2009 ICT discussion

[cvdwightw]

I think I get what Matt is saying.

When "muh-NAY" requires a spelling, any of four situations can happen: the answer's Monet, the player meant Monet, the moderator heard Manet; the answer's Manet, the player meant Manet, the moderator heard Monet; the answer's Manet, the player meant Monet, the moderator heard Monet; the answer's Monet, the player meant Manet, the moderator heard Manet. In two of these situations, the player gets points he wouldn't have otherwise gotten; in the other two, the player isn't clued into anything other than that the answer's either Manet or Monet, which, as he's buzzing in with the wrong one of those two choices, won't help him get points anyway. Therefore, we end up rewarding real knowledge, but not rewarding lack thereof.

Matt is arguing that "pom-pah-doo" only occurs in the following situations: the answer's Pompadour, the player meant Pompidou, the moderator heard Pompidou; the answer's Pompadour, the player meant Pompadour, the moderator heard Pompidou. If we accept Matt's argument, then by requiring spelling, the moderator automatically conveys to the player that the correct answer must be Pompadour - if the answer was Pompidou, we wouldn't require spelling. Therefore, we run into the double dillemma of possibly punishing real knowledge (if we just outright disallow anything that does not clearly have an r sound at the end, which may actually disadvantage native speakers of French), or possibly rewarding a player for lack of real knowledge (player figures out why he's being prompted for spelling, realizes his first answer was wrong, changes answer). I'm not sure that prompting is a reasonable alternative, either - it's not like the two possible answers are in any way related to each other, so prompting might result in hilariously bad "what does this prompt want" attempts at divination.

[jagluski]

For benefit of those who did not attend the ICT: NAQT had exactly 12 lower- and lobby-level meeting rooms (we were promised 13 when we signed on with the hotel; it's unclear to me how 13 became 12). Rather than split up a pod, we decided to have games in those 12 rooms, and a control room on the 11th floor. All of DII, plus 25% of DI, were on the 11th and 12th floors. We decided that the least problematic way to handle scoresheet intake and packet distribution for those 12 rooms was for one of the 12 to be the point of contact for all 12. I had two staff with me (by the way, thank you from the bottom of my heart, Alabama people), one who kept score and one who stood outside the door to prevent the scoresheet/packet exchange from interrupting games. For obvious reasons I think NAQT should never again resort to this if it can be at all avoided, though in general it worked.

I'd like to note that there was also a DII room on the 6th floor, and one on the 5th. I know it blew for me on the 6th floor to have my stats room on the 11th, and it was probably even worse for whoever was on the 5th floor. This logistically was pretty lame, although being able to call up to the stats room from our room when we had the crazy D2 teams going to the wrong room madness was useful.

Moving to that, was it totally impossible to print a new schedule for DII? The playoffs schedule confused a ton of teams who were in the bracket with the byes, and I know at least one game had to be re-read...

There was a room on the 5th floor, but no games were held in it. Your room was the only room with games not on the lower levels, 11th, or 12th floors. We were supposed to have all the rooms on the 11th or 12th floor, but we discovered during a pre-tournament walkthrough that all the rooms weren't soundproof, even though we had previously told the hotel this was needed. Thus, we had to go to a room on the 6th floor.

I also admit that I made an error in forgetting to tell teams where the BYE team would be (8R on the grid) and that new teams would have a bye in the playoff rounds that weren't made aware of the fact. The staff should have received schedules for their game rooms and definitely will in the future.

[Important Bird Area]

I'd be curious to see the conversion rates of the Pendergast tossup. That strikes me as very hard.

Digging this up from page 1 now that we have conversion stats.

This was converted in 3 out of 15 rooms, which is enough to conclude "too hard for ICT, should have been the third part of a bonus."

[grapesmoker]

Thanks to Jeff for posting this tossup. I'd like to contrast this question with a Hegel tossup that I wrote for T-party earlier in this academic year.

Augusto Vera worked to popularize this man in France in the 19th century, while an Italian school devoted to him was led by Bertrando Spaventa. During the 1930s, an influential course on the correct reading of this thinker was delivered in Paris by Alexandre Kojeve. Moncure Conway was a leader of the (*) Cincinnati school devoted to this thinker, though the St. Louis school is better known. For 10 points--name this German philosopher who wrote The Science of Logic and The Phenomenology of Spirit.

This thinker criticized Newton and argued that Plato was right in holding in the Timmaeus that there could be no planet between Mars and Jupiter in a doctoral dissertation entitled De Orbitis Planetarum. This thinker interpreted the miracles of Jesus in a naturalistic way in his early work “The Life of Jesus,” and he distinguished between natural and positive religion, the latter of which he viewed as authoriatrian, in his The Positivity of the Christian Religion. J.M.E. McTaggart wrote a notable commentary on this man's Science of Logic, which constituted the first part of his proposed Encyclopedia that would describe his philosophical system, while a more famous work by this man describes the state as “the actuality of the ethical idea,” and “the march of God in the world.” For ten points, identify this philosopher who detailed the rise of the master and slave moralities and outlined the dialectical process in history in such works as Phenomenology of Mind and Philosophy of Right.
Answer: Georg Friedrich Hegel

