NAQT's Post-ICT Survey

Old college threads.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

cvdwightw wrote:One question that I would like to see is regarding NAQT's chemistry distribution. It is generally divided into elemental and non-elemental. This works well at the high school level, but I think that at the college level this distribution does not accurately reflect what people learn through coursework (namely, people study elements in a context and not, so to speak, in a vacuum). Perhaps a survey question could ask whether people would prefer a chemistry distribution divided into organic/biological and non-organic (in which element questions could still definitely be written, but would be more relevant to what is taught in college chemistry courses).
This gives a somewhat misleading view of NAQT's chemistry subdistribution at the college level; while it is true that we have a quota for questions about elements (though that can be interpreted quite broadly), there are also much larger subcategories for physical chemistry, organic chemistry, thermodynamics, and "anything except elements" questions.
msaifutaa
Lulu
Posts: 37
Joined: Tue Feb 12, 2008 6:40 pm

Re: NAQT's Post-ICT Survey

Post by msaifutaa »

rhentzel wrote:
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 sets; does that mean there should be 18/18 in a set? No. It seems to me that ACF feels similarly: philosophy, mythology, and religion belong in a set, but each one isn't given 18/18, instead each is given 12/12 (approximately) so that not every packet is the same.
That's not quite how ACF works, though. Since ACF is packet submission, it is the case that every submitted packet will have 1/1 Mythology, 1/1 Philosophy, 1/1 Religion. However, since ACF is not timed and instead stops at question 20, the questions after 20 will not be read. Thus, in, for instance, a 24/24 packet submission (reasonably typical), you get a 5/6 chance to read the mythology toss-up in a normal round. Now, even if not, it's still there--it might be used to break ties or as an extra question if the moderator screws up. I feel that this is a fundamentally different paradigm to not having set distribution requirements.

On a related note, I really dislike foreign language bonuses. And it's not that I don't know words in foreign languages--as a bonus, they usually just seem a cheesy way to get a free 30. This even from someone who enjoys trying to fraud the ends of legitimate academic topics (like John Chrysostom) by making up words from Latin or Greek.
Mark
MIT
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

cvdwightw wrote:I think it would be extremely relevant to find out players' levels of experience with pyramidal questions. Players who play or have played formats with longer questions (e.g. PACE and ACF) may have radically different views on things than players with no experience or whose previous experience has been something like CBI or :chip: . I see NAQT as sort of a "transitional" format, combining high-quality question writing on relevant subjects (a staple of ACF) with a timed tournament and a higher non-academic content that are more common to CBI. Accordingly, the format is not entirely foreign to people accustomed to either ACF-style or CBI-style tournaments, and the ability to please both audiences is a large part of its appeal.
This reply doesn't have much to do with the survey, but I thought I'd point out, for the sake of discussion, that this is not how NAQT sees itself.

We don't view ourselves as a "transitional format" through which players and schools must/should evolve from primitive, timed formats toward sophisticated, untimed, purely academic formats; we view ourselves as simply "a format," but hopefully the one that brings the most education and enjoyment to the most people, regardless of their skill level.

We certainly understand that some people prefer other formats, and we're happy that they have opportunities to compete in tournaments that more closely fit their interests, values, and personalities; part of the strength of the current quiz bowl circuit is its diversity.

In a similar vein, we don't view the audience of potential players as falling into two "pools," those who would prefer to play CBI (but are willing to play NAQT) and those who prefer to play ACF (but are willing to play NAQT), but as a much more complex landscape that includes a large number of players who actually prefer to play NAQT.

I think it's wrong to view NAQT as simultaneously trying to craft policies as a "balancing act" to get pro-CBI players to sign up without alienating pro-ACF players; while we certainly want the first to happen and the second not to, we make our decisions based on the idea that we are offering the best form of quiz bowl.

For instance, NAQT doesn't have a clock to appeal to CBI teams. NAQT has a clock because it helps us run our tournaments on time and, in our opinion, gives games a quicker pace and makes them more exciting. On the flip side, we don't have mostly academic content to appease ACF teams, but because we believe that that's what quiz bowl should (mostly) be about.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

Brian Ulrich wrote:Just a note on the CE: I don't remember details, but right now the distribution is divided into "U.S." and "World," with these being subdivided into politics, social, and miscellaneous. The line between politics and social is pretty vague, and I suspect its initial purpose is to get people to ask about stuff other than Senators and world leaders.
This is true. As currently formulated, "political" current events are those about actual politicians and elections, while "social" is everything else (except business, art, and science), including things with a strong political component like Supreme Court decisions or gun policy.

Brian is correct that the original motivation was to reduce the number of "world leader X was re-elected" questions. "Social" was probably a poor name for the category.

I'll add in passing that current events is one of a number of areas in which NAQT is chronically short of questions; if you enjoy current events and think that you could do as good (or better!) job, please send us a writing sample. We could use questions at all levels of play from A-level invitational series to the DI ICT (active DI players are welcome to write for the DII ICT, invitational series, and the HSNCT).
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

ToStrikeInfinitely wrote:I'd personally think that NAQT should simply bring in a good bio/chem/physics editor and allow him or her to set the distribution as they see fit.


NAQT subject editors already have relatively wide latitude to set up the distributions in their subject areas as they see fit.

NAQT is also always on the lookout for new writers and editors; we make no promises that anybody that applies would be immediately made a subject editor or given sole control over the ICT, but people willing to put in the time can be assured of rapid advancement through NAQT's ranks (such as they are), including membership.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

ToStrikeInfinitely wrote:In general, I feel this super-rigid distribution in the sciences categories will hurt the production of a good tournament rather than help it. (perhaps this can be worked into a question?)
I'm curious as to why you feel this way; from my point of view, the subdistributions that NAQT uses contribute to packet quality in a number of ways: They help ensure that our questions don't focus too heavily on the specific interests of high-production writers, they eliminate some of the time that set editors would have to spend determining whether a category was balanced, and they suggest to writers very clearly what sorts of questions are stilled needed in the final set.

