Are 20 Questions Enough?

[Sulawesi Myzomela]

Excluding NAQT, most good quizbowl sets in both high school and college are written to be played in a 20/20 format. One thing I've always liked about NAQT is they have a little more distribution flexibility due to having more than 20/20 in every packet. So that brings me to the title question of this thread:

Are 20 questions enough?

I think quizbowl could benefit from having more sets with 24/24 (or even more) questions per game. One benefit would be the ability to squeeze subjects like RMP, Fine Arts, Social Science, Geography, and Current Events into each packet without having to choose between which academic questions get in and which don't. It can also allow questions sets that include Pop Culture to minimize its impact in each game. Adding more questions raises the chances that the more knowledgable teams will win, because those teams will have more chances to prove their knowledge each game. Not only would this help distinguish between the best teams, but also reduce upsets where a more knowledgable team has bad luck and loses to teams they are better than. This will also increase the number of questions written in each set, meaning schools can collaborate on a set (which they should already be doing) while still giving players more opportunities to improve through question writing.

This is something I've been thinking about quite a while, and is simply an idea. There are probably good reasons to not consider this, but I'm curious as to what the community at large thinks about the idea.

[Charles Martel]

I've thought this same thing before. It seems to me that every round should include at least 4 history, 4 science, 4 literature, 3 fine arts, 2 RM, 1P, 1 SS, .5 Math, .5 Geo, .5 current events, .5 trash. This adds up to 20.5, and those numbers are minimums in my opinion. The only options seem to be to reduce 1 or 2 categories each round, which would cause the result of games to fluctuate more, or to raise the number of questions in a match.

[tintinnabulation]

IHSA has switched from 30 questions a round to 24 questions. One of the leagues we play in has changed its format accordingly. (The other one we play in hasn't--we're still playing twenty.) I don't know if this enables us to distinguish better between the top team and the...um...bottom team or just drags it out longer. (Our team doesn't have any competition in that league. Not trying to brag or anything; it's the truth.)

When IHSA switched the number of questions per match, some coaches complained because they had won matches on the twenty-ninth or thirtieth question, but some matches will be decided on the last question no matter how many teams you have. And perhaps having the larger amount of points at an "earlier" point in the game makes you the better team. Good teams who go against each other are just better at a certain point in time. We've all seen teams lose and then win against each other in the same tournament. The team who won was simply better during that match. I don't know whether making the matches longer will help.

[Smuttynose Island]

These are just some of my thoughts on the matter so take them as you will:
It also has the potential to make sets harder. By increasing the number of questions that you have to write in "hard" areas, you start writing on even harder answerlines, which can negatively impact the game. If you believe that you can expand a packet's length in such a way that this doesn't happen, or is very minimized, then go ahead, and do so, but don't if you start having to stretch to fill historically difficult categories.
Also I think that there is a lot to be said about having a consistent national format as it allows teams to quickly familiarized themselves with how the game works.
Lastly, yes more questions can reduce the possibility of upsets, but where do you draw the line? 20, 24, 30, 100 questions? Eventually you are going to have to arbitrarily chose a cutoff point. Right now 20 TUs and 20 bonuses seems to work well and allows tournaments to finish at reasonable times. At some point the length of each round comes into play. NAQT manages to avoid long round lengths by having shorter questions, something that many people complain about and allow for, arguably, less gradation of knowledge, which can possibly override the purpose of adding more questions in this post's original thread, namely to reduce upsets.

[Charles Martel]

I don't think that increasing the number of questions purely to increase the length of the game is a good objective. The reason to increase the number of questions is that the high school distribution lends itself better to 21/21 or 22/22 than 20/20, and 24 is the next nice round number.

[Mewto55555]

1P, 1 SS

I have heard this part of the distribution is most accessible and quite easy to write as well!

[Smuttynose Island]

I don't think that increasing the number of questions purely to increase the length of the game is a good objective. The reason to increase the number of questions is that the high school distribution lends itself better to 21/21 or 22/22 than 20/20, and 24 is the next nice round number.