I want to compare these two questions because I think such a comparison can shed some light on a problem people often have with NAQT questions, and that's the absence of middle clues. Consider the first Hegel tossup; I believe it was written by Andrew, and Andrew obviously knows his Hegel, I'm sure. In terms of how the clues are arranged, it's clearly pyramidal, but the problem comes in the transition between the harder and easier parts of this tossup. The St. Louis Hegelians are probably well known to people who have read up on their Hegel popularizers, but I would argue that they're still a relatively difficult clue for most of the teams in the field. While I'm sure that the top several teams can buzz there, most teams are going to experience an abrupt drop in difficulty between "St. Louis school" and "The Science of Logic." I'm not going to claim that my own question is some paradigmatic tossup on Hegel (in retrospect I think the clue on "Science of Logic" came a little early) but I think what the absence of the character limit allows me to do is to put in a lot more transitional clues; for example, the description of the state and the master/slave dialectic are certainly recognizably Hegelian things (which I thought should go after the title of "Science of Logic") and putting those kinds of clues in allows someone who is familiar with some of the content of Hegel's work to buzz before someone who just knows the two most famous titles by Hegel.

I'm pointing this out not because I think I can write a better Hegel tossup than Andrew but rather because I think the character limit contributes significantly to the creation of these types of questions. It's great if you're good enough that you know about Hegel from his popularizers, but what if you're not (even though you might be familiar with the contents of Hegel's works)? The NAQT tossup essentially fails to reward that, and it's not the fault of the writer, it's the fault of the format restriction. I've played two NAQT-style tournaments this year (SCT and FICHTE, and many more in my career) and I see this issue come up repeatedly: half the question will be something hard that most teams can't get, and the second half ends up triggering buzzer races. People have made some abstract arguments about how this is the fault of the implementation rather than the theory, but in my experience, the empirical evidence is too strong for that view to stand up. It's simply a problem that arises much more often in NAQT events than in any other ones.

I personally hold, as I've continued to hold for many years now, that the character limit and its equally annoying attendant the clock are both anachronisms of a bygone age. Literally no one but NAQT employs these things, and I think I've presented some useful evidence for why at least one of them should be abandoned. I would also suggest that too many people are fixated on preserving some kind of difference between NAQT and the rest of the circuit (which is overwhelmingly, perhaps entirely, mACF in format) for reasons which most of the time reduce to being different for the sake of difference (this includes things like the distribution as well). My own belief is that NAQT ought to move to a soft character limit which should be increased to probably about 600 or 650 characters (my tossup above should be relatively easily reducible to this count) and eliminate the clock.

[yoda4554]

Thanks to Jeff for posting this tossup. I'd like to contrast this question with a Hegel tossup that I wrote for T-party earlier in this academic year.

Augusto Vera worked to popularize this man in France in the 19th century, while an Italian school devoted to him was led by Bertrando Spaventa. During the 1930s, an influential course on the correct reading of this thinker was delivered in Paris by Alexandre Kojeve. Moncure Conway was a leader of the (*) Cincinnati school devoted to this thinker, though the St. Louis school is better known. For 10 points--name this German philosopher who wrote The Science of Logic and The Phenomenology of Spirit.

This thinker criticized Newton and argued that Plato was right in holding in the Timmaeus that there could be no planet between Mars and Jupiter in a doctoral dissertation entitled De Orbitis Planetarum. This thinker interpreted the miracles of Jesus in a naturalistic way in his early work “The Life of Jesus,” and he distinguished between natural and positive religion, the latter of which he viewed as authoriatrian, in his The Positivity of the Christian Religion. J.M.E. McTaggart wrote a notable commentary on this man's Science of Logic, which constituted the first part of his proposed Encyclopedia that would describe his philosophical system, while a more famous work by this man describes the state as “the actuality of the ethical idea,” and “the march of God in the world.” For ten points, identify this philosopher who detailed the rise of the master and slave moralities and outlined the dialectical process in history in such works as Phenomenology of Mind and Philosophy of Right.
Answer: Georg Friedrich Hegel

I want to compare these two questions because I think such a comparison can shed some light on a problem people often have with NAQT questions, and that's the absence of middle clues. Consider the first Hegel tossup; I believe it was written by Andrew, and Andrew obviously knows his Hegel, I'm sure. In terms of how the clues are arranged, it's clearly pyramidal, but the problem comes in the transition between the harder and easier parts of this tossup. The St. Louis Hegelians are probably well known to people who have read up on their Hegel popularizers, but I would argue that they're still a relatively difficult clue for most of the teams in the field. While I'm sure that the top several teams can buzz there, most teams are going to experience an abrupt drop in difficulty between "St. Louis school" and "The Science of Logic." I'm not going to claim that my own question is some paradigmatic tossup on Hegel (in retrospect I think the clue on "Science of Logic" came a little early) but I think what the absence of the character limit allows me to do is to put in a lot more transitional clues; for example, the description of the state and the master/slave dialectic are certainly recognizably Hegelian things (which I thought should go after the title of "Science of Logic") and putting those kinds of clues in allows someone who is familiar with some of the content of Hegel's work to buzz before someone who just knows the two most famous titles by Hegel.

I'm pointing this out not because I think I can write a better Hegel tossup than Andrew but rather because I think the character limit contributes significantly to the creation of these types of questions. It's great if you're good enough that you know about Hegel from his popularizers, but what if you're not (even though you might be familiar with the contents of Hegel's works)? The NAQT tossup essentially fails to reward that, and it's not the fault of the writer, it's the fault of the format restriction. I've played two NAQT-style tournaments this year (SCT and FICHTE, and many more in my career) and I see this issue come up repeatedly: half the question will be something hard that most teams can't get, and the second half ends up triggering buzzer races. People have made some abstract arguments about how this is the fault of the implementation rather than the theory, but in my experience, the empirical evidence is too strong for that view to stand up. It's simply a problem that arises much more often in NAQT events than in any other ones.

