Tossup length

[Birdofredum Sawin]

The ACF Fall discussion is getting a little wild, so I thought I\'d start this as a separate thread.

Some of the comments about tossup length seem to indicate confusion about how the game is played. That is, they seem to assume that every buzz is of the following nature: a player recognizes a clue he knows cold (\"oh, \'Adam Bede\' is by George Eliot\") and buzzes accordingly. If that were the case, then it might well seem silly to have so many clues prior to the giveaway.

However, such reflex buzzes (the ones that reward list memorization) are only part of the game. Another part is figuring out the answer, or that the answer can only be one of a few things, even though you don\'t know for sure what it is. Consider the following tossup (yeah, it\'s long -- it\'s from last year\'s ACF nationals):

The conclusion to this work mocks Bayle and Polybius for imagining that there could be nations that wouldn’t need religion, because the maxims of philosophers are useless to make people be virtuous. It begins with an “explanation of the picture placed as frontispiece,” in which we learn that the “lady with the winged temples” on top of a globe represents metaphysics, while the “luminous triangle with the seeing eye” is of course God. Its conclusion describes an “eternal natural commonwealth” which was “ordained by divine providence,” and follows the fifth book, which considers the “latest barbaric history” and the connection between “feudal law” and “ancient Roman law.” The fourth book claims that ancient jurisprudence was “a severe kind of poetry,” while poetic wisdom is the subject of Books 2 and 3, the latter of which deals with the “discovery of the true Homer.” First published in 1725, it proposes a cyclical theory of history. FTP, name this major work of Giambattista Vico.
Answer: The New Science or La scienza nuova

Sure, if you\'ve read The New Science recently you\'re going to get this fairly early. And if you\'ve never so much as picked up a copy of the book, you may well not get it until one of the last two clues (\"cyclical theory of history\" or even \"Vico\") if at all. So why bother with all the intermediate clues?

Well, even if you don\'t happen to know any of those clues cold, they allow you to engage in a process of winnowing based on your knowledge of the subject in general. The first sentence tells you it\'s probably a philosophical work of some sort; if you know who Bayle was, you can guess that the work was written in the late 17th century at the earliest; you can also guess that it\'s a European work of philosophy, because books in the analytical tradition don\'t usually made nods to people like Bayle or Polybius. The second sentence discusses its frontispiece, which you might know about without having read the book (whole essays, e.g. one by Angus Fletcher, have been written on it), but even if you\'re still not buzzing you can feel more confident of the \"this is a weird book\" theory (something staid like \"A Theory of Justice\" has been ruled right out by now). The third and fourth sentences narrow down the subject matter of the book, touching on more and more famous parts of it and allowing the player more time to discard possible answers and narrow in on the right one. Then we get to the solid giveaways, which allow for reflex buzzes (or for the player who has been biding his time to go in with complete confidence).

My point is: A good player listening to this tossup (or any other) is not sitting there, his mind a blank slate, waiting to hear the first trigger word that will cause him to make a knee-jerk buzz. Instead, he\'s mulling over what he\'s been told, formulating possible answers, ruling some of them out, and waiting until he hears something that will tip one of those possible answers from \"possible\" to \"sufficiently probable for me to take a chance on it.\" If he\'s lucky, he\'ll hear some sort of confirmatory word (like \"cyclical\") as he takes the plunge. If he\'s unlucky, he\'ll hear something that suggests his thinking was terribly off-course. Either way, an active process of deduction has been going on the whole time.

That\'s one of the main reasons for the tossups in ACF to be as long as they are: to allow time for this to happen. At NAQT, there\'s basically no time for this kind of thinking to proceed: the game moves much faster, the tossups are much shorter, and people are either buzzing on clues they know or on instinct. ACF is simply a slower, more ruminative game. Some people only like the speedier, minimal-thought-required style; some people only like the slower, more deliberate style; some people enjoy both. (I myself like them both, while recognizing that they are different things.) But it\'s important to note that there is a difference, and that asking for NAQT-length ACF questions is to ask that ACF stop being ACF and be something else.

Andrew

[recfreq]

Theoretically speaking, the more clues the TU has, the finer the difference b/t two players playing on the question you'll be able to make. So in a sense, long (in terms of clues) is always better, but it's just that logistics like time and uniformity dictates that we use shorter questions. I think may be one way is to have denser but shorter questions--that is, lots of clues but as few words to be read as possible. I've tried this in the past, and ended up neglecting that thing we call grammar.

[recfreq]

I neglected to mention though, that this idea of shorter denser question does take away the element of reflection that Andrew was pointing out, which I also find to be a type of thinking on a higher order from the NAQT reflex buzzer race, so may be we _do_ want to write complete sentences after all.

[vandyhawk]

Mike, what are your thoughts on regionals question length? Figured I might start writing some questions here and there in the near future, so some guidelines would be appreciated, and I imagine will help you guys too if everyone submits stuff of similar length. Thanks.