There's clearly no guarantee that questions matching NAQT's subdistribution will be good questions or not flawed in other ways . . . but if a computer can enforce some aspects of a good distribution, why not let it and free up man-hours for subtler issues?
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

ToStrikeInfinitely wrote:The high content of trash and CE in a tournament that bills itself as "National ACADEMIC Quiz Competitions". Considering the NAQT current events editor himself has posted on this forum that he finds it difficult to find this many relevant current events topics, why is there no talk of seriously scaling back the CE distribution?
Possibly for the same reason that the fact that we have problems filling our physics distribution hasn't led anybody to suggest that we scale back the amount of physics. You should trust me that you would not enjoy playing on packets that NAQT found easy to fill. :-)

More likely the lack of discussion is because most people assume that NAQT will include its standard "Should there be lots more, more, the same, fewer, or many fewer" question in the survey so they will get a chance to express their opinion on the issue.

For the record, NAQT believes that awareness of ongoing current events is an important part of a general education and belongs in quiz bowl.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

ToStrikeInfinitely wrote:I would like to see the character length of NAQT questions addressed, and I'd like to see what people's positions on the amount of time given are. While I personally have no qualms (inasmuch as they're an interesting way to mix it up) with the clock and powers, several editors have told me that writing a good question within the character limit is very very difficult. Why not extend the time and the question length a little? I feel this would preserve the fast-pacing that makes NAQT fun while allowing for meatier questions.
If people would rather spend more time at the ICT, it could either be spent playing more matches, playing more questions in each match, or playing longer questions in each match.

I haven't done any mathematics to work this out, but I suspect that playing more questions in each match would do the most to increase the amount of "fairness" (though that's a difficult-to-measure concept).

We've added a survey question to address this issue, so we'll have a chance to see what people would prefer.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

ToStrikeInfinitely wrote:The conflation of mythology with literature. Who advocates this? Why not have a true RMP distribution?
What makes an "RMP distribution" true? Does mythology really have more in common with philosophy than literature?

I also don't understand why players would even care that mythology was part of literature; it gets the same number of questions, the only difference is that when the packet is 0/1 or 1/0 "short" on novels, plays, and poems that "shortfall" is likely to be made up with mythology. This strikes me as a sensible way to approach things.
And why not have a more extensive non-western myth distribution, as most other tournaments do?
Basically out of fears that non-classical, non-Norse, non-Egyptian myth would, broadly speaking, prove too difficult to meet our conversion targets. Or, conversely, that requiring more of it while meeting our conversion targets would result in endless repetition of the same answers.

For instance, at the DI ICT only 60% of the rooms converted Bragi (9/15), 63% (7/11) converted Ptah, and 84% (11/13) converted Jimmu Tenno. 100% converted Eumenides, Marduk, Medea, and Paris, and 92% converted Cassandra.

If we were seeing 90% conversion rates on all of those answers, NAQT would have no problems whatsoever delving into harder mythoi; we have no objection to them on principle, particularly if we learn from the survey that nobody really cares if the SCT and ICT have the same distribution. This is an area where I think the players, as a collective whole, need to step up. :-) If you want harder mythology questions, start answering the ones we do ask at a 90% clip.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:Every year on these surveys, I've selected the "Drastically Increase" (or whatever it's called) option for the Myth/Literature category and been dismayed that I didn't have the option to say "There's enough Lit, but there needs to be more Myth".
In cases like this, players should feel free to send us a free-form e-mail explaining exactly what you would like to vote for. One person requesting "drastically more" mythology won't change much, but if we get half-dozen such e-mails, we'd certainly try to formulate a survey question to address the issue. We can't make every change that people want, but, at a minimum, we'd want to know what people/certain segments want that we aren't giving them.
And I agree that the within-category distribution is pretty lopsided, though they seem to be improving the myth questions themselves since my first ICT as a frosh where we lost a bracket-deciding game to Valencia on a myth toss-up with wrong clues. I haven't seen this year's ICT set yet, but the myth at SCT this year, while pathetically low in number and biased in content area, was at least correct when it appeared, so that's something.
Can you e-mail NAQT the specifics of this at [email protected]? If I have the year correct, this error has never been reported and fixed in our packets.
As someone who has sat out of NAQT tournaments for years now to act as a reader instead, I would be more likely to play if RMP received its own distribution (combined with the suggestions called for by pretty much everyone to tighten the within-packet distribution). That said, I'm not sure if I'm the NAQT target audience anyway, so take my ideas with a grain of salt.
As a player, how could you tell whether or not NAQT had turned RMP into a separate category in its distribution files? And why would you care?

On the one hand, it would be relatively easy for NAQT to make this change and, if it would suddenly make many critics happy, there's no reason not to do it. I suspect, however, that nobody would actually be able to tell if we did make this change and that the complaints have a deeper (and more significant) origin. Something akin to "There should be more mythology questions" or "There should be more non-western mythology" questions rather than "RMP should be a top-level distribution category."
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

Brian Ulrich wrote:
setht wrote:Brian, do you feel that a science CE question has to have all clues/bonus parts related to current events, or would you be willing to get science CE tossups that (say) open with a couple sentences on recent stuff and then move on to science background that isn't necessarily based on recent events?
This has always been a tricky issue, and while I have no idea what other people at NAQT think about this, I tend to go by a sort of informal "how it feels" test. Another way of looking at it has been whether someone who has been in a coma for the past year or two could still answer it before the giveaway. As examples of other directions, though, the category has historically included diseases, endangered or newly extinct species, pollution, climate change, cloning and other genetic breakthroughs, and that sort of thing. For ICT, I tried to write on MRSA and the Wilkins Ice Shelf, but couldn't get either to work.
Historically, NAQT's current events questions have almost always included some number of non-CE clues; in fact, I think they usually grade from "really recent events" into "last couple of years clues" and finally into largely time-invariant material. I've certainly written a lot of CE bonuses where one of the parts had no CE content at all (but which might conceivably be better known due to its association with recent events).

