A Treatise on the Common Link

[theMoMA]

I'd argue that there is something wrong with Jerry's question on children. It was very possible, perhaps even likely, to buzz with knowledge of one or more of the paintings and get the question wrong. If this were magical fairyworld where everyone always understood exactly what a tossup was driving at, the question would be fine. But it's not, and we're consequently not concerned with some kind of bizarre standard of quality that depends solely on the paper-existence of a question. By the nature of the game, we must concern ourselves the very pragmatic and not always ideal concept of "playability." If a question doesn't play well, even if that's due to our imperfect human nature or our imperfect world in general, it's a bad question. We can't simply look at the "children" tossup and decide that hey, even though a bunch of people with knowledge negged this tossup, it's alright because they shouldn't have done that. We can't start by claiming that the question-on-paper is good and conform reality to fit our theory that the question is good. We must start with the reality of the question-as-played and use it to understand why the tossup was good or bad.

People seem to be confusing a few things about common link tossups, so let me explain why the outright badness of the "children" tossup does not mean the death of common links everywhere. The thing that seems to cause the most confusion is that there're some kinds of tossups that are driving at something specific and occasionally neg people when they don't really know the answer. Remember the discussion we had about the tossup on Voortrekkers? People heard South African sounding things and said Boers, were prompted, and didn't know enough to get the tossup. We can look at that and say, hey, that's not great, and we should try to figure out a way to write that question so that people don't do that. Maybe we'll write on the Great Trek itself, using the same clues, since a playability issue that we didn't foresee happened. Maybe we'll decide that the people who negged probably didn't know anything about the Voortrekkers and were just trying to linguistically fraud the tossup, and write dozens more Voortrekker tossups while sipping Grey Goose, smoking blunts, and doing triple salchows on jetskis into our combination jacuzzi-foam pits.

Regardless, there's a big difference between the "Voortrekker" tossup and the "children" tossup. If you're prompted on "Boer," and you actually know the clue applies to "Voortrekker," you're going to say that. If you're buzzing with some promptable answer on the "children" tossup, it's not readily apparent that the word that the question desires you to say is "children."

To further the argument, if you buzz on "Piet Retief" with "Boers," you either know or don't know that Piet Retief was the leader of the Great Trek, and if you do, that's the readily apparent subclass of the thing you've already said (Boers) that you will say. On the other hand, if you buzz on, say, the "Daughters of Edward Darley Boit" clue with "people," or "siblings," if we were prompting on that at that point, there is no readily apparent subclass of things. You know a lot of classes of people that the daughters of E. D. Boit fall into if you've seen that painting, and there's really no clear hierarchy of those classes. Is "daughters" a subclass of "children," which is a subclass of "people?" Where does "siblings" fit into that order? These are not the kind of thing that we could definitively tease out in an extended discussion, let alone considerations that players should be forced to make while playing quizbowl.

I will say that I think it's not the best possible question if people can make the "Boers for Voortrekkers" neg, but as people may have noticed, there's quite a bit of reductionism to this problem. What I mean is, at some point, you have to make an arbitrary decision that "people who say more general class X but cannot produce the specific subclass Y of X upon prompting simply don't deserve points" (i.e. if you buzz with "mammals" but can't come up with "dogs"). You can argue that "Boers for Voortrekkers" is cutting it too close, and that may be true in practice, but I don't think it's provably true in a theoretic sense. And arguing that you can't write common links because people might do the "neg after being prompted on mammals for dogs" is absurd, because we know that there are plenty of common link tossups that are playable, and anything that tells us otherwise contradicts with reality (i.e. that thing we're dealing with).

Essentially, what I'm proposing is that there are three basic kinds of common link: "dog," "Voortrekker," and "children." "Dog" common links are the kind where there's basically a one-to-one mapping of the clues to the answer. While Garm is a mammal, you're always going to respond "dog" when you hear "Garm is this type of creature." These questions are just as playable as any tossup. "Voortrekker" common links might cause some confusion about the specificity of the answer (i.e. do you want fullerenes or nanotubes?) but if a player answers "fullerenes" and knows that the clue applies specifically to "nanotubes," there's no confusion about where the prompt is leading, because the classes of things in question nest in a specific and known hierarchy. These questions might have playability issues if they are particularly transparent with regards to the larger class of things, causing a lot of people to buzz who don't know that the clues specifically apply to the smaller subclass. The "children" common links are the problematic ones, because there are a bunch of possibly right classes of things that don't nest within each other and don't lead to readily available subclasses upon prompts; instead the question assumes the player will understand that the classes overlap in such a way to produce one correct answer. This causes all sorts of playability issues for reasons that I won't restate.

