Page 1 of 1

Conversion stats

Posted: Fri Jul 25, 2014 11:59 am
by theMoMA
I've computed the conversion stats based on the scoresheets that Kay provided, and I'll post a spreadsheet of them sometime this weekend on Google docs. I need to go through and fill in the answers before posting; right now, it's just listed by packet and question number.

Re: Conversion stats

Posted: Mon Jul 28, 2014 3:40 am
by theMoMA
https://docs.google.com/spreadsheets/d/ ... sp=sharing

Bonus stats by category:

18.77 Other art
18.24 Phil
17.70 Relig
17.27 Lit
16.76 Other
16.26 Hist
16.17 Music
16.10 Paint
16.13 Overall
15.49 Myth
14.88 SS
14.35 Chem
13.96 Physics
13.23 Bio
11.58 Other sci

Overall tossup conversion rate: 87%
Overall tossup neg rate: 21%

Re: Conversion stats

Posted: Mon Jul 28, 2014 9:41 am
by Galadedrid Damodred
I proudly present to you the most ridiculously hard tossup I wrote for this tournament:

3 TU2 Tawba surah 0 1 0 0 0 1 0 0 1 1 1 0.14 0.57

Yeah, that was probably a bad idea. At least it wasn't sulfonate!

Interesting to note that the much-discussed torsion bonus actually fell right in the middle of a cluster of 6 physics bonuses that all averaged a respectable 14.3 points. (Minor labeling nitpick: the rocket equation bonus is Other Sci.) Judging by the stats, you'd think that people would be complaining more about this one:

7 Bonus13 Physics Jarzynski equality, reversible, fluctuation theorem 0 1 0 0 0 0 1 2.857142857

Apparently no one (and especially not Chris Ray) can use the Jarzynski equality or fluctuation theorem...

Re: Conversion stats

Posted: Mon Jul 28, 2014 10:37 am
by Gautam
Galadedrid Damodred wrote: Yeah, that was probably a bad idea. At least it wasn't sulfonate!
LOL, I knew it would shade harder, but not this hard. Victor P. really advised me against using it. But it was a great submission from M. Hausinger! Oh well.

Re: Conversion stats

Posted: Mon Jul 28, 2014 10:55 am
by theMoMA
The four science categories generally had lower bonus conversion rates. I'm guessing this is partly because I'm not a scientist, so the general conception of bonus difficulty I had for the tournament was tougher for me to impose on science bonuses (although I did make several of them easier, mostly by adding comments for the other editors in Google docs and having them change the easy parts to be a little more forgiving). This is subjective, but I think that the gap between science and humanities was lower for the top half of the field, so I'm guessing another chunk of the conversion gap is that the lower-half teams tended to have better coverage of humanities than science (especially on some of the tougher easy parts, and those 0s make a difference quickly).

But I wonder if higher-level science is by nature a bit tougher than higher-level humanities. It seems to me that science players tend to know humanities better than non-science players know science. Perhaps at CO level, where middle parts are tough and often require collaboration between teammates to convert, a humanities bonus where 3 or 4 players have knowledge of the subject matter is going to be easier to play than an equivalently difficult science bonus where only 1 or 2 players can help out in the discussion. ("Equivalently difficult" defined as "equally difficult for the cohort of players with knowledge in the particular category.")

Maybe some science people could chime in on whether they thought that the science at CO was markedly harder than the humanities (and perhaps include some suggestions for future editors on bringing these divergent conversion numbers more in line, or if that's even a worthwhile goal).

Re: Conversion stats

Posted: Mon Jul 28, 2014 12:11 pm
by vinteuil
Yeah, so I'll apologize here for the polytextuality/De Vitry/Four/four bonus; I probably should have at least had a "description acceptable" tag for the first part, since I thought I was making it pretty clear that the pieces described had more than one text. I am a tiny bit surprised that two teams got Pli selon pli (if you are one of them, I like you, we should hang out some time) and none got De Vitry, but maybe the "wrong team got that bonus" in most rooms. In any case, I hope the last part was actually an easy part.

In general though, it looks like I did an OK job with middle parts and tossup answerlines at least (sorry for not making the French Horn giveaway easier! I thought that that piece was super-famous); so these conversion stats make me happy—thanks to Andrew for posting them!

Re: Conversion stats

Posted: Mon Jul 28, 2014 2:19 pm
by Ike
theMoMA wrote:The four science categories generally had lower bonus conversion rates. I'm guessing this is partly because I'm not a scientist, so the general conception of bonus difficulty I had for the tournament was tougher for me to impose on science bonuses (although I did make several of them easier, mostly by adding comments for the other editors in Google docs and having them change the easy parts to be a little more forgiving). This is subjective, but I think that the gap between science and humanities was lower for the top half of the field, so I'm guessing another chunk of the conversion gap is that the lower-half teams tended to have better coverage of humanities than science (especially on some of the tougher easy parts, and those 0s make a difference quickly).

But I wonder if higher-level science is by nature a bit tougher than higher-level humanities. It seems to me that science players tend to know humanities better than non-science players know science. Perhaps at CO level, where middle parts are tough and often require collaboration between teammates to convert, a humanities bonus where 3 or 4 players have knowledge of the subject matter is going to be easier to play than an equivalently difficult science bonus where only 1 or 2 players can help out in the discussion. ("Equivalently difficult" defined as "equally difficult for the cohort of players with knowledge in the particular category.")

Maybe some science people could chime in on whether they thought that the science at CO was markedly harder than the humanities (and perhaps include some suggestions for future editors on bringing these divergent conversion numbers more in line, or if that's even a worthwhile goal).
I think what you're saying about lower level teams is true and that CO science is by nature designed to be challenging; however I overshot the mark on some of the bonuses for Other Science, and I apologize. It seems the tossups were mostly fine in terms of gettability at the end, but I'll tone down the bonuses if I do CO other science for any future iterations. I'll note that other science is the most nebulously (harhar pun) defined category in quizbowl - I don't think anyone in QB has ever made claims to being an "other science" expert because that would require phds in at least 4 fields.

I personally think it's okay to have science be a few conversion points lower than some of the other categories. It's just going to happen because it's so much less "fun" to learn about proteins than it is about history for most of the field - and for the ones that actually do want to learn about proteins, CO is the place to challenge them.

Re: Conversion stats

Posted: Fri Aug 08, 2014 12:00 pm
by Sen. Estes Kefauver (D-TN)
Who 30ed my Siachen Glacier bonus? Or the one about Indian Christians?