It's hard to draw an exact line, but if one had a four-line tossup with two lines of "recent developments" followed by two lines of "textbook physics" clues, I suspect that, more often than not, it would end up filling the CE:S (current events, science) distribution rather than the physics distribution.

Ideally, I think, current events questions would have some sort of appeal to non-specialists. Seth's GRB example is a great one: There's something intrinsically interesting about the farthest/brightest object in the universe that would go along way toward justifying that as a CE question (rather than physics or astronomy). On the other hand, reporting that experiments have refined the parameters used in the liquid-drop model might always end up in physics even if the same number of clues of the same recency appeared in each.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:Also, I'm not positive if you're right about the fixed number. It could be, and I readily admit that you're in a better position to know than me, since I only have empirical evidence, but if so, the fixed number is precipitously low. In my room at SCT, I read every toss-up in every round except three rounds where I missed one (and looked at the one I missed), and the number, which I used to know but have since forgotten, so this could be off now, in the whole tournament was something like 4/3. I could swear I've seen more than that (though not by much) in previous years. It may be a fixed number that they discuss and change slightly each year though.
NAQT's packets do have a fixed amount of mythology (that is, a fixed fraction of the literature quota is mythology and the literature fraction has changed very little). There have been slight changes over the past four to five years, but nothing particularly major.
msaifutaa
Lulu
Posts: 37
Joined: Tue Feb 12, 2008 6:40 pm

Re: NAQT's Post-ICT Survey

Post by msaifutaa »

rhentzel wrote:Can you e-mail NAQT the specifics of this at [email protected]? 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 have just misspoken in a very odd way) but it claimed that Anubis was Isis's father. That's not just wrong, it's chronologically impossible since Anubis is the son of Nephthys, Isis's younger sister. There was another question somewhere else in the set that I'm 80% sure claimed that Asgard was a mead-hall but it might have claimed that Valhalla was a realm/plane--whichever way, it had it backwards.
As a player, how could you tell whether or not NAQT had turned RMP into a separate category in its distribution files? And why would you care?

On the one hand, it would be relatively easy for NAQT to make this change and, if it would suddenly make many critics happy, there's no reason not to do it. I suspect, however, that nobody would actually be able to tell if we did make this change and that the complaints have a deeper (and more significant) origin. Something akin to "There should be more mythology questions" or "There should be more non-western mythology" questions rather than "RMP should be a top-level distribution category."
As long as it passes the Turing test for RMP being its own distribution, I don't really care. I was under the impression, however, that there is no sub-distribution within packets (so if you keep Mythology as part of lit, then four or five lit in one packet could, in theory with bad luck, be mythology, and then no mythology for several more rounds, or something like that). If there's another option that puts Myth in there reliably and more often, I totally don't mind the way you do it. Heck, if you have to put Myth in with Science for whatever bizarre reason, if it increases Mythology questions and makes them more reliable in appearance, that's cool with me.
Mark
MIT
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

rhentzel wrote:As far as distribution issues go, changing subdistribution quotas (e.g. types of mythology) is more likely than "top-level" quotas (e.g., literature versus science), but we've certainly changed the latter before (including last year), so anything is possible.
I should update my original post: It wasn't last year (2006-2007) that we altered the distribution of top-level categories, but the year before.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

setht wrote:I'll certainly vote for more myth and philosophy in the survey; perhaps we can cut down some of the general knowledge? The Clio tossup was fine, but if it's being used as an excuse to put more myth content into the tournament, why not scale back the general knowledge and bump up the myth content? In general, my impression is that the last two ICTs have had substantial numbers of general knowledge questions that were primarily myth or philosophy or fine arts or social science or other academic subjects. I definitely prefer this to trivial general knowledge, but I'd happier getting a "pure myth" question on Clio without the stuff on the Clio Awards. I'm fine with leaving in a reduced number of general knowledge questions, and I hope they continue to consist primarily of mixed-subject academic and occasional trash questions, but as a writer I think it's easier to produce high-quality single-subject questions, and as a player I enjoy those more.
NAQT likes to use mixed-topic questions because it reduces the number of tossups (particularly at the ICT) on which people need to put down their buzzers because the question is simply outside of their field of expertise. Similarly, mixed-topic bonuses provide greater chances for a team to collaborate (and, hopefully, to avoid taking a zero for lack of a specialist).

The Clio question also included clues suggesting that David McCullough has never won one; if you know your history (and the basic fact that Clio was the muse of history) you can get it off that even if you know next to nothing about Clio's actual attributes in mythology.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:I think that there's definitely a large swathe of Mesopotamian and a reasonable amount of Aztec that's toss-upable (perhaps not quite enough--maybe Mayan and Incan can be thrown in here, though I'm not sure that the Mayans or the Incans really have very much toss-upable, which was why I initially threw them in other), though admittedly maybe not as much African (perhaps throw in Caribbean religions of African origin and you can get a reasonable selection for 1/1 though).
Is there good empirical evidence for this? NAQT's biggest concern when editing sets is choosing tossups and bonuses that are answerable (and not just on giveaways, etc.) In a field where questions on Bragi and Ptah go unanswered in one-third of the rooms (thereby completely failing to differentiate those teams), I'm a little leery of incorporating Incan or Mayan answers that almost everybody would agree are even harder.