Are there cases where a "Voortrekker" tossup might be bad? Are there muddled cases between "Voortrekker" and "children" type tossups? Of course. That's why we have to look at our questions with playability in mind. We should look at a tossup on "children" and say, look, when we eye the paper we can see that all these things are factually true of paintings that depict children, and in the ideal world everyone playing would understand exactly what we want, but it would be pretty easy for people to get confused here, so I'm either going to find a bunch of paintings with the actual word "children" in their title, or better yet, just convert this into a bonus or something. And we should look at every tossup (not just common link tossups) critically to make sure that it will play well, because a question that plays poorly is a bad question.

[setht]

Jerry's mentioning my myth tossup from last year's MO on "_talking heads_" in the EFT discussion thread reminded me of a point I wanted to make about common-link myth tossups. First off, I would say that an answer line of "_talking heads_" falls somewhere in the "Voortrekker" or "children" classes Andrew has described. It's not a great answer line, and the question would be improved by simplifying to an answer line of "talking _heads_". That's something I should have caught as I was writing the question, and I apologize (again) for not doing that. Having said that, I'd like to point out that a tossup on "talking _heads_ in myth" may do a better job of rewarding "real myth knowledge" (of the academic sort) than most myth tossups do (see this link, especially the first mention of the "vital head" motif, for some good academic sources). I think it's fine for us to write tossups on Mimir, or on Bran, or on Orpheus, focusing on primary or reputable secondary sources--in particular, writing that sort of "character in myth" question seems to me to be analogous to writing a literature tossup on a work without including lit crit clues (which seems acceptable to me)--but writing a tossup on a theme/motif/archetype in myth that is studied as such in the academy may well do a better job of rewarding real knowledge.

-Seth

[Cheynem]

Did a lot of people neg the "children" tossup?

[grapesmoker]

I'll just go ahead and reproduce this question here:

In a painting by Ingres, Henry IV is depicted receiving the ambassador of Spain while occupied with three of these people. Philip Otto Runge painted a painting named after the Hulsenbeck ones of these, three of whom are depicted in it. Rolling a hoop, riding a barrel, and hitting each other with sticks are some of the activities in which these people are engaged in a Breughel painting about these kinds of games. Two female exemplars of this type of person with the surname Barnard were depicted holding Chinese lamps by John Singer Sargent, who also painted four other female examples who were related to Edward Darley Boit. For ten points, identify these diminutive humans, one of whom was depicted playing with a mirror while on her mother's lap by Mary Cassatt.
Answer: children (prompt on “boys” or “girls”, accept kids)

I'm going to look at the fact that some people may or may not have negged that tossup and declare I'm fine with it. I also find it improbable that anyone buzzed and said "people," in a tossup in which the word "people" was repeatedly mentioned.

Look dude, you got given the proper title of a painting (The Hulsenbeck Children) and the proper title of another painting (Children's Games). In fact, at least two of the clues, including the second one, told you that these people shared a common last name. The notion that you could guess "siblings" for this given the fact that straight away I told you this information is odd; no pyramidal tossup on "siblings" (if such a thing could exist) would be written this way! And then the clue about "Carnation, Lily, Lily, Rose," told you that these were female examples of this type of person, which means that "siblings" doesn't even fit the clue semantically. Even if the only thing you recognized was the clue about the Boit daughters, it still wouldn't make sense to say "daughters," since I told you these were female people (are there male daughters?) so you're left with a pretty restrictive set of possible answers. In fact, I'm going to argue that if you're familiar with the paintings in the clues and were listening to the words of the question, you couldn't have really thought that something like "peasants" or "siblings" would have been an acceptable answer. It looks to me like you buzzed with a general idea about what was going on, and you got negged because either you didn't know the specific things in those paintings or you confused yourself somehow or whatever; none of this is really my fault or my problem.

I don't know where this idea comes from that you should be able to buzz on whatever you vaguely recognize and get points for it without having to listen to what the question is asking for. I almost yelled at my teammates in practice this week for doing just this very thing, and I'm not at all sympathetic to your arguments here. It's not my problem you didn't parse the entirely uncomplicated pronouns provided for your benefit in that question or that you weren't able to use contextual information from earlier clues to eliminate obviously ridiculous answers.

