therefore saying it as four syllables should be perfectly fine-as a Turkish speaker I'm inclined personally to add in th...
by msaifutaa
Thu Mar 06, 2008 1:42 am
Forum: College area archives
Topic: Pronunciation question
Replies: 24
Views: 6255

Re: Pronunciation question

So I've come to accept that correct/precise pronunciation is not necessary to demonstrate "knowledge" for the purpose of quizbowl. But where do you draw the line? Is knowing the consonants + number of syllables enough to demonstrate "knowledge"? To me, this standard seems utterl...
by msaifutaa
Thu Feb 28, 2008 10:42 pm
Forum: College area archives
Topic: College quizbowl 4 years from now...
Replies: 17
Views: 4266

Re: College quizbowl 4 years from now...

Neither of these approaches is necessarily a bad system. What you could do is a have a "packet plus" approach, where everyone writes a single packet and then edits one area of their specialty. This can also iron out any biases in the packets that might be present. For example, for EFT, De...
by msaifutaa
Thu Feb 28, 2008 4:33 pm
Forum: College area archives
Topic: College quizbowl 4 years from now...
Replies: 17
Views: 4266

Re: College quizbowl 4 years from now...

Let me offer some suggestions on how this can be done. I would encourage everyone in the club to work on writing a single packet over the summer. Three months should be plenty of time for just about anyone to put together a decent packet, and then come September, you can get together with people an...
by msaifutaa
Thu Feb 28, 2008 3:45 pm
Forum: College area archives
Topic: College quizbowl 4 years from now...
Replies: 17
Views: 4266

Re: College quizbowl 4 years from now...

Beyond that, there seem to be young teams who know lots of things and have pretty good reputations for question-writing, like MIT, Harvard, Drake, and Dartmouth, who have untapped editing potential. I think one initiative that should be undertaken is to have these teams (and more) start running hou...
by msaifutaa
Fri Feb 22, 2008 7:14 pm
Forum: College area archives
Topic: POLL
Replies: 38
Views: 9542

Interesting curiousity; if you take out Chris' ballot, then MIT remains at 11. However, then Chicago, Maryland, and Brown are all tied at exactly 240 points. Yup, Chris's wasn't the one missing MIT though, so it isn't a surprise that we stay at 11 if you remove it. If you take out Chiego's ballot, ...
by msaifutaa
Fri Feb 22, 2008 6:56 pm
Forum: College area archives
Topic: POLL
Replies: 38
Views: 9542

Actually Dwight, Eric Mukherjee was retarded enough to leave Vandy off his ballot. And Jerry, yeah, I got a couple of top ten ballots, but I didn't count those as "rejected" since by rejected I really mean OMFG THE WEST IS THE BEST LET'S RANK UC-DAVIS #4. And yeah, I have gotten that crap...
by msaifutaa
Fri Feb 15, 2008 7:07 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

This can probably be split off into a new thread, but Mark, you have a good point. On one hand, it can be very disheartening for new players to play against Jason Keller and co (in fact, its disheartening for anyone); at the same time, you don't want to baby your freshman into solely playing novice...
by msaifutaa
Fri Feb 15, 2008 1:30 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

Would emailing NAQT immediately and asking them to take you off the waitlist "because you can't travel" or something like that work? NAQT is always willing to listen to arguments in favor of exceptions or exemptions, but in this case we would be unlikely to accept them. Our principal conc...
by msaifutaa
Thu Feb 14, 2008 7:47 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

stuff about Brandeis having experience As far as experience on Brandeis's team goes, we had one freshman, two sophomores, and myself (a senior). One of the sophomores did not play last year at SCT, while the other did (as did I), but since we did not qualify last year, it's irrelevant. We hosted SC...
by msaifutaa
Thu Feb 14, 2008 7:27 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

Wow, MIT was certainly not expecting this, especially not the DII team. That's pretty amazing, but ironically, it actually could be pretty bad for some of our DII players, particularly our frosh--does being on the waitlist disqualify the whole team from DII next year? Being waitlisted does not, but...
by msaifutaa
Thu Feb 14, 2008 7:19 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

Wow, MIT was certainly not expecting this, especially not the DII team. That's pretty amazing, but ironically, it actually could be pretty bad for some of our DII players, particularly our frosh--does being on the waitlist disqualify the whole team from DII next year? Being waitlisted does not, but...
by msaifutaa
Thu Feb 14, 2008 6:59 pm
Forum: College area archives
Topic: NAQT SCT Results and ICT Invitations
Replies: 52
Views: 10952

Wow, MIT was certainly not expecting this, especially not the DII team. That's pretty amazing, but ironically, it actually could be pretty bad for some of our DII players, particularly our frosh--does being on the waitlist disqualify the whole team from DII next year? We initially had two DII teams ...
by msaifutaa
Tue Feb 12, 2008 9:26 pm
Forum: College area archives
Topic: Cardinal Classic Discussion
Replies: 46
Views: 9471

It may be taught at Maryland, but I wouldn't know since AI isn't a required course here. Anyways, relatively advanced topics from upper level courses are typically not the greatest things to introduce into the canon at non-national events. I'm all for expanding the CS canon, but I personally don't ...
by msaifutaa
Tue Feb 12, 2008 9:05 pm
Forum: College area archives
Topic: Cardinal Classic Discussion
Replies: 46
Views: 9471

(incidentally, what makes you think everyone in CS takes a machine learning course?). It's not a machine learning course--it's an Intro AI course, and a requirement to graduate for those on the CS side of EECS at MIT. If the (admittedly incorrect) assumption that it was typical was correct, then it...
by msaifutaa
Tue Feb 12, 2008 8:37 pm
Forum: College area archives
Topic: Cardinal Classic Discussion
Replies: 46
Views: 9471

Hi, Did anyone actually get the AdaBoost question? It went dead in every game situation that I've heard of so far. My own opinion is that AdaBoost is too obscure and I wish you wouldn't write tossups on it again. MaS Ironically, due to the discussion here, I think that the topic has gained addition...
by msaifutaa
Tue Feb 12, 2008 7:30 pm
Forum: College area archives
Topic: Cardinal Classic Discussion
Replies: 46
Views: 9471

For any set of weak-classifiers, this process is guaranteed to converge to a solution that classifies at least as well as the best classifier. Developed by Yoav Freund and Robert Schapire, this meta-algorithm involves repeatedly reclassifying the training data with the best of the set of original c...