If there are lots of mythology answers that we could count on being converted 85% of the time (and powered 15% of the time) by DI ICT teams, I don't think there's any doubt that we would look at adding such material as a required part of our distribution, at least at the ICT level.

I'm somewhat skeptical, but I've been proven wrong about such things before and am always willing to learn.

Alternatively, if you don't plan on playing NAQT in the future, send us a list of answers for us to digest and evaluate.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

trphilli wrote:As a moderator, I met some teams (not usually represented here) that had concerns about the current wording of the forfeit rule.

Their train of thought:

Would you prefer that in a situation where a team fails to arrive within five minutes of a stated game time that the first half begin at that time and the late team uses their time out to stop the game when they arrive compared to a straight forfeit?
As a tournament director, my concern with just beginning the game and letting the other team start playing when it arrives is that if the team has good reason to be late (i.e., it was my fault or they were assaulted in the street or something), I may very well want to consider requiring that the game be made up. If the packet has already been (partly) read, that might make it difficult to do.

What were the teams' complaints? Just that losing the entire game because they were, say, ten minutes late is a penalty that is incommensurate with the offense?
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

grapesmoker wrote:Hey NAQT people, on a totally unrelated note, can we have a question in the survey about whether the 2-day ICT should be shortened to one day? Personally, I find the Friday play hugely inconvenient and I can't be the only one.
As in past years, this will appear on our policy survey.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:That's not quite how ACF works, though. Since ACF is packet submission, it is the case that every submitted packet will have 1/1 Mythology, 1/1 Philosophy, 1/1 Religion. However, since ACF is not timed and instead stops at question 20, the questions after 20 will not be read. Thus, in, for instance, a 24/24 packet submission (reasonably typical), you get a 5/6 chance to read the mythology toss-up in a normal round. Now, even if not, it's still there--it might be used to break ties or as an extra question if the moderator screws up. I feel that this is a fundamentally different paradigm to not having set distribution requirements.
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 that any different, in effect, from saying that the tournament as a whole will have 15/15 mythology in 18 packets? That's identical in spirit to how NAQT's distribution works, we just have fewer mythology questions.

I agree that there's a chance that questions could be used for tiebreakers or as extras . . . but the chances of this happening seems very small and unlikely to greatly affect the total number heard. Something like 1 in every 50 games in NAQT's database has gone to tiebreaker questions; I doubt the percentage is much higher in ACF.

If NAQT included an extra mythology question in every packet (#27) to be used as the final tiebreaker, would that "really" count as an extra 1/0 myth per packet? I would assume not.

Perhaps more to the point, the 2008 ACF Nationals distribution included 1/1 "Extra Big Three" with the result that a packet of 24/24 could have 6/5 literature, 5/6 literature, or 5/5 literature. How is that fundamentally different from NAQT's model in which every packet can have 4/5, 5/4, or 5/5?
On a related note, I really dislike foreign language bonuses. And it's not that I don't know words in foreign languages--as a bonus, they usually just seem a cheesy way to get a free 30. This even from someone who enjoys trying to fraud the ends of legitimate academic topics (like John Chrysostom) by making up words from Latin or Greek.
I haven't tallied bonus conversions for the ICT yet; would you want to wager that the conversion on the LA:FL bonus was above the mean bonus conversion for the tournament by a statistically significantly amount?
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

As long as it passes the Turing test for RMP being its own distribution, I don't really care. I was under the impression, however, that there is no sub-distribution within packets (so if you keep Mythology as part of lit, then four or five lit in one packet could, in theory with bad luck, be mythology, and then no mythology for several more rounds, or something like that). If there's another option that puts Myth in there reliably and more often, I totally don't mind the way you do it. Heck, if you have to put Myth in with Science for whatever bizarre reason, if it increases Mythology questions and makes them more reliable in appearance, that's cool with me.
As currently implemented, a certain fraction of literature questions must be mythology and an attempt is made to distribute those as evenly as possible within the literature questions. It shouldn't be possible for a packet to ever have more than 1/1 mythology (not counting "soundalike" mythology clues/answers that may appear as part of literature questions on the <i>Aeneid</i>, say, or in "mixed-topic" bonuses).

By grouping it with literature, it means that the system will first assign a literature count for the packet (say 4/5) and then, within that, 1/0 will go to mythology, 0/1 to religious literature, and 3/4 to the rest of literature.

The benefit of doing it this way is that we won't have a case where the packet gets 5/5 literature (the maximum possible literature) AND also gets 1/1 mythology (the maximum possible mythology) making it heavy in both fields simultaneously. If it has 4/5 novels/poems/plays, a packet can only have 1/0 mythology.
User avatar
magin
Auron
Posts: 1005
Joined: Fri Oct 27, 2006 5:50 pm
Location: College Park, MD

Re: NAQT's Post-ICT Survey

Post by magin »

rhentzel wrote:
msaifutaa wrote:I think that there's definitely a large swathe of Mesopotamian and a reasonable amount of Aztec that's toss-upable (perhaps not quite enough--maybe Mayan and Incan can be thrown in here, though I'm not sure that the Mayans or the Incans really have very much toss-upable, which was why I initially threw them in other), though admittedly maybe not as much African (perhaps throw in Caribbean religions of African origin and you can get a reasonable selection for 1/1 though).
Is there good empirical evidence for this? NAQT's biggest concern when editing sets is choosing tossups and bonuses that are answerable (and not just on giveaways, etc.) In a field where questions on Bragi and Ptah go unanswered in one-third of the rooms (thereby completely failing to differentiate those teams), I'm a little leery of incorporating Incan or Mayan answers that almost everybody would agree are even harder.