[Mike Bentley]

I don't know, Jerry. You're giving people a lot of credit of keeping track of a lot of information in the early clues, something that can be pretty hard to do when you're having questions ready quickly at you. Most tossups work by describing a clue you know something about, at which point you buzz in. You then have 5 seconds to put together in your mind what is being asked for. If this was a common link tossup on artist from America, for instance, there would be no ambiguity on buzzing on the Sargeant question with "America" after hearing a description of Carnation, Lily, Lily, Rose. However, in this tossup you need to go against your normal quizbowl instincts of matching a kernel of knowledge ("this is a John Singer Sargeant painting that I know the details of being described") and wait until you're divined the answer in the context of the other clues in the question (something I argue is difficult to keep track of based on descriptions of early clues).

I don't think that this is the worst common link tossup ever, but I agree with other people that it's asking a lot of the audience to correclty identify what is being asked about and perhaps did unfairly penalize people for the knowledge you were otherwise looking to reward (i.e. knowing what's going on Carnation, Lily, Lily Rose).

[grapesmoker]

I don't know, Jerry. You're giving people a lot of credit of keeping track of a lot of information in the early clues, something that can be pretty hard to do when you're having questions ready quickly at you. Most tossups work by describing a clue you know something about, at which point you buzz in. You then have 5 seconds to put together in your mind what is being asked for. If this was a common link tossup on artist from America, for instance, there would be no ambiguity on buzzing on the Sargeant question with "America" after hearing a description of Carnation, Lily, Lily, Rose. However, in this tossup you need to go against your normal quizbowl instincts of matching a kernel of knowledge ("this is a John Singer Sargeant painting that I know the details of being described") and wait until you're divined the answer in the context of the other clues in the question (something I argue is difficult to keep track of based on descriptions of early clues).

I'm asking people to keep track of basic pronouns and their referents, something virtually every other question demands. If you are playing a tossup on Diomedes and hear a clue that goes "his father did this and that" referring to Tydeus and you buzz and say Tydeus, you will be negged. The same thing will happen in literature questions, history questions, and so on. Yes, you are expected to keep track of words in the question in a way that allows you to give the correct answer and not guess on hearing the first thing that you recognized. Sorry if that's an onerous burden for some people.

I'd also like to note how this discussion continues the trend of Andrew Hart not understanding common link tossups at EFT; last year, we had a protracted discussion about my tossup on "law," which Andrew complained was alternately vague, too easy, or too confusing, despite the fact that it was anchored by references to actual titles with the word "law," in them. I'm not sure if the problem here is listening to the question properly or what, but the words in them that aren't explicit clues are there for a reason; you would do well to pay attention to them. In any case, I'm going to take this complaint somewhat less seriously than I normally would.

[Mike Bentley]

I'm asking people to keep track of basic pronouns and their referents, something virtually every other question demands. If you are playing a tossup on Diomedes and hear a clue that goes "his father did this and that" referring to Tydeus and you buzz and say Tydeus, you will be negged. The same thing will happen in literature questions, history questions, and so on. Yes, you are expected to keep track of words in the question in a way that allows you to give the correct answer and not guess on hearing the first thing that you recognized. Sorry if that's an onerous burden for some people.

My argument is that it's a lot harder to do this in common-link questions like the one on children than in non-common link tossups and in "straightfoward" common-link tossups (by this I mean a common-link tossup on Poland or Inca gods or whatever). You don't need to go back multiple clues to piece together information about what's being asked for in a question on Diomedes when it says "his father did this which influenced this figure", since it's right in the clue. You just need to know the clues about Tydeus to get points there.

[grapesmoker]

My argument is that it's a lot harder to do this in common-link questions like the one on children than in non-common link tossups and in "straightfoward" common-link tossups (by this I mean a common-link tossup on Poland or Inca gods or whatever). You don't need to go back multiple clues to piece together information about what's being asked for in a question on Diomedes when it says "his father did this which influenced this figure", since it's right in the clue. You just need to know the clues about Tydeus to get points there.

Ok, see the thread I just started for multiple instances of people failing to use any notion of context in non-common link tossups! I continue to be highly unsympathetic to your view.

[cvdwightw]

Oh yay, another common link discussion!

My view remains essentially unchanged: every tossup question is a common link tossup. Some questions just have much tighter between-clue links than others, or lend themselves more easily to a clear answer choice. No one would write a tossup on "adventures of Loki," but there's no reason why one wouldn't write a tossup on the adventures of Loki with the answer line "Loki."