I personally hold, as I've continued to hold for many years now, that the character limit and its equally annoying attendant the clock are both anachronisms of a bygone age. Literally no one but NAQT employs these things, and I think I've presented some useful evidence for why at least one of them should be abandoned. I would also suggest that too many people are fixated on preserving some kind of difference between NAQT and the rest of the circuit (which is overwhelmingly, perhaps entirely, mACF in format) for reasons which most of the time reduce to being different for the sake of difference (this includes things like the distribution as well). My own belief is that NAQT ought to move to a soft character limit which should be increased to probably about 600 or 650 characters (my tossup above should be relatively easily reducible to this count) and eliminate the clock.

I don't think the character limit is really what's causing a supposed lack of middle clues here. Consider, for instance, this question I wrote without a cap limit for the last lit singles:

Jack Lynch edits the annual devoted to the “Age of” this writer. One of his essays claims men should be jarred from procrastination by the “certainty that life cannot be long, and the probability that it will be much shorter than nature allows,” while another ridicules the education of critic Dick Minim. His major critical work claims Lycidas had “no art, for there is nothing new,” begins with an essay on Richard Savage, and popularized the term “metaphysical poetry” in a section on Cowley. He commented that “Nothing odd will do long” regarding Tristram Shandy and wrote that Shakespeare lacked a “moral purpose,” because he was a “poet of nature,” in the Preface to his 1765 edition of the Bard’s work. For 10 points, name this author of the Lives of the English Poets, who compiled a Dictionary of the English Language.
ANSWER: Samuel Johnson

Now, as it happened, I wrote this questions this way because we were short a criticism question, short on 18th century, and because within English Johnson's critical work is generally more important than its quiz bowl reputation. However, this is a question where we have almost none of the traditional Johnson stuff--nothing from his poetry, Rasselas, Boswell's biography, etc. We have, instead, a lot of significant, oft-quoted, very hard clues about his criticism. Looking at this, I don't think the right evaluation of it is that I didn't put the right middle clues into a Samuel Johnson tossup, even though I had the room--it's that I wrote a tossup on Johnson's criticism, which is going to just be pretty hard most of the way. A lack of character limit had no influence here.

I think the same things going on with the Hegel tossup. This does not seem to me to be a Hegel tossup without middle clues--it's a tossup on Hegelians, which, though (I assume) it is a very significant topic all the way down to philosophy people, is just going to be a hard tossup for people working from traditional Hegel canon fodder. I also think that it's fine to have a few tossups like this at nationals.

[Important Bird Area]

Dave, I think Jerry's argument is that the tossup with Hegelians is so hard that it produces a cliff on the well-known works at the end. By contrast, your Samuel Johnson tossup lacks a cliff (if it lacks a cliff? I don't actually know anything about this one) because players at a lit singles can be expected to know something about Johnson's criticism, and that the distribution of buzzes would therefore be better. Similarly, raising the character limit to 600 (or whatever) would have hypothetically allowed the insertion of an additional middle clue about Hegel's works while preserving all of Andrew's hard clues about various kinds of Hegelians in the leadin (whereas restructuring it under the existing limit requires a choice of what to cut; it's 497 characters long, almost exactly half the length of Jerry's tossup.)

[matt979]

That should be allowed. You can't expect people to pronounce Shona correctly, especially the whistled sibilants. Thinking that the weird [ts] glide would be marginalized is rational.

I've been thinking about this, in the context of the guideline that "a plausible or phonetic pronunciation is usually acceptable, unless it demonstrates a fundamental lack of understanding about the correct answer," and by comparison to Pompadour.

Is it fair to characterize this as more of a by reasonable people, for reasonable people point than a deep linguistic analysis? I don't mean to marginalize the latter; rather, I'm wondering about the threshold level of linguistic expertise needed reliably to resolve pronunciation protests at a national championship.

When we're talking about plausible and/or phonetic, I think general linguistic principles are more applicable than specific language rules -- even for the latter, if the answer given isn't technically correct (for example, the CHAHN-gih-R-EYE consensus) then in principle we're talking about what we reasonably expect a speaker of (American) English to grok.

For example, even it's incorrect, is it at all rational for someone who speaks no French (but has passing familiarity with words of French origin like "ballet" or "bourgeois") to think that the 'r' at the end of a French name is sometimes silent?

(I guess trailing r's could also plausibly present a problem of American dialect, though I assume the actual case I don't think anyone had the Boston accent in mind.)

[Birdofredum Sawin]

Yeah, for what it's worth, I did in fact write that "Hegel" tossup, and wrote it as a tossup on "schools of Hegelian thought which will have the name 'Hegel' as the answer" (for reasons of answerability which have been covered recently in the "common link" thread).