If there are lots of mythology answers that we could count on being converted 85% of the time (and powered 15% of the time) by DI ICT teams, I don't think there's any doubt that we would look at adding such material as a required part of our distribution, at least at the ICT level.
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-interest to have less mythology in favor of other topics), I really enjoy hearing questions on myth. They're usually interesting (in a way that questions on chemistry or philosophy are not), and there's a large amount of answerable topics to write on. Also, I don't see why non-Greek or Roman mythology doesn't come up more often; there are many accessible answers there.
Jonathan Magin
Montgomery Blair HS '04, University of Maryland '08
Editor: ACF

"noted difficulty controller"
msaifutaa
Lulu
Posts: 37
Joined: Tue Feb 12, 2008 6:40 pm

Re: NAQT's Post-ICT Survey

Post by msaifutaa »

magin wrote: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-interest to have less mythology in favor of other topics), I really enjoy hearing questions on myth. They're usually interesting (in a way that questions on chemistry or philosophy are not), and there's a large amount of answerable topics to write on. Also, I don't see why non-Greek or Roman mythology doesn't come up more often; there are many accessible answers there.
Yeah, I was about to suggest exactly this (looking at what rooms had a neg and a then no answer as opposed to just no answer). It's probably also instructive to look at whether the question was during a prelim or a playoff round, since in the playoff round, I'd quite expect the very bottom bracket to have stuff go dead.

That said, to address rhentzel's earlier question, Bragi is (at least in my mind) reasonably difficult for Norse. I mean, back when I was much younger and Age of Mythology came out, Bragi was the only god from all three featured pantheons (Greek, Norse, Egyptian) who I didn't already know. I would expect certain low-difficulty deities from obscure mythologies to fare better than Bragi.
Mark
MIT
User avatar
bsmith
Tidus
Posts: 586
Joined: Thu Dec 09, 2004 10:33 am
Location: Ottawa, ON

Re: NAQT's Post-ICT Survey

Post by bsmith »

rhentzel wrote:
trphilli wrote:As a moderator, I met some teams (not usually represented here) that had concerns about the current wording of the forfeit rule.

Their train of thought:

Would you prefer that in a situation where a team fails to arrive within five minutes of a stated game time that the first half begin at that time and the late team uses their time out to stop the game when they arrive compared to a straight forfeit?
As a tournament director, my concern with just beginning the game and letting the other team start playing when it arrives is that if the team has good reason to be late (i.e., it was my fault or they were assaulted in the street or something), I may very well want to consider requiring that the game be made up. If the packet has already been (partly) read, that might make it difficult to do.

What were the teams' complaints? Just that losing the entire game because they were, say, ten minutes late is a penalty that is incommensurate with the offense?
I obviously was not there, but I'm assuming this incident was Toronto's DII forfeit. They likely thought that a game could start and a team could join in progress. We have done this in some independent Canadian events, but only when McGill is late in the early rounds because they haven't arrived yet. The game starts with one team only, and if the late team arrives during the first half, they have some ground to make up. If the late team arrives after the half (or not at all), it's a forfeit. This was intended to penalize tardiness, not encourage good teams to take a break and show up late for easy opponents. Certainly this behaviour would not be acceptable at a more formal tournament. I apologize if they blame me for instilling this idea of showing up late, but I would never recommend they try it at a national tournament.
Ben Smith
Ottawa '08 & '10
msaifutaa
Lulu
Posts: 37
Joined: Tue Feb 12, 2008 6:40 pm

Re: NAQT's Post-ICT Survey

Post by msaifutaa »

rhentzel wrote: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 that any different, in effect, from saying that the tournament as a whole will have 15/15 mythology in 18 packets? That's identical in spirit to how NAQT's distribution works, we just have fewer mythology questions.

Perhaps more to the point, the 2008 ACF Nationals distribution included 1/1 "Extra Big Three" with the result that a packet of 24/24 could have 6/5 literature, 5/6 literature, or 5/5 literature. How is that fundamentally different from NAQT's model in which every packet can have 4/5, 5/4, or 5/5?
I definitely don't want to speak for the editors here, but by inspection, it seems to me that mythology is not often dropped from the first 20 (and when it is, I always notice), perhaps because the editors realise that this is the only myth toss-up for the round (it seems that often the Extra Big Three are used to pick out the best Big Three and drop the rest, or in case of repeats, which are more common in the Big Three simply because there are more questions in them). When I've seen mythology pulled out, it almost always seems to be over a repeat or bad writing. On this, however, I could be totally wrong.
I haven't tallied bonus conversions for the ICT yet; would you want to wager that the conversion on the LA:FL bonus was above the mean bonus conversion for the tournament by a statistically significantly amount?
I dunno about averages--I can't speak for other teams. I have a feeling that they are closer to the extremes though. It just doesn't feel like an accomplishment to get those points. For instance, some time ago I came across a foreign language bonus in Russian (maybe not even NAQT, it was back quite a bit), and one player from Russia 30ed it, but he said he felt dirty doing it, rather than the good feeling you get from 30ing a tough academic bonus due to deep knowledge.
Mark
MIT
User avatar
trphilli
Lulu
Posts: 37
Joined: Thu Feb 10, 2005 3:12 pm

Re: NAQT's Post-ICT Survey

Post by trphilli »

rhentzel wrote: What were the teams' complaints? Just that losing the entire game because they were, say, ten minutes late is a penalty that is incommensurate with the offense?
Yeah the feeling was for minor tardiness they were deprived of an entire game. And theoretically they wanted the chance to mount a comeback.
Tom
Univ of Illinois '05 & '06
Lincoln-Way Community High School '01
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

magin wrote: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-interest to have less mythology in favor of other topics), I really enjoy hearing questions on myth. They're usually interesting (in a way that questions on chemistry or philosophy are not), and there's a large amount of answerable topics to write on.
I won't say that conversion percentage is NAQT's gold standard for answerability, but it's pretty close. I mean, it was the actual measured value under match conditions among the DI teams. Yes, I suspect Maryland could have answered Bragi from the giveaway clues (but negged themselves out of it), but their opponent clearly couldn't.