I think what Andrew (and possibly to some extent Mike) is arguing is that several common link tossups, possibly including the one on children, include Clues that cannot stand alone. Quoting from the QBWiki, "The logic that a player should know not to buzz in with the incorrect answer based on previous clues is not sufficient; since the player did not buzz on those clues, it must be assumed that they are of no use to them." Inevitably, this is the major problem with bad common link tossups. For instance, in the infamous "assassinating a Roosevelt" question, no middle clue could uniquely point to the answer ("so...this is discussing an attempted assassination of TR...but the first sentence mentioned two people...what does the second one entail?").

Another issue with the "clues that cannot stand alone" thing is when people recognize a specific clue, then have to think back on previous clues to figure out what association with that clue is being sought (cf. Jeff's account of the Skorzeny tossup in this thread). I think this is what Mike's arguing - it's not that people don't understand the context of clues, it's that people who recognize a specific clue must use clues they don't know to figure out what the link is.

Now Jerry, what I think you're arguing is that the context clues in the "children" tossup should have been sufficient that someone who recognized a specific clue could have buzzed confidently based on knowledge + context (and, indeed, it appears that people did so). However, I hope you realize that you're also a very good question writer and understand how to use those context clues such that someone who is paying attention and then earns knowledge from a clue should be able to answer confidently. In the hands of a poorer editor or writer, I have complete faith that this kind of tossup would have been disastrous.

I also think that your argument earlier in this thread is contradicted by your answer line. You're instructing the moderator to prompt on "boys" (even after you mention "female exemplars;" are there female boys, Jerry?)! You apparently allow for a prompt on one answer that is "clearly wrong" according to your definition of "context clues." Furthermore, you're also instructing the moderator to prompt on "girls," and then arguing that it wouldn't make sense to say a roughly analogous answer ("daughters") after being told there are "female exemplars" (I don't know, are there male girls, Jerry?).

I think the moral of this discussion is: common link tossups are really damn difficult to write well, and not-well-written common link tossups are far more dangerous than other types of tossups.

[grapesmoker]

I should have been more specific in my prompt instruction: I wanted it promptable before the point where it became clear that both male and female exemplars were being discussed. In particular, you could have buzzed on the Hulsenbeck Children clue (all boys) and said "boys," and gotten the prompt. It was my way of trying to be generous to someone who got the gist of what was happening and needed a little nudge to get it. Of course, if I were being very strict about the whole thing, I would not allow for a prompt at all.

As to your other point, well, the fact is that questions need to be written competently. I'm not responsible for someone else's failure to write a good common link tossup any more than I am responsible for someone else's failure to write a good tossup on a chemical element. If you don't think you can do it well, don't do it.

[Skepticism and Animal Feed]

it's not that people don't understand the context of clues, it's that people who recognize a specific clue must use clues they don't know to figure out what the link is.

Except that this is how every tossup ever works. Again, if you hear "FTP, name this German", and you know none of the previous clues, you have to use the previous clues to determine whether you should be buzzing with Bismarck or with Johann Friedrich Overbeck.

If you don't know the early clues, then you are taking a risk each and every time you buzz on any subsequent clue that can apply to two or more things. This is inherent to quizbowl.

[Cheynem]

I think there is a difference between the obvious "broad filler space" phrases like "Name this German" or "He fought in World War II" and actual, specific clues. That's why tossups don't end with "FTP, name this German." They end with "FTP, name this German leader known as the Iron Chancellor who famously unified Germany."

A common link tossup is inherently different because generally speaking, it requires multiple "meaty" clues to even narrow down what the answer is. This isn't true of all common links, which is why I think some of the more successful ones specifically refer to a title or state very clearly that they want an _occupation_ or a _title_ or whatever.

Comically Stupid Tossup time: Let's just say you wrote a common link tossup on "birds." Your first clue is something like "In the film Sabotage, Mrs. Verloc reacts to her brother's death while watching a film about the apparent death of one of these entities." This clue is describing the Disney short "Who Killed Cock Robin?" about the supposed death of the bird Cock Robin. At this point, if I know what the clue is referring to, I could theoretically buzz with _bird_, _robin_, _Disney cartoon characters_, _cartoon stand-ins for Bing Crosby_, or various other things. I require at very least the second clue to narrow down my options (although I would assume a moderator would accept both _birds_ or _robins_ for the first clue, not so much the other two).

That was a stupid example, but I think the dilemma holds for most common links. I mean, this is sort of a given of the very nature of common links, isn't it? It's asking you to determine the common link between more than one thing, so I would at the very least need two clues.