My thinking behind the question was that the first sentence would be very very hard; the second sentence would be hard to QB people but is important and gettable for philosophy people (Kojeve's Hegel interpretation was a huge influence on both Leo Strauss and on French philosophers of the second half of the 20th century); the third sentence, though less important in the history of philosophy, is one which I thought would be better known to QB people (the St. Louis Hegelians have -- I thought, perhaps mistakenly -- come up with some frequency in ACF tournaments of the past decade). In retrospect, perhaps the last judgment was wrong, and this would have been better as a tossup whose third sentence modulated to a discussion of Hegel himself (e.g. "This man talked about [more or less Hegelian sounding things] in his book The Science of Logic. For 10 points--name ...").

I actually think this tossup (with the possible revision I suggest above) is pretty much fine, and I don't see that adding a few hundred characters worth of clues would help it that much. I think that no matter how long a tossup of this type gets, there is still going to be a "difficulty cliff" between "very hard clues which only bona fide experts on the subject know" and "first well-known title, or similar reflexive buzz point." I guess Jerry is arguing that, by having so many more words, he can include a wealth of "transitional clues" which allow players who are neither "experts on the subject" nor "people who can only buzz on titles" to get their innings. No doubt a good writer can do that, but in the wrong hands longer tossups just meander or experience exactly the same kinds of abrupt difficulty shifts (e.g., Jerry's Hegel tossup has the same jump from "hard clues" to "The Science of Logic" that mine does, albeit the jump comes earlier in the tossup). I would argue that a length-restricted tossup can also feature enough transitional clues, though in the wrong hands it will lack sufficient transitional clues or be flawed in other ways. For instance, take this other philosophy tossup from this year's ICT (already mentioned in this thread, I think):

This man collaborated with Jean-Loup Thebaud on a work comprised of seven days of dialogues about the idea of justice. In addition to ~Just Gaming~, this one-time member of the ~Socialism or Barbarism~ school used Lacanian ideas to attack structuralism in ~Discours, figure~. He introduced the idea of the (*) "{differend}," but is best known for a "report on knowledge" published in 1979. For 10 points--name this French thinker who was incredulous toward {meta-narratives} and wrote ~The Postmodern Condition~.

answer: Jean-Fran\,cois _Lyotard_

I'd say that this does a better job than my Hegel tossup of modulating smoothly from "description of very obscure work" to "best-known title," with a nice progression of transitional clues (placing him in 20th-century France, describing his intellectual position, giving catchphrases from his major work).

The empirical question, I guess, is whether this Lyotard tossup (if you accept that it's solid) is actually something of an outlier. Another way of putting the empirical question would be to ask: Do 1,000-character tossups in general tend to be better structured than 500-character ones, and if so, is that because of an inherent limitation in the latter? I don't know how one would go about generating the empirical data required to back up any arguments about this. (For instance, I haven't seen any study of the tossups in, say, the most recent ACF regionals with percentages on how many of the tossups had significant "difficulty cliffs," or how many used their length to meander around without winnowing pyramidally to the answer, etc.)

My own sense, from writing a ton of both NAQT's length-capped tossups and non-NAQT length-unlimited tossups (see, e.g., 2005's ACF nationals), is that it is possible to write "good enough" pyramidal tossups under the character limit. As I've argued elsewhere, the character limit helps create a certain kind of game (e.g. one which demands a less leisurely deductive process from the player) which is different from the ACF game. I don't think that it's inevitably inferior, though.

[No Rules Westbrook]

Well, Andrew, I think there's a pretty big advantage to having a higher character limit - as you suggest, it gives a lot of fairly good players a chance to get the question before it drops to a very low level. Here's the thing - it's very hard to give a lead-in clue and confidently say "if you're an expert, you should know this!" - and it's very hard to give a middle clue where you can confidently say "if you have a fair amount of knowledge, you'll know this!" All you can do is make educated guesses.

Let's look at it this way - On any given tossup about Thing X, you might have:
1. A group of players who know a whole lot about it (maybe it's in their general area of interest/study, but still they probably don't know every "expert clue" about it - unless they're a true scholastic expert, e.g. they've studied Thing X in depth as a scholar, which is quite rare)
2. A group of players who know a fair amount about it (but this is QB, so a lot of these people don't necessarily learn clues exactly in order from easy to hard - so this group of people includes people who may know a random but substantial collection of hard, medium, and easy clues about Thing X - that is, you can't truly rely on them knowing one clue just because it's easier than some other clue they know, but they do have a decent body of knowledge)
3. A group of players who know fairly little about it (in QB, this group is usually composed of people who haven't done much clue study for Thing X, maybe they're generally familiar with its definition or what it is, but usually they're very hard-pressed to buzz beyond an "easy" clue)
4. A group of players who know nothing or almost nothing (perhaps they've heard of its name and can get it off some kind of base giveaway)

Now, increasing the character limit (in the hands of a good writer, who doesn't just blabber on, but gives real clues) gives you a lot more wiggle room to allow group 1 to get the answer before group 2, and more space for group 2 get the answer before group 3. You can throw in more clues that the great player "might know" and then more clues that the good player "might know" before dropping to a lower level. I say "might know" because that's the reality of the situation - I always laugh when people declare "well, there was a good hard clue in there! - if you know so much about this Thing, why didn't you buzz there, huh mr. smarty pants?!" Because that's just not the way QB works - unless you're that rare scholastic expert, most solid players know a random scattering of collection of clues for most topics, and all you can do is make an educated guess. Lengthening questions a little bit gives you a little more room to layer in clues, so that pyramidality is enhanced.

