Tossup length

[magin]

After looking at some tournaments this year which had very long tossups (more than ten lines long in 10 point Times New Roman with 1 inch margins), I thought I'd post a thread encouraging people not to write them. I used to be in favor of very long tossups at high difficulty levels because I thought such length is necessary to truly differentiate among top teams, but after spearheading a tournament filled with such tossups (Gaddis 2), I've had a change of heart.

First of all, I find that such very long tossups tend to use that extra space for more leadin-type clues that few, if any players will answer, instead of middle clues that will be buzzed on by more than one player. Realistically, I don't think that's a good use of clue space at any tournament.

Secondly, very long tossups and very lengthy bonus parts (more than 10 lines, and more than 3 lines, respectively) make rounds take longer to read, especially for newer readers. This is probably less of an issue at tossup-only side tournaments, but for regular tournaments, it's going to result in rounds taking more time.

Finally, instead of using stuffing all of the neat clues you find about, say, Atahualpa into one mammoth eleven-line tossup, you can write a seven-line tossup on Atahualpa and save some of those unused clues to make your life easier if you decide to write Atahualpa questions in the future.

Questions should certainly contain leadins that reward very knowledgeable players. However, if you're using forty or fifty percent of your tossup space for clues that almost no one will answer, I think you're devoting too much of your tossup space to leadins. This phenomenon might be caused by writers assuming that people will buzz on all the clues they have found, overemphasizing what people actually know when playing. Personally, I've tended to assume that people will automatically know clues just because I'm staring at them on a computer screen or a textbook, and usually, that holds false more often than not. Also, in the past I've felt the need to make tossups longer to prevent early buzzes from being on the first or second clue; now, I recognize that that need is irrational, and people buzzing on leadins or early clues with knowledge is absolutely fine.

I recommend using the first one or two lines of a question for leadin material, and then move on to the middle clues, which should make up the rest of the tossup until the giveaway. In practice, I think six to seven line tossups mostly consisting of middle clues (in Times New Roman 10 point font) and bonuses where each part is around two lines or fewer is a good length for questions. Five or eight line questions are probably fine too, but greater than eight is pushing it, and fewer than five lines may not differentiate between teams well.

Do these guidelines seem reasonable? Is there an argument for very long questions that I haven't considered? I think, as a community, issues such as question length have been seen as common assumptions, but I'd like to talk about them openly to prevent possible miscommunication.

[Mechanical Beasts]

I think you've described very well some of my errant thought processes while bludgeoning the community with fourteen line delights at HI last year, Jonathan. I can say that several of the dozen tossups I've written for this year's edition are ten lines; none are more (and none are in their final, edited form).

One cogent argument against over-long tossups is purely the fact that frequently fourteen line tossups aren't that long because someone found twenty clues: they're that long because someone is writing without adequate concern for economy of language--and often that lack of care goes along with a disregard for other important facets of editing--the sheer exercise of ensuring your tossups don't use unnecessary words (and the tremendous majority of behemoth tossups do, though it's certainly not a property inherent to them) helps make them better in other ways.

[Skepticism and Animal Feed]

The most compelling reason for me is that marginal quizbowl teams and players seem to greatly prefer shorter tossups and are disproportionately spooked by longer tossups. So you are shrinking your audience if you write too long.

[MLafer]

Perhaps it would be helpful to ask why people write questions over 8 lines. I don't get it myself, and Magin's reasoning perfectly coincides with mine, along with the fact that it wastes editing/writing time in an era of sloppy & late tournaments.

So, to those of you who insist on writing long-ass tossups, WHY?

[Magister Ludi]

I usually agree with Jonathan, but on this issue I have to differ. I find that many question sets that adhere to strict line limits suffer from a lack of context in clues. When one is trying to write "concisely" I often find that this philosophy translates to questions that are overly generalized. Often I find myself not being able to buzz on things that I'm very familiar with because a description lacks context. I find myself saying, "Thats kind of what happens" after listening to a tossup that tries to summarize an incident from a work in an oversimplified fashion. The push to have set line limits encourages writers to summarize complicated plot clues as curtly as possible rather than as accurately as possible. A lot of people praised FICTE 2008 as an example of a great short questions, but I thought that set was mediocre because it suffered from short/over generalized descriptions and a lack of middle clues. I can provide examples from tournaments with line limits (such as Penn Bowl 2009 or FICTE) to show how I believe these line limits lead to vague clues.

While Jonathan stated in his post that he hopes questions will consist mainly of middle clues, I fear that a move towards shorter questions will lead many people to fill their questions with too many leadins. Once again look at FICTE 2008 (which suffered from many, many difficulty cliffs) to illustrate how shorter questions often lead to fewer middle clues. I will be interested to read this year's T-Party to see if difficulty cliffs from a lack of middle clues is a recurrent problem.

Also, I think that if you are writing a tossup on something very difficult (like many of the subjects in Gaddis) then one doesn't need twelve line tossups. This idea was made clear to me after KLEE. However, if we want to continue the trend of writing on easier "core academic" subject matter I think it is necessary to have longer questions to appropriately distinguish between the top players at high level tournaments. I hope at this year's ACF Nationals there will be more questions on very well known books that many literature students read and study (the Jane Eyres and Madame Bovarys of the world) and fewer questions on books that people don't really study very much. Accordingly to make these more accessible questions viable for top players, I believe we need long questions that differentiate between many different layers of knowledge. For example I imagine that there are actually many different levels of knowledge for a work such as Mrs. Dalloway. I guarantee you that there are some people who have read criticism and have very intimate knowledge of the quotes/ minor incidents, then there are people who read the book once and know the major plot points, then there are those people who have written a question on it or have read a summary, then there are those you just know a few random clues about the work. It seems to me that to differentiate between these fine gradations of knowledge about well known subjects, we really do need longer questions.