A dead tossup doesn't differentiate teams at all (or perhaps only marginally by awarding one a -5); a tossup that goes dead in 40% of the rooms is not pulling its weight in terms of differentiation, regardless of the number of the interrupts.
Also, I don't see why non-Greek or Roman mythology doesn't come up more often; there are many accessible answers there.
I think the reason is that NAQT, in its collective person, disagrees that there are many accessible answers out there. Maybe we're wrong; that's certainly possible, but those tossup conversions for Bragi and Ptah reinforce my belief that non-classical mythology is not as well known as some people think (and that classical mythology is quite well known).

If we were writing the ICT for the top eight teams, it would probably look a lot different. If the worst team in the tournament is Harvard or UC-Irvine, a lot more areas and answers become tossup-worthy.

If that doesn't seem fair, sign up as a writer and start sending us tossups at our acceptability threshold that are so interesting that we simply can't reject them; if we measure their conversion and find that they are actually hitting 85%, I doubt a single person in NAQT would argue that there shouldn't be a place for them in our distribution and at our tournaments.
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:Yeah, I was about to suggest exactly this (looking at what rooms had a neg and a then no answer as opposed to just no answer). It's probably also instructive to look at whether the question was during a prelim or a playoff round, since in the playoff round, I'd quite expect the very bottom bracket to have stuff go dead.
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 it really matters how many of the rooms were "neg and no answer" and how many were simply "no answer"; in either case, the tossup is (probably) not doing its job of distinguishing which team knows more about the subject.
That said, to address rhentzel's earlier question, Bragi is (at least in my mind) reasonably difficult for Norse. I mean, back when I was much younger and Age of Mythology came out, Bragi was the only god from all three featured pantheons (Greek, Norse, Egyptian) who I didn't already know. I would expect certain low-difficulty deities from obscure mythologies to fare better than Bragi.
This surprises me somewhat; anecdotally, I would have said that Bragi was quite easy as a Norse tossup answer. Not as easy as Odin or Thor, obviously, but I would have expected a DI ICT conversion of over 90% for "Name this Norse god of poetry."

That said, it's nice to be talking with another Age of Mythology player . . .
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

bsmith wrote:I obviously was not there, but I'm assuming this incident was Toronto's DII forfeit. They likely thought that a game could start and a team could join in progress. We have done this in some independent Canadian events, but only when McGill is late in the early rounds because they haven't arrived yet. The game starts with one team only, and if the late team arrives during the first half, they have some ground to make up. If the late team arrives after the half (or not at all), it's a forfeit. This was intended to penalize tardiness, not encourage good teams to take a break and show up late for easy opponents. Certainly this behaviour would not be acceptable at a more formal tournament. I apologize if they blame me for instilling this idea of showing up late, but I would never recommend they try it at a national tournament.
I actually agree that this seems like a classier and less draconian way to handle it . . . if you know that the team is late of its own volition and is deserving of (some degree of) chastisement. Without knowing that, however, I think it's probably safer to tentatively declare a forfeit and save the packet to make sure the questions are available if the round does need to be replayed in the name of fairness.

Possibly this is an area where better communication with teams could help. If a tournament director had cell phone numbers for all teams and could ascertain that they were running late for their own fault, then perhaps the "start the room and let them start playing when they show up" policy could work out.
msaifutaa
Lulu
Posts: 37
Joined: Tue Feb 12, 2008 6:40 pm

Re: NAQT's Post-ICT Survey

Post by msaifutaa »

rhentzel wrote: 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 it really matters how many of the rooms were "neg and no answer" and how many were simply "no answer"; in either case, the tossup is (probably) not doing its job of distinguishing which team knows more about the subject.
Weird about the conversion not changing much. That's pretty cool.

As for "not doing its job of distinguishing which team knows more", I would say that the case of an overeager neg (a neg where Team A negs and Team B doesn't even know the answer but Team A would have known it later) shouldn't count against the question. In fact, that means that the question was otherwise a very good way to distinguish between the better team and the worse team, since only the better team knew it--the neg took that away though. It's arguable that when a second team barely bounces back a giveaway clue after an overeager buzzer on the first team nearly powers the question, that in that case the team with greater knowledge in the subject didn't get the toss-up (though of course, neg and lockout is a very good system--that's the first team's punishment for losing a gamble).
This surprises me somewhat; anecdotally, I would have said that Bragi was quite easy as a Norse tossup answer. Not as easy as Odin or Thor, obviously, but I would have expected a DI ICT conversion of over 90% for "Name this Norse god of poetry."

That said, it's nice to be talking with another Age of Mythology player . . .
You probably don't want this thread to become a list of all the Norse things that I think are easier than Bragi (lest it also become an answer list for memorisers or something), but he isn't, for instance, on the You Gotta Know list (or even mentioned on it, as some others are who don't make the list), and I think I could probably rattle off at least thirty answers for Norse mythology that I think are better known than Bragi. And it's not that I don't know Bragi. As a myth-lover, my team would expect me to know things of Bragi difficulty (and higher). But unlike, say, Odin, I wouldn't expect my other teammates to know him.