There are plenty of other advantages to a larger character limit too, I think. For one really important thing, it lets you provide full interesting descriptions of what you're talking about - you don't have to condense and squeeze and chop up grammar and flow just for the sake of a character limit. Don't get me wrong, I'm no fan of some writer just blabbering on about something because he likes to hear himself talk...but a lot of times, in order to provide an interesting non-trite developed clue, you need space that you don't have under a character limit.

[Birdofredum Sawin]

For a change, I actually more or less agree with Ryan's discussion here. That is, I do think that it's clearly the case that having longer tossups can allow for better distinctions between players with varying degrees of knowledge. As an abstract principle, at least, it is certainly true that "longer tossups will, in the hands of good writers, allow for finer-grained distinctions between players." But the question is, How "fine-grained" do we, as a community, want or need those distinctions to be?

To approach this, I'll note that I think that Ryan incorrectly implies that there are two forms of quizbowl, one with a character limit on tossups (NAQT) and one without such a limit (ACF). Instead, I would say that NAQT has a hard character limit which is stipulated by internal rules (i.e. "No tossup can exceed 500 characters, or the NAQT system won't accept it"), whereas the circuit has a "soft" character limit which is stipulated by social norms (i.e. "No tossup can exceed X characters [where "X" could be extrapolated by a survey of the longest tossups at tournaments like ACF nats and CO], or any competent editor will trim it drastically").

So which limit is "better"? Well, one can imagine a form of quizbowl in which the average tossup is around 2000 characters long. I submit that such a form of quizbowl would do an even better job than the hardest ACF tournaments we have ever seen of differentiating between players with varying degrees of knowledge. I also submit that such a form of quizbowl would be virtually unplayable, for perhaps obvious reasons (moderator exhaustion, inability to pay attention to 12-sentence tossups, etc.).

We're also all familiar with forms of quizbowl in which the average tossup is much shorter. And we all agree that a form of quizbowl in which the average tossup is around 200 characters long is only really suitable for novices, because it's basically impossible to write tossups of that length which will satisfactorily distinguish between knowledgeable players.

The question of "where to set the limit" is going to depend on a number of factors, including how fine-grained we want our differentiations of player knowledge to be, how much endurance we think our target audience is going to have, etc. My view is that, for the time being, both the NAQT and ACF "limits" are within the bounds of "good quizbowl." (So, if ICT dropped down to like 350 character tossups, it would no longer be making sufficient differentiations between players; but if ACF's cultural practices allowed tossups to start creeping up to the 1300 character range, all but the absolute hardest-core players would likely lose patience for the questions.)

I'll also say that my experience of reading for many of the top teams in the country on last weekend's ICT questions doesn't do much to convince me that NAQT's current 500 character limit is unacceptable. That is, I didn't see many buzzer races on early clues, but rather a dispersed pattern of buzzes which, to my eye, looked comparable to the kind of pattern one would see at a circuit tournament.

[cvdwightw]

In my experience, the hard cap length means that writers must be more careful when selecting clues. Thus, a well-written NAQT tossup can be as good as a well-written ACF tossup; however, if clues are selected poorly, it is much more apparent in NAQT than in ACF.

Let's say you're writing a tossup on Kobo Abe, and you decide you want to lead in with plot clues from either Inter Ice Age 4 or The Man Who Turned Into a Stick. In an NAQT tossup, you'd better choose one or the other, and in either case you'll end up rewarding only the players with knowledge of that Abe work. In an ACF model, you can have both. In the hands of a competent writer, either of these models work (and in either case, you're not rewarding people who only have knowledge of The Suitcase); we can similarly go down with second-tier Abe works, then down to his famous stuff. So you end up with an ACF question that mentions 6 to 8 works and an NAQT question on 3 or 4; as long as there isn't a ridiculous difficulty cliff (Description of The Man Who Turned Into a Stick, description of Inter Ice Age 4, title drop those titles, title drop Woman in the Dunes), neither one of these is necessarily a bad question. As Ryan said, people have a smattering of knowledge about most things, and so you're inevitably going to prevent people who have knowledge of things not mentioned in the tossup from buzzing on those things.

A good ACF tossup can always be cut down to a good NAQT tossup, and a good NAQT tossup can always be expanded into a good ACF tossup. As long as people are careful with clue selection, there's nothing that says that an ACF tossup must necessarily be better.

[at your pleasure]

I think that for our purposes, it might be more useful to discuss the utility of longer(6-8 line) questions. Let's take your Abe example. I have no doubt that it is possible to write a reasonably pyrimdial ICT or ACF nats tossup on Abe. The question is "Is the relative utility of greater ability to determing Abe knowledge worth the tradeoff in gameplay time ?". I think that at a national championship where the field is strong(and more likely to be close in their levels of Abe knowledge) and happy to make the tradeoff due to greater dedication, it's definitely worth it to make your tossup more fine-grained*. However, at a hypothetical competition where the field is unlikely to be close in their Abe knowledge and less dedicated**, it may not be worth it to make the tradeoff.
*I suspect that this is in fact true of the vast majority of collegiate circuit quizbowl.

[Mechanical Beasts]

In my experience, the hard cap length means that writers must be more careful when selecting clues. Thus, a well-written NAQT tossup can be as good as a well-written ACF tossup; however, if clues are selected poorly, it is much more apparent in NAQT than in ACF.