I think that you might have misinterpreted what I said, although if you did not then I apologize, although I do feel like what follows is probably still a worthwhile point to repeatedly expound. I am absolutely against making rounds longer just to make the game longer. That's what I cautioned against in my last paragraph. At some point the extra length added to each match becomes more detrimental to the game than the reduction in the probability of an upset occurring serves as a gain.

[Sulawesi Myzomela]

These are just some of my thoughts on the matter so take them as you will:
It also has the potential to make sets harder. By increasing the number of questions that you have to write in "hard" areas, you start writing on even harder answerlines, which can negatively impact the game. If you believe that you can expand a packet's length in such a way that this doesn't happen, or is very minimized, then go ahead, and do so, but don't if you start having to stretch to fill historically difficult categories.
Also I think that there is a lot to be said about having a consistent national format as it allows teams to quickly familiarized themselves with how the game works.
Lastly, yes more questions can reduce the possibility of upsets, but where do you draw the line? 20, 24, 30, 100 questions? Eventually you are going to have to arbitrarily chose a cutoff point. Right now 20 TUs and 20 bonuses seems to work well and allows tournaments to finish at reasonable times. At some point the length of each round comes into play. NAQT manages to avoid long round lengths by having shorter questions, something that many people complain about and allow for, arguably, less gradation of knowledge, which can possibly override the purpose of adding more questions in this post's original thread, namely to reduce upsets.

The reason I proposed 24 is because there is already a precedent for having that many questions in a match. 20 and 24 are both already arbitrarily chosen cutoff points, but nearly every single non-NAQT set uses the former. I don't think questions have to be as short as NAQT to make this work on time, with a reasonable linecap of sets that use 24/24 packets, this would add at most 15-30 minutes to a tournament.

The idea isn't simply to improve the chances for the more knowledgable team to win, but to allow for more variety in sets that want to include more topics. It seems like we can avoid a lot of debates like "this tournament had too few religion and too many music questions", "why didn't you take out the 6 trash questions in the whole set to include a little bit more economics", and "this tournament included geography questions but no current events." Because all of those topics have a place in quizbowl, and right now they're fighting for position in each set, with the only other (unwanted) alternative being to take away from the Big Three subjects.

By no means do I think quizbowl sets should have 30 or 100 questions per game, and I don't even think every set has to have more than 20. Some tournaments are born to have 6-7 line tossups with a strict mACF distribution and that's completely fine.

[Charles Martel]

I don't think that increasing the number of questions purely to increase the length of the game is a good objective. The reason to increase the number of questions is that the high school distribution lends itself better to 21/21 or 22/22 than 20/20, and 24 is the next nice round number.

I think that you might have misinterpreted what I said, although if you did not then I apologize, although I do feel like what follows is probably still a worthwhile point to repeatedly expound. I am absolutely against making rounds longer just to make the game longer. That's what I cautioned against in my last paragraph. At some point the extra length added to each match becomes more detrimental to the game than the reduction in the probability of an upset occurring serves as a gain.

I was actually responding to Brittany Trang when I said this. I also forgot to mention that there is no reason for IHSA to be 24/24 as opposed to some other number of questions, as it uses a distribution completely different from the rest of high school quizbowl.

EDIT: Clarity

[Skepticism and Animal Feed]

This might be a good discussion to have in a world where high school question writers don't have problems writing 20-question packets that are not too hard on time. We don't live in that world.

[ether a go-go]

I've thought this same thing before. It seems to me that every round should include at least 4 history, 4 science, 4 literature, 3 fine arts, 2 RM, 1P, 1 SS, .5 Math, .5 Geo, .5 current events, .5 trash. This adds up to 20.5, and those numbers are minimums in my opinion. The only options seem to be to reduce 1 or 2 categories each round, which would cause the result of games to fluctuate more, or to raise the number of questions in a match.

It's pretty easy to drop down to .5 P, .5 SS, and 0 trash and fold the math into the science, which adds up to 18.5, and it wouldn't be a crime to drop fine arts to 2.5. Essentially, I don't think you can make a convincing argument to increase round lengths because of the distribution.