What I'm saying, I guess, is that failure to convert Bragi is not a condemnation of Norse myth as too hard--Bragi is just difficult enough (not really something I would call super difficult or even close to that, but comparatively harder than the ones that got 100% conversion) that I would expect something like the numbers you gave me.

Also, keep in mind (probably just a quick oversight on your part) that your list of Classical myth that 100% of people got right includes Marduk. Marduk is Mesopotamian, and I'm not surprised that 100% of rooms answered it--if anything, Marduk is good proof that you can do non-Classical mythology, whereas Bragi is more like proof that the field lacks knowledge of Norse myth to a certain depth. If you had asked, for instance, a toss-up on the dude who put Telemachos in front of Odysseus's plow to prove Odysseus wasn't insane and make him go to the Trojan War, I bet despite that story being relatively common that you would have seen 20% or lower conversion.
Mark
MIT
Rayford Smuckles
Lulu
Posts: 13
Joined: Tue Feb 07, 2006 2:44 am
Location: California

Re: NAQT's Post-ICT Survey

Post by Rayford Smuckles »

magin wrote:
Maryland negged the Bragi tossup, so it went unanswered, but we would have answered it had we not negged
Not that this necessarily takes away from your point, but I was actually the person who negged Bragi in that match, in a buzzer race with Chris Ray somewhere around the "cup," clue, I believe. (I remember it pretty clearly because I stupidly and inexplicably blanked on the answer and got timed out) The tossup then went to the end and the guy on your team who I didn't know (Jeremy, presumably) buzzed in and said something incorrect. In a way, this is even more interesting, as it shows that there can be situations where people on both teams know the answer and yet it just doesn't get entered into the stats as an answered tossup for some reason. Over an entire tournament's worth of results, stuff like this is going to happen with enough frequency to where percentages on individual tossups should be taken with a grain of salt, or at least interpreted as a fairly flexible number (i.e. anything from 60-80% conversion is at a pretty much equivalently acceptable difficulty.)

MFS
Mike

Stanford
rhentzel
Rikku
Posts: 260
Joined: Thu May 15, 2003 4:20 pm
Location: Minneapolis, MN
Contact:

Re: NAQT's Post-ICT Survey

Post by rhentzel »

msaifutaa wrote:As for "not doing its job of distinguishing which team knows more", I would say that the case of an overeager neg (a neg where Team A negs and Team B doesn't even know the answer but Team A would have known it later) shouldn't count against the question. In fact, that means that the question was otherwise a very good way to distinguish between the better team and the worse team, since only the better team knew it--the neg took that away though. It's arguable that when a second team barely bounces back a giveaway clue after an overeager buzzer on the first team nearly powers the question, that in that case the team with greater knowledge in the subject didn't get the toss-up (though of course, neg and lockout is a very good system--that's the first team's punishment for losing a gamble).
We don't currently track "overeager negs," and I'm not sure that we could even if we wanted to. NAQT attempts to correct for this phenomenon by pegging our desired real-life conversion at 85%, but saying that we'd like 90% of our tossups to be answered if read to completion in all rooms. The 5% discrepancy is intended to account for exactly this sort of case: random attrition due to overeager negs, moderator mispronunciations, spaced-out players, and so on.
User avatar
magin
Auron
Posts: 1005
Joined: Fri Oct 27, 2006 5:50 pm
Location: College Park, MD

Re: NAQT's Post-ICT Survey

Post by magin »

Rayford Smuckles wrote:Not that this necessarily takes away from your point, but I was actually the person who negged Bragi in that match, in a buzzer race with Chris Ray somewhere around the "cup," clue, I believe. (I remember it pretty clearly because I stupidly and inexplicably blanked on the answer and got timed out) The tossup then went to the end and the guy on your team who I didn't know (Jeremy, presumably) buzzed in and said something incorrect. In a way, this is even more interesting, as it shows that there can be situations where people on both teams know the answer and yet it just doesn't get entered into the stats as an answered tossup for some reason. Over an entire tournament's worth of results, stuff like this is going to happen with enough frequency to where percentages on individual tossups should be taken with a grain of salt, or at least interpreted as a fairly flexible number (i.e. anything from 60-80% conversion is at a pretty much equivalently acceptable difficulty.)
Yeah, you're right about that. Sorry; it indeed happened the way you describe.

Also, in theory, teams studying for ICT could either decide not to learn non-Western mythology, since it makes up so small a part of the distribution, or not be exposed to many commonplace answers in non-Western myth due to their relative paucity in NAQT. Having a relatively small distribution in certain areas (such as, for instance, non-Western myth) may lead to a self-fulfilling prophecy where teams at ICT are less exposed to those areas, decreasing their conversion rates.
Jonathan Magin
Montgomery Blair HS '04, University of Maryland '08
Editor: ACF

"noted difficulty controller"
User avatar
grapesmoker
Sin
Posts: 6345
Joined: Sat Oct 25, 2003 5:23 pm
Location: NYC
Contact:

Re: NAQT's Post-ICT Survey

Post by grapesmoker »

I think the idea behind separating myth and literature is in order to create space for both distributions. mACF formats allow more space for things "traditionally" regarded as literature (I put that in quotes because you can talk about the Bible as literature, the Iliad, etc.) and more space for myth. By the way, RMP is just a supercategory that sort of got built up because it groups religion and myth (which are close) and philosophy probably more appropriately belongs in the social science distribution or something of the sort. Anyway, what matters is not the labels but how much representation each category gets. Since literature is a "big three" category, it gets equal representation with history and science, which I think is right. I think both the myth and the philosophy distributions need to be expanded at ICT, and the myth needs to be much more varied. Right now the emphasis is too much on classical stuff and not enough on things in other cultures.
Jerry Vinokurov
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
User avatar
setht
Auron
Posts: 1205
Joined: Mon Oct 18, 2004 2:41 pm
Location: Columbus, Ohio

Re: NAQT's Post-ICT Survey

Post by setht »

rhentzel wrote:
ToStrikeInfinitely wrote:...why not have a more extensive non-western myth distribution, as most other tournaments do?
Basically out of fears that non-classical, non-Norse, non-Egyptian myth would, broadly speaking, prove too difficult to meet our conversion targets. Or, conversely, that requiring more of it while meeting our conversion targets would result in endless repetition of the same answers.