As a writer I don't like to have these question length concerns in the back of my mind. I simply want to write a question with the freedom to make it as long as I think it needs to be.

[Matt Weiner]

I think Ted's post illustrates the common fallacy of conducting all normative discussions of tossup construction through the lens of high-level literature questions, when there are many other categories and difficulty levels to be concerned about.

If you're writing a tossup on Jane Eyre for ACF Nationals, then yes, surely you need it to be a long question, because you're trying to discriminate among people who are very good players, using a question about a very famous work. But, since it's Nationals, you can write an 8 line tossup and that's fine. Writing a tossup on Reza Shah for Regionals or on Odin for Fall does not require 8 lines, because of the nature of those events. Writing a tossup on Soul Mountain for the ICT doesn't either, because there are far fewer levels of knowledge to worry about as answers become more difficult.

[Mike Bentley]

One thing to note is that there's a pretty big difference in tossup length between FICHTE 2008 and even a 6-line 10 point TNR question. FICHTE 2008 was using the old ICT cap of 425 characters per question (which was bumped up to 500 characters last year), while a 6-line 10 point TNR question can have up to 700 characters in it. I've generally felt that Penn Bowl has done a pretty good (but not perfect) job with its answer line cap of 6 lines.

Also, I'd argue that there's definitely a point of diminishing returns in regards to question length. Increasing a question from 4 lines to 6 lines will almost certainly create more distinctions in where people buzz, whereas increasing a question from 8 to 10 lines only rarely does this.

[ValenciaQBowl]

Going back to Lafer's question about why some people do this, I'll hazard that one possible answer might be pressure felt from feedback to be received later on this board. I think a lot of writers might rather take a little heat for writing toss-ups that are too long than for displaying difficulty cliffs or too-easy lead-ins. It's the same principle, I think, that goads some folks into choosing overly difficult answers on which to write, too.

I know I felt some of this pressure when writing Delta Burke this year, as I knew it would be mirrored at other four-year college sites. I've always probably had some lead-ins that were tougher than necessary for the field, as learning some hard clues was one reward I offered myself for writing much of the set, but in the last two years my questions have gone from probably averaging 5-6 lines of 11-point TNR to 7-8. Considering what I've seen thus far from stats from DB itself and the mirrors, I doubt I'm differentiating much, as the "middle clues" probably aren't often coming till the fourth line in. And I agree with Jonathan that opening clues which one can be confident will not be answered by anyone are just uselessly lengthening the experience, particularly for newbies (as Bruce also mentioned).

As a regular participant at Chicago Open, I have to wonder how many questions are answered by anyone before the first period appears. Again, I think this comes from social pressure to make sure one's questions are hardcore and not easily fraudable, etc. Though some of this pressure is surely positive, it may also be driving folks to err on the side of too hard and too long.

PS--no jokes about that last line, please.

[No Rules Westbrook]

For once, I'm really on the side of people like Magin here. I don't understand this balooning tossup length phenomenon, and I'd really appreciate it if people took the care to create sharply-worded 6-7 line tossups for most events. I can understand slipping into more lines on a few tossups like philosophy, for instance, because those tossups often contain clues that take up a lot of space - even if you're doing a good job of being very concise (which you should - you should have clear pronouns, but you don't need to say "The protagonist of this work" six times, for instance). Conversely, subjects like mythology or geography are often ones where you can get out a high volume of clues in a hurry.

Even a very experienced and hardcore moderator like myself sighs at the prospect of reading through some of these 11-line behemoths people are cranking out. It really inflates the length of matches unless every moderator you have is very good at breezing through thickets of language quickly.

I think people should focus on "give a few hard clues, a few more medium clues, then a serviceable giveaway - and get out." Often, it seems like writers who aren't very confident of the difficulty of the clues they're giving just tend to go "uh, I'll throw all of these clues in there, and at some point people will buzz!" Regularly, these kinds of bloated tossups have clues out of order - a fairly famous clue followed by 5 lines of really unfamous clues, etc. And, you don't need to give every conceivable giveaway - I've seen plenty of tossups which will drop something super well-known, and then just continue for 2-3 more lines of giveaway. I don't see the point in doing that. If you give a clue that will almost surely result in everyone who actually knows the answer buzzing in and giving it, it's time to stop the tossup. The same thing with bonuses - it's fine to give a few harder clues up front to make bonuses interesting, but end the prompt with the easiest clue - don't drop the easy clue and then spend 2 lines babbling about less-famous works.

Lately, I've kind of been slipping into longer tossup (and bonus) length as well. It's something I'm deliberately going to try to stem, and particularly try to cut back on the volume of clues in the front half of the tossup that are lead-in material. I've begun to plan Experiment III, and it will focus on shorter tossup length (with bonuses, hopefully). So, this is an issue that I think applies to all difficulty levels.