In my mind, the important consideration is match length vs number of matches. If you had an eight team tournament, I wouldn't see anything wrong with playing a full round robin of 30/30, and there probably are other scenarios that would lend themselves to similar solutions. With larger tournaments, playing lots of matches becomes necessary, so it makes sense that question writers who want their questions used at several sites are going to keep producing lots of 20/20 rounds.

[jonpin]

The reason I proposed 24 is because there is already a precedent for having that many questions in a match. 20 and 24 are both already arbitrarily chosen cutoff points, but nearly every single non-NAQT set uses the former. I don't think questions have to be as short as NAQT to make this work on time, with a reasonable linecap of sets that use 24/24 packets, this would add at most 15-30 minutes to a tournament.

I dispute this. In my opinion, turning a standard twelve-round 20/20 tournament into a tournament using 24/24 will either (a) shorten your tournament to eleven rounds, maybe even ten, or (b) add almost an hour to the tournament.

[Sulawesi Myzomela]

The reason I proposed 24 is because there is already a precedent for having that many questions in a match. 20 and 24 are both already arbitrarily chosen cutoff points, but nearly every single non-NAQT set uses the former. I don't think questions have to be as short as NAQT to make this work on time, with a reasonable linecap of sets that use 24/24 packets, this would add at most 15-30 minutes to a tournament.

I dispute this. In my opinion, turning a standard twelve-round 20/20 tournament into a tournament using 24/24 will either (a) shorten your tournament to eleven rounds, maybe even ten, or (b) add almost an hour to the tournament.

I was thinking along the lines of a 10 round tournament, so you're right that my estimate is a low. The questions will have to a bit shorter than the average housewritten set to offset some of the time added to a tournament, although I think producing questions that are 4-5 lines long instead of 5-7 lines long would not really reduce quality of a set.

In my mind, the important consideration is match length vs number of matches. If you had an eight team tournament, I wouldn't see anything wrong with playing a full round robin of 30/30, and there probably are other scenarios that would lend themselves to similar solutions. With larger tournaments, playing lots of matches becomes necessary, so it makes sense that question writers who want their questions used at several sites are going to keep producing lots of 20/20 rounds.

This is a good point. The main reason I'm bringing this up is because I think it would be nice to see more variety in housewritten sets when it comes to things like this, but I can see why question writers want to make their tournaments more uniform to appeal to people seeking to mirror their event. I think that if KAAC ever gets around to having straight TU/B format in Kentucky then 30/30 should be the way to go for them because they can't get rid of mathcomp, usually have a lot of 3-5 line tossups in their sets, and district and regional tournaments usually have 4 and 8 teams with state competition involving 32 teams over 2 days.

[Smuttynose Island]

The reason I proposed 24 is because there is already a precedent for having that many questions in a match. 20 and 24 are both already arbitrarily chosen cutoff points, but nearly every single non-NAQT set uses the former. I don't think questions have to be as short as NAQT to make this work on time, with a reasonable linecap of sets that use 24/24 packets, this would add at most 15-30 minutes to a tournament.

Yeah, this is extremely optimistic. Let's say that you do move to 24/24 rounds. That means that it is reasonable to expect that every 5 rounds you've essentially added an extra round to your tournament. That means that you've added atleast 30 minutes and most likely 40-50 minutes to your tournament. At this point it is reasonable to expect a tournament to run 10 rounds, so that means that the additional questions add two whole 20/20 rounds to your tournament, which adds an extra hour to hour and 40 minutes to your tournament. Adding this amount of time to a tournament will probably have a detrimental effect on team participation as we already have problems getting some teams to stay for entire tournaments and newer teams may just get exhausted or bored by the length of a tournament, especially when they could be playing more rounds against more equal opponents in the same time.