For instance, at the DI ICT only 60% of the rooms converted Bragi (9/15), 63% (7/11) converted Ptah, and 84% (11/13) converted Jimmu Tenno. 100% converted Eumenides, Marduk, Medea, and Paris, and 92% converted Cassandra.

If we were seeing 90% conversion rates on all of those answers, NAQT would have no problems whatsoever delving into harder mythoi; we have no objection to them on principle, particularly if we learn from the survey that nobody really cares if the SCT and ICT have the same distribution. This is an area where I think the players, as a collective whole, need to step up. :-) If you want harder mythology questions, start answering the ones we do ask at a 90% clip.
Like Maryland, Chicago negged on Bragi and watched it die, but the conversion statistics are still clearly lower than the goal, and I think it's reasonable to want tossups where more teams can pick up answers at the end if their opponents neg.

Having said that, I don't think you're comparing apples and apples, R. The Eumenides, Medea, Paris and Cassandra are all really well-known characters in Greek myth, appearing in famous works of literature/art (which makes them great subjects for tossups). Marduk is a similarly prominent figure in Babylonian myth. It's not at all surprising that these tossups had great conversion rates. Bragi is not the Norse equivalent; he's a second- or third-tier character. The same holds for Ptah in Egyptian myth. There are clearly much fewer well-known characters in Norse myth or Egyptian myth than in classical myth. Given your tossup conversion goals, this clearly leads to a distribution with less Norse than classical myth, less Egyptian than classical myth, etc. I think everyone's fine with this. The thing is, I think there are enough well-known answer choices in non-classical myth as a whole to allow having more of it than the 5/4 that appeared in this year's ICT set. If you want a bunch of examples, I wrote 155 non-classical myth tossups (well, I guess 13 of those are mixed, and some of those include classical clues) for a myth singles tournament back in 2005. Some of those answers probably wouldn't get 90% conversion (e.g., my own tossup on Ptah), but I think a large fraction would.

Apparently this year's ICT set had 9/9 myth in 18 packets, half classical and half non-classical. As I said earlier in this thread, I'd like to see something like 12/12 to 15/15 myth in an 18 packet set. I'm okay with half classical and half non-classical, but let's say you decide to do slightly more non-classical than classical. If you jump up to 15/15 myth, with 6/6 classical and 9/9 non-classical, I think there are enough well-known non-classical answers out there to furnish 9/9 at SCT and ICT without cycling through the same answers too frequently, especially if you encourage multiple-mythos common link questions (and if you have these, there's no reason not to have questions that start with harder non-classical clues and eventually get to better-known classical clues, if that'll help conversion statistics). In fact, I'd advocate something like 6/6 classical, 6/6 non-classical, and 3/3 whatever (can be multiple-mythos) for a set of 15/15 myth.

This may not be true in DII--I don't know what they're teaching kids in schools these days--but I'm pretty sure it's the case in DI, and I think there's a fairly sizable portion of the (DI) circuit that would like to see more myth in general and more non-classical myth in particular.

I don't think people are necessarily calling for harder myth tossups, or that non-classical myth tossups are inherently harder. If next year you had a Norse tossup on the Norns, an Egyptian tossup on Horus, and a Greek tossup on Silenus, I'm pretty sure the conversion numbers would be pretty much the opposite of this year's numbers. I think everyone calling for more non-classical myth is fine with getting tossups on the Norse/Egyptian/Aztec/whatever equivalent of Medea.

-Seth
Seth Teitler
Formerly UC Berkeley and U. Chicago
President of NAQT
Emeritus member of ACF
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: NAQT's Post-ICT Survey

Post by Mechanical Beasts »

setht wrote: This may not be true in DII--I don't know what they're teaching kids in schools these days--but I'm pretty sure it's the case in DI, and I think there's a fairly sizable portion of the (DI) circuit that would like to see more myth in general and more non-classical myth in particular.
I'm not terribly good at myth, to be honest, but I think to speak for DII to a certain extent--why not treat the difference between DI/DII myth the same way one does other things? Instead of asking for Mjollnir, ask for the dude who wielded Mjollnir, Thor. A tossup on Mjollnir is going to see at least 90% DI conversion, and I'd say that a tossup on Thor should get the same, if not better.

A lot of DII players are going to be DI players someday, so they might as well hear the same myth distribution as the DI players, to prepare them for their DI future, right? Otherwise, you could make the argument that there shouldn't be non-Western lit in DII if it doesn't get converted enough, or that there shouldn't be organic chemistry (noooooooooooo!) because there'll be a lot of DII teams with a lot of freshmen, and even if they have a chem major, not every chem major takes a first-year orgo course as a freshman.

Even if you get worse conversion--or even if you choose to give DII the same Mjollnir tossup--the DII players will get better from hearing the whole tossup, then, and the next match they play, they'll remember that Mjollnir is Thor's hammer, or that Bragi's the Norse god of poetry, and so forth. They might not know harder clues (yet), but they'll know something.
Andrew Watkins
Locked