Let's say you're writing a tossup on Kobo Abe, and you decide you want to lead in with plot clues from either Inter Ice Age 4 or The Man Who Turned Into a Stick. In an NAQT tossup, you'd better choose one or the other, and in either case you'll end up rewarding only the players with knowledge of that Abe work. In an ACF model, you can have both. In the hands of a competent writer, either of these models work (and in either case, you're not rewarding people who only have knowledge of The Suitcase); we can similarly go down with second-tier Abe works, then down to his famous stuff. So you end up with an ACF question that mentions 6 to 8 works and an NAQT question on 3 or 4; as long as there isn't a ridiculous difficulty cliff (Description of The Man Who Turned Into a Stick, description of Inter Ice Age 4, title drop those titles, title drop Woman in the Dunes), neither one of these is necessarily a bad question. As Ryan said, people have a smattering of knowledge about most things, and so you're inevitably going to prevent people who have knowledge of things not mentioned in the tossup from buzzing on those things.

I don't think it's as important that you can mention more works, especially when several are of the same difficulty (though sure, it's better when you can have players buzz on The Red Cocoon, then The Box Man, then Inter Ice Age 4, then...) It's better, rather, to be able to include deeper clues about the most important works (and this is probably most important with philosophy, where necessarily short clues about works makes them disproportionately less buzzable, I think). I think more NAQT cluespace is taken up with titles--which probably is better than, like, just having a bunch of seven word summaries of works and then two titles, but which is not as good as the alternative.

I think Andrew's argument that both length ranges can constitute a legitimate form of quizbowl is certainly a valid one. (I certainly believe that, executed ideally, 500 character tossups create as good a result as 1000 character tossups within the margin of error that upper level play requires.) I do believe, though, that it doesn't accomplish as much in terms of learning (I don't think many people read old ICT sets to learn novel clues, because so many novel clues are in other sets). And that's sort of a shame, because I know I'll learn a lot by reviewing the science, for example, in this set, but I'd probably have learned more if the tossups had more content.

I have succeeded in saying much that's tautological, but I think it's still important.

[No Rules Westbrook]

Well, Andrew's no doubt right that we have to make some reasonable choices about where to cut off length and fine-grainedness of clues. But, I reject his and Dwight's relativistic premise, stated all too casually, that one approach isn't necessarily better than the other approach.

To me (and others may disagree if they're inclined), asking me to edit or write NAQT questions would be like asking me to edit or write questions with one hand tied behind my back. With the hard cap and timed round in place, I don't feel that I can create questions which fully meet what should be the standards of modern qb (that is, I don't feel that I could present the quality and quantity of interesting clues necessary to produce optimal questions, at least for discerning players). And, in my mind, that makes the NAQT structure inferior.

This can be coupled with what I (and others) believe to be the ridiculous overrepresentation of things like current events and geography and trash in NAQT and underrepresentation of things like fine arts and social science. In my opinion, the result is that when a serious academic player decides to write for or play NAQT, they are in effect choosing an old broken-down pickup truck over a nice new Cadillac. You can scrub up the truck and stick some new tires on it (see: hire better writers), but it ain't never gonna be a Cadillac. I don't care if you wanna get dolled up in your Sunday best and roll into town on the old junker once a year, just don't pee on my leg and tell me it's not necessarily any worse than rain.

[AKKOLADE]

they're not cadillac nutz

[cherrycolouredfunk]

FWIW, I consider myself something of a philosophy player, though my knowledge of Hegel is pretty thin. (Phenomenology of Mind is gathering dust on my shelf; hopefully this summer I'll read a bit in conjunction with Charles Taylor's Hegel.)

I powered the ICT TU off Kojeve, though I've never heard of the St. Louis or Cincinnati schools. For the analytic philosophy crowd, a better clue would have been the contemporary Pittsburgh school (http://ndpr.nd.edu/review.cfm?id=12925) centered around McDowell, Brandom, Haugeland, et al. I would have put this clue before Kojeve, actually; his influence on Sartre and, more popularly, Fukuyama, is pretty well known.

Quick googling reveals that these St. Louis and Cincinnati schools are from the nineteenth century. I suspect they're mainly of historical/antiquarian interest today. A side note: Personally, I'd prefer if fewer questions were written about such schools, and more on contemporary philosophy. An unscientific google search restricted to quizbowlpackets.com gives two results for "St. Louis Hegelians," but only two results for the rather more important Churchlands, zero for Dummett, and zero for Grice. This, however, is probably a subject for another thread, though I invite you to imagine an economics answer space dominated by nineteenth century Ricardians.

I guess Jerry's TU is a bit better structured, though I encounter "Hegel's philosophy falsified by the solar system!!!111" pretty frequently. I guess the conclusion is, uh, philosophy TUs are hard to write.

[No Rules Westbrook]

My post above notwithstanding, I want to say that I actually agree with Yaphe that the St. Louis Hegelians are kind of "famous" (in a quizbowl sense in my experience, if not any other). I would've liked to see that question elaborate a little bit on "There exists a St. Louis school" - give a few of the names associated with it (I know there are like three dudes assoc with it) - I'd like to see that both in the interest of pyramidality and in the interest of making the question seem fuller and possessed of more interesting facts. And, I'd probably like to see a few more developed clues after St. Louis, instead of just "name the author of work and work." But, that's all part of my argument against the character cap again, and why I think you can improve substantially on questions without it.