[JordanKuhn]

YAGHH I typed up a whole 4 paragraphs and then my computer froze so I lost all I had. But I'll restart, with a shorter version;

Yeah, this is extremely optimistic. Let's say that you do move to 24/24 rounds. That means that it is reasonable to expect that every 5 rounds you've essentially added an extra round to your tournament. That means that you've added atleast 30 minutes and most likely 40-50 minutes to your tournament. At this point it is reasonable to expect a tournament to run 10 rounds, so that means that the additional questions add two whole 20/20 rounds to your tournament, which adds an extra hour to hour and 40 minutes to your tournament. Adding this amount of time to a tournament will probably have a detrimental effect on team participation as we already have problems getting some teams to stay for entire tournaments and newer teams may just get exhausted or bored by the length of a tournament, especially when they could be playing more rounds against more equal opponents in the same time.

Think about it this way, though; playing 2 extra rounds adds the time of, duh, 2 extra rounds. In between any rounds there is always an inevitable halt caused by the breaks. Simply adding 4 questions to a round would not add any extra score-calculating, any extra manouvering of people, no room switching, no more bracket figuring, etc. Adding 4 questions to a packet, and therefore adding 40 questinos through the tournament, is a lot easier and less time consuming than adding the same 40 questions in 2 whole new rounds. I propose that adding a simple 4 questions per round won't change the total length of the tournament by an hour and a half, but by substantially less, maybe less than 45 minutes.

I also agree with Leucippe and Clitophon, who wrote that it would depend on how many teams you had as to how many questions you put. A tournament with 30 teams would of course not add any questions, but a small, local tournament of less than 10 teams would be fine adding even up to 10 or so questions per round, not that that would be desirable or desired.

Anyway, those are my shortened opinions. Have at them if you wish, I'm just a simple boy :P

[Smuttynose Island]

YAGHH I typed up a whole 4 paragraphs and then my computer froze so I lost all I had. But I'll restart, with a shorter version;

Yeah, this is extremely optimistic. Let's say that you do move to 24/24 rounds. That means that it is reasonable to expect that every 5 rounds you've essentially added an extra round to your tournament. That means that you've added atleast 30 minutes and most likely 40-50 minutes to your tournament. At this point it is reasonable to expect a tournament to run 10 rounds, so that means that the additional questions add two whole 20/20 rounds to your tournament, which adds an extra hour to hour and 40 minutes to your tournament. Adding this amount of time to a tournament will probably have a detrimental effect on team participation as we already have problems getting some teams to stay for entire tournaments and newer teams may just get exhausted or bored by the length of a tournament, especially when they could be playing more rounds against more equal opponents in the same time.

Think about it this way, though; playing 2 extra rounds adds the time of, duh, 2 extra rounds. In between any rounds there is always an inevitable halt caused by the breaks. Simply adding 4 questions to a round would not add any extra score-calculating, any extra manouvering of people, no room switching, no more bracket figuring, etc. Adding 4 questions to a packet, and therefore adding 40 questinos through the tournament, is a lot easier and less time consuming than adding the same 40 questions in 2 whole new rounds. I propose that adding a simple 4 questions per round won't change the total length of the tournament by an hour and a half, but by substantially less, maybe less than 45 minutes.

No, 30-40minutes is how long it actually takes people to straight up read 20/20. That time does not include room switching or any of the other items that you mentioned.*

*The exception is timed NAQT rounds.

[Sulawesi Myzomela]

30-40minutes is how long it actually takes slow moderators to straight up read 20/20.

That's more accurate. I can read a straight up 20/20 round in 20-25 minutes max, and I'm not as fast of a moderator as a lot of people who reguarly read quizbowl tournaments. If you're tournament doesn't have any bad moderators (a very rare feat) then this would be less of an issue.

[Standard-winged Nightjar]

A bigger question is, are upsets that bad a thing? If everyone basically knew going into each round who was going to win, it would take more of the excitement out of the game. There's obviously a balance to be struck here; it shouldn't be random (like having potential 220 point swings on the basis of one tossup in the second quarter in Questions Unlimited format). But I don't think this is really a problem in any legit quizbowl format right now.

[Grams's Go-Go Boots]

Upsets happen often enough on 20/20 where they are a possible result without making it a meaningless "anyone can win regardless of talent level" thing. I'm not sure if "But I don't think this is really a problem in any legit quizbowl format right now." is trying to say that upsets are impossible on 20/20; if it is, I'd have to disagree.