[Captain Sinico]

I'm going to ask that you aim for reasonable length in the frame of previous ACF Regionals: my guess would be that we'll conclude your median length should be 6.5 or so, and your max should be 8 (similar to Fall.) The hard, fast stuff will be released shortly, with all the other Regionals info. Obviously, if people write good questions that are longer, we won't discard them; we will, however, do our best to make their length more appropriate, so you can save us the work of shortening and yourself the work of writing more than will be heard by keeping it below 8 at all times.
In fact, I'd like to announce a strong commitment to length control, given that a number of people are unhappy with the length and given that I sincerely believe that well-written questions can differentiate teams of the highest levels at a length of 6 or 7 lines. I want to be able to say, if someone says "What was with all the 9 and 10 line questions at Regionals," that there was not a single one of those.
However, length is secondary: the important things in this regard are that your questions be clue-dense, comprehensible, and concise (not necessarily short, that is, but also not wasting words.) In other words, alongside the other principles of good question writing (no hoses, uniquely identified answer as soon as possible, pyramidality, etc.) you should write so that everything you're saying conveys some meaningful (in the sense of moving precipitately towards the answer) piece of information. If you're writing on those principles, and if you're not systematically exceeding 8 lines, you can be confident you're writing good questions for Regionals.

MaS

[First Chairman]

There's the vantage point of writing with respect to the player. How long is a normal person's attention span when it comes to playing this game? It is pretty hard to keep one's attention beyond 20 seconds, and for readers, anything that looks like an entire page of justified text is mentally exhausting.

I don't subscribe to writing "See Dick run" length questions, but there is an optimal length for a sentence that is read to an audience. The more difficult you make it for the reader, it is inevitable that the difficulty will be transmitted to the player. The more complex your sentences are, the more difficult it may be for a reader or a player to figure out what you are going for.

If I could make a request, please don't have more than 20 words in a sentence. Please put as many clues as you want, but don't overwhelm me with how much you know. Leave some clues behind for the rest of us to write for the next tournament. :)

[matt979]

I agree (not just assent but strong, outspoken support) with most of what Andrew Yaphe wrote to begin this thread. A few related observations:

--
It's still quite possible (though a lot more challenging) to write NAQT-length tossups that reward the deductive thinking Yaphe describes. More to the point, even though the timeframe is compressed, I suspect that good players very frequently carry out this evaluation process on NAQT-length tossups and buzz with something less than cold clue-specific knowledge but more (and much better) spot analysis than the word "instinct" connotes.

Bad shorter tossups are painfully obviously bad because they're random, abrupt, and clue-sparse. But it's deceptively easy to skate by writing deceptively bad longer tossups, where the quality just doesn't justify the verbiage.

--
I'm not at all familiar with <i>The New Science</i> and would not have gotten this tossup even at the end. I don't claim to be anywhere near a good player on academic questions, though many people observing this debate are (I imagine) newer players deciding how much of QB is worth that time.

(Hmm, that may be a bad assumption given that if you made it to this forum you've already demonstrated enough interest that you're probably not going anywhere. But still, there exist enough people who ought to play ACF questions (yet might unwisely choose not to bother) that Yaphe's point is well worth making to a wider audience.)

In any case, Yaphe's same points could easily be made on a tossup about something or someone "easier," one that ends with "everyone knows that" clues but is also solvable from the lead-in. The best question in the ACF Fall packets (in my opinion) is a fantastic example, especially around the seque from his works of poetry to his presence at the Miracle on Ice game as an official translator.


--
Quoting from weblog comments: Brian Rostron noted, quite correctly, that "by providing interesting and substantive information [good ACF-style tossups] may actually encourage people to learn more about the subject, or in this case, read the work."

In the best case, quite so (though I'm not convinced that ACF length is a prerequisite for fitting interesting facts). Unfortunately, some long tossups aren't all that interesting, and many more that are moderately interesting have their appeal squashed by the mumbled, mispronounced monotone of a lot of invitational readers.

(This isn't a knock on readers, especially since they're volunteering their time (our time: I read ACF fall), but sight-reading text is difficult, especially sight-reading text written at a very high reading comprehension level.)

I think many longer questions are written in the same prose style as academic treatises. Given that questions must be sight-read, this is often a mistake. Something with a "voice" more like newspaper prose (NY Times if you prefer) will make it much more apparent that this interesting, awesome thing you're asking about really is that compelling. I realize this will sounds like a call for "dumbing down" questions; please understand that it's not. I also realize that as advice it's not immediately useful for lack of specific suggestions on how to tighten one's prose.

To improve both prose and content: Those of you in a position to get question feedback from Yaphe et al (mainly if you wrote ACF questions), I strongly recommend that you solicit this feedback regardless of how much writing experience you have. Yaphe's commentary on specific questions iwill greatly improve your writing.