[Birdofredum Sawin]

I'd be happy to post excessively in a side-thread about the proper constitution of philosophy questions in quizbowl, but for now I'll just say that I strongly disagree with the possibly implied premise that "many (if not most?) philosophy questions should be on things of interest to contemporary (Anglo-American analytic?) philosophers." If you take this position, maybe you see it as analogous to the view that "science questions should be on things of interest to contemporary scientists, not antiquated theories which people don't take seriously any more." But that analogy doesn't really work, because while basically all contemporary scientists share the same outlook (i.e. none of them take phrenology or phlogiston seriously), you can't say the same of "all contemporary philosophers." Some of them are hardcore analytic types who don't give a damn about anyone before Frege or maybe Kant; some of them are fuzzy Continental types who don't give a damn about any American thinker after Peirce; and some of them are historically minded, and actually take the views of pre-1900 thinkers seriously.

My own goal when filling a philosophy distribution is to produce a mix of historical stuff, contemporary analytic, and contemporary Continental. So in ICT there were questions on, e.g., Epictetus and Plato on the historical side; the Chinese Room argument and David Lewis on the analytic side; Heidegger and Lyotard on the Continental side.

Anyway, the aim of this post was to nip in the bud any possible "philosophy questions should start falling in line with science questions" mindset. Again, if that seems to require further argument I'd be glad to take it up in a separate thread.

[Mechanical Beasts]

Yeah, I think that's reasonable. Having the other perspective, I also think that the fields are of entirely different natures. If people publish articles, today, analyzing [REDACTED] from [OLD TIME PERIOD] (as I now know they do), then they are relevant, today, to the modern state of philosophy (or of science). This is true for [REDACTED] but not for phlogisticated air.

[setht]

Well, Andrew's no doubt right that we have to make some reasonable choices about where to cut off length and fine-grainedness of clues. But, I reject his and Dwight's relativistic premise, stated all too casually, that one approach isn't necessarily better than the other approach.

To me (and others may disagree if they're inclined), asking me to edit or write NAQT questions would be like asking me to edit or write questions with one hand tied behind my back. With the hard cap and timed round in place, I don't feel that I can create questions which fully meet what should be the standards of modern qb (that is, I don't feel that I could present the quality and quantity of interesting clues necessary to produce optimal questions, at least for discerning players). And, in my mind, that makes the NAQT structure inferior.

This can be coupled with what I (and others) believe to be the ridiculous overrepresentation of things like current events and geography and trash in NAQT and underrepresentation of things like fine arts and social science. In my opinion, the result is that when a serious academic player decides to write for or play NAQT, they are in effect choosing an old broken-down pickup truck over a nice new Cadillac. You can scrub up the truck and stick some new tires on it (see: hire better writers), but it ain't never gonna be a Cadillac. I don't care if you wanna get dolled up in your Sunday best and roll into town on the old junker once a year, just don't pee on my leg and tell me it's not necessarily any worse than rain.

I'm inclined to disagree. When I started writing questions for SCT I found it slower going than writing for (m)ACF events, but I soon realized that I should spend a little more time per question working on wording to make things fit under the character limit without being unclear or sacrificing good clues, and a little less time per question collecting good clues--there just isn't space for three really hard clues in an SCT tossup, so once a writer's found one or at most two good, hard clues they should move on to researching middle-level clues. After some practice I think I was able to produce NAQT tossups faster than I typically produce (m)ACF tossups. In my experience, writing bonuses also went a bit quicker and I also had a bit of a transition period where I went from taking about the same amount of time per bonus to taking less time. Here, the main things for me were (again) to remember not to bother looking up as many clues per bonus part, and to skip writing lots of clues into each part. In particular, at some point I figured that at SCT tossups can't afford to have lots of "learning clues" that are interesting for top-level players but largely unbuzzable, and that bonuses really can't afford to have "learning clues" in every part--in fact, most bonuses should have at most one or two such clues overall.

It's possible that these restrictions (say, maximum 1-2 "learning clues" per tossup/bonus) immediately rule out the possibility of producing a set that meets "what should be the standards of modern qb." I don't think these restrictions preclude production of sets that accurately distinguish between players and teams in a sufficiently fine-grained manner. From a purely competition-based standpoint, then, I think the length cap is not at odds with the standards of modern qb. I guess the reduced amount of learning per question at a timed event with a hard length cap might turn off Ryan or other people, but I personally enjoy having a couple tournaments a year that emphasize competition at a good pace over learning per question--and I also enjoy having a bunch of tournaments that emphasize competition at a slower pace with more learning per question, and I enjoy having a couple vanity tournaments that emphasize learning per question over competition.

Moving on to Ryan's final paragraph, I wanted to point out that the ACF and NAQT distributions actually appear to have pretty much the same amounts of geography and social science. I'd be happy to see geography decrease and social science increase, but it doesn't seem correct to say that ACF is way ahead of NAQT in handling those categories. I'd also like to see NAQT decrease its trash and current events distributions a bit in favor of bumping up fine arts, RMP, and literature. In any case, while I prefer the ACF distribution and longer questions, I certainly don't feel that NAQT is as big a drop-off from ACF as pee vs. rain on my leg; I don't know if Ryan really does feel that way or if this is just another example of his country fried hyperbole.

-Seth

[Important Bird Area]

Moving on to Ryan's final paragraph, I wanted to point out that the ACF and NAQT distributions actually appear to have pretty much the same amounts of geography and social science. I'd be happy to see geography decrease and social science increase, but it doesn't seem correct to say that ACF is way ahead of NAQT in handling those categories.