[Scaled Flowerpiercer]

A bigger question is, are upsets that bad a thing? If everyone basically knew going into each round who was going to win, it would take more of the excitement out of the game. There's obviously a balance to be struck here; it shouldn't be random (like having potential 220 point swings on the basis of one tossup in the second quarter in Questions Unlimited format). But I don't think this is really a problem in any legit quizbowl format right now.

I basically agree with Dave on this matter; upsets are bad if they are incredibly common and there seems to be a near 0 correlation between a team being good at quizbowl and a team winning at quizbowl. However, an occasional upset, as a result of very good playing on the part of the team and a couple of tossups coming up that luckily went a team's way, are fine, and contribute not only to the excitement of the game, but the ability for quizbowl to be called a "game" (or sport, or whatever you like).

As someone who has played his fair share of QUnlimited, I know that upsets are for more prevalent there than anyone else - in a tournament yesterday, my team's victory - by a 90 point margin, mind you, was probably almost entirely determined by a second quarter tossup, a second quarter tossup which boiled down to "what is 6!/4!" Now, the NHBB format might be a bit more open to upsets than 20/20, due to the lack of bonuses meaning that tossups actually do end up almost always being woth the same number of points (except for 30-20-10 powers), but it similar seems fair and there aren't upsets that are not the results of exceptional play.

As for whether or not 20/20 should be extended, I guess this could reduce upsets, by making distribution more constant, but I really think that 20/20 is fine - it has few enough upsets that in general the best teams win if they play well, but they are not immune to upsets to the point where quizbowl ceases to be fun.

EDIT: and for the record, not counting 60-second round result changes, 2nd quarter QU tossups are only 120 point swings.

[Charles Martel]

Upsets are only acceptable if they don't happen for certain reasons. If you played a round of quizbowl that suddenly switched to 100% literature in the middle of a tournament, leading to a defeat, this would clearly be a problem. I think the problem with 20/20 is that we're forced to choose between (A) punishing certain categories, such as philosophy and math, and (B) varying the distribution from round to round, which is also undesirable.

A separate issue of whether to length 20/20 is about the compromise between length and result consistency. Anyone who's played on a half-packet before would oppose switching quizbowl to 10/10, yet no one would think about 40/40. The range of 20-25 is probably about right.

[Charbroil]

I think the problem with 20/20 is that we're forced to choose between (A) punishing certain categories, such as philosophy and math, and (B) varying the distribution from round to round, which is also undesirable.

What does this mean? Neither of these things happen currently*, unless you consider the current philosophy and math distributions to be "punishing" those subjects. As people have mentioned before, it's not really possible to write more questions in those subjects while maintaining decent conversion levels.

* Unless you consider NAQT, but I don't think NAQT distributes by the set rather than by the round to avoid punishing categories, whatever that means.

[Standard-winged Nightjar]

Upsets happen often enough on 20/20 where they are a possible result without making it a meaningless "anyone can win regardless of talent level" thing. I'm not sure if "But I don't think this is really a problem in any legit quizbowl format right now." is trying to say that upsets are impossible on 20/20; if it is, I'd have to disagree.

No, that's not what I meant. Upsets are of course possible on 20/20 and indeed do happen all the time. And if you went to 24/24, of course they could still happen too. I just don't see the level of upsets as a problem, and if anything, a bit of uncertainty probably adds to the fun of it for many teams.

[Grams's Go-Go Boots]

Upsets happen often enough on 20/20 where they are a possible result without making it a meaningless "anyone can win regardless of talent level" thing. I'm not sure if "But I don't think this is really a problem in any legit quizbowl format right now." is trying to say that upsets are impossible on 20/20; if it is, I'd have to disagree.

No, that's not what I meant. Upsets are of course possible on 20/20 and indeed do happen all the time. And if you went to 24/24, of course they could still happen too. I just don't see the level of upsets as a problem, and if anything, a bit of uncertainty probably adds to the fun of it for many teams.

Good! Because I was really confused by the version of your post I create via misreading it or whatever.

[Grams's Go-Go Boots]

You can find the thread split so that the math discussion is here.

THANK YOU FOR HOLDING