I've been reading a lot of ICT feedback lately, and this seems at variance with the opinions expressed by a lot of other players.

The ACF distribution calls for 1/1 geography in each submitted packet of 24/24.

NAQT's distribution resulted in the following for ICT: 3 rounds with 1/1 geography and 15 rounds with 2/1 or 1/2 geography. There were no rounds with fewer than two geography questions, as would be expected to happen to a few ACF rounds (from repeats or just mediocre submissions).

Therefore the complaint that "these packets had too much geography" I've been seeing these week has a strong basis in fact (that is, current NAQT packets have roughly 50% more geography than their ACF counterparts).

I happen to think that more geography is a good thing, but it's fairly clear that that's a lonely opinion these days.

[Cheynem]

I like well written, interesting geography and not just ALMANAC FUN stuff. Hopefully Harvard International and the upcoming all-geo tournament will show us the exciting future of geography questions.

[setht]

Moving on to Ryan's final paragraph, I wanted to point out that the ACF and NAQT distributions actually appear to have pretty much the same amounts of geography and social science. I'd be happy to see geography decrease and social science increase, but it doesn't seem correct to say that ACF is way ahead of NAQT in handling those categories.

I've been reading a lot of ICT feedback lately, and this seems at variance with the opinions expressed by a lot of other players.

The ACF distribution calls for 1/1 geography in each submitted packet of 24/24.

NAQT's distribution resulted in the following for ICT: 3 rounds with 1/1 geography and 15 rounds with 2/1 or 1/2 geography. There were no rounds with fewer than two geography questions, as would be expected to happen to a few ACF rounds (from repeats or just mediocre submissions).

Therefore the complaint that "these packets had too much geography" I've been seeing these week has a strong basis in fact (that is, current NAQT packets have roughly 50% more geography than their ACF counterparts).

I happen to think that more geography is a good thing, but it's fairly clear that that's a lonely opinion these days.

Well, I'm not sure whether the ACF distribution described in the qbwiki page I linked to above is official, but if it is, it calls for 1/1 geography in the first 20/20 of every packet. It's true that the submission guidelines call for 1/1 in each submitted packet of 24/24, but I gather that Matt (and possibly other people--maybe even including myself, although I don't remember ever voting on this) doesn't want to have less than two geography questions in any ACF round. I think the appropriate method of comparison for the ACF and NAQT distributions is percentage of the first 20/20 in an ACF packet vs. percentage in a full NAQT set, since that's the best representation of what people actually play in the vast majority of games; under that metric, the geography and social science numbers are very similar for the two formats (I guess NAQT has about 6% more geography relative to ACF, and about 8% less social science; these differences seem negligible to me).

As I said before, I would like to see the geography distribution decrease a bit, but I feel that way about both the NAQT and ACF distributions. If lots of people feel that NAQT has too much geography I hope they'll let NAQT know, and I also hope they'll let ACF know that 5% geography in the first 20/20 is more than people want.

-Seth

[setht]

While I'm at it, I thought I'd add that I feel this year's set was the best ICT so far. I thought the 2008 set was pretty good; I think this one was better. I think the tossup answer selection ranged a bit deeper than it has in the past, as did the hard parts of bonuses, and I liked that.

Looking specifically at some astro questions: the extinction tossup probably should have axed the first sentence (which doesn't seem unique to me) and put in another sentence transitioning to the bio definition (perhaps some stratigraphy-related stuff on the Permian-Triassic or K-T?); aside from that, I like the question. The color excess and reddening clues are both great, the A-sub-lambda notation is as close to universal as you're going to get in science, and dust extinction is really important in astronomy--it's always nice to see some real astronomy questions. Given the paucity of astronomers playing the game, this question probably would have done better in misc. science with a sentence or two of astronomy material followed by an expanded treatment of extinction in biology, but I certainly enjoyed playing it as it was.

I think the quasar question starting off with Gunn-Peterson troughs fell victim to a recent upswing in the frequency of Gunn-Peterson clues/bonus parts in circuit events. I think I might have written one of the first such questions, and I think at the time people thought I was crazy for doing so; a couple mentions in various sets later, it's possibly too easy for a lead-in. If it weren't for the random appearance of G-P in recent sets, I'd think the clue was fine--this isn't something people see in intro courses or anything like that. It's also real astronomy material, and certainly quasars are of great interest in astronomy (and I think they have some cross-over appeal in the general public, something that dust extinction probably doesn't have going for it, sadly).

Finally, I'll call out the tossup on Rhea as not being a good idea. In general there wasn't much solar system geography in the set, and some of what was there was fine--e.g. the bonus on Kuiper belt objects and the bonus on transits of Venus, which are of current and historical interest. I felt the hard parts of those bonuses were possibly overly-hard, but the basic content seemed fine for solar system questions. The Rhea tossup, by contrast, didn't seem to have much going for it. I guess Rhea might be of some interest since there are signs of a ring system, but my impression is that even the ring system isn't enough to make Rhea all that interesting, and it probably shouldn't be the topic of a tossup. In general I think "features of the solar system" (e.g. the Kuiper belt) tend to make better solar system tossup subjects than "name this chunk of rock and ice that has maybe one thing distinguishing it from the next chunk of rock and ice."

-Seth