While any sort of tossup might at one point have clues that apply to two or more things, the general idea is that it is possible to buzz on the first clue with deep knowledge, knowing it only applies to one historical figure, one author, one book. I would argue that it is not quite possible to do that with a common link tossup unless it narrowly defines itself quickly. This is why I am in favor of the tossup doing so (which actually I think the children one did quite well) or being a little more generous on the answer line if the common link is not so easy to determine from one or two clues.

EDIT: Marnold has a good point that the few terrible common link tossups that seep through (including my Skorzeny one!) warp people's perceptions of common links, making them expect potentially insane stuff like _German leaders who killed themselves_ or stuff like that.

[marnold]

No one would write a tossup on "adventures of Loki,"

Well, as CO's _operations_ of Otto _Skorzeny_ proved, you shouldn't be so confident!

Anyway, I think these discussions have now passed Computational Math for "Threads Where The Most Obvious Conclusions Are Lost Amidst Absurdly Protracted Discussions." There are some crappy common link questions and there are some that are pretty good; the issue is usually decided over whether they point clearly to what kind of thing they want and whether that's a good answer choice in the first place. Some people really don't like common link questions and some people love them; the tournaments produced by the first group won't have them, but you can expect to see them in packets by the second group. Some people are just not very good at the knowledge-independent gameplay part of quizbowl and will forget pronouns, get confused during questions, reflex buzz off clues they shouldn't, etc.; those people (and unfortunately I'm often one of them) will get questions wrong because of it. Generally, I think very little can be abstracted from the discussions of particular questions. Instead of trying to tease out general theories or categories of common links, more can be accomplished from discussions of specific questions that illustrate things like "this is a terrible answer choice" or "this clue doesn't aim people to the right answer" and everything else that makes some common link questions a pile of suck.

[cvdwightw]

it's not that people don't understand the context of clues, it's that people who recognize a specific clue must use clues they don't know to figure out what the link is.

Except that this is how every tossup ever works. Again, if you hear "FTP, name this German", and you know none of the previous clues, you have to use the previous clues to determine whether you should be buzzing with Bismarck or with Johann Friedrich Overbeck.

If you don't know the early clues, then you are taking a risk each and every time you buzz on any subsequent clue that can apply to two or more things. This is inherent to quizbowl.

Bruce, you are failing to reductio my absurdum. I would hope that 100% of players recognize "German" in this context as meaning "a person originally from, or possibly living in, the area that we associate with the present-day country of Germany, or the area we associate with a historical concept of 'Germany' that was contemporaneous with this person's achievements" and furthermore recognize that there are hundreds of realistic answers based on that single clue. This is not the same as recognizing a specific clue as pointing to a very specific thing ("this clue is clearly a description of Carnation, Lily, Lily, Rose") and trying to figure out what link this very specific, known, thing has to do with the other not-known clues in the tossup.

No one else in this thread is trying to make an association between a very general clue that is not intended to stand alone and a very specific clue that under normal playing circumstances would stand alone. I would go as far as to argue that "Name this German" is simply a context clue - it provides no actual information to entice a player to buzz, unless I suppose you're debating between Bismarck and Metternich or something. I sincerely hope that you're being contrarian for the sake of being contrarian and do not actually believe that "this is a description of a very specific thing that people have knowledge of, and which under normal playing circumstances they would answer with that thing and earn 10 points" and "this is a very general description used to provide context" are the same thing.

In good common-link tossups, the context clues surrounding each very specific thing should allow someone with knowledge of that very specific thing to buzz in with the exact link that the context clues imply. In some questions, often due to transparency concerns, this is not the case, but a player should be able to buzz in with the most general plausible answer that applies to the clue he recognizes and continue to be prompted until he attains the appropriate level of specificity (of course, what constitutes a plausible answer is a debate for another thread, but it suffices to say that it is incumbent on the question writer to determine that level; to some extent, this is analogous to prompting on "Louis" in a tossup on a specific King of France named Louis, since "Louis" is the most general plausible answer that applies to any clue in that question).

[Mechanical Beasts]

I would go as far as to argue that "Name this German" is simply a context clue - it provides no actual information to entice a player to buzz, unless I suppose you're debating between Bismarck and Metternich or something.

Yeah; I've generally found that you either have

and he fought against the Lombard League. For 10 points, name this German.

in which having fought against the Lombard League is the giveaway, and

and he fought against the Lombard League. For 10 points, name this German, who had a red beard and stuff.

in which the red beard is the giveaway. I think we use "name this German" simply because it's as easy to write as "name this person" and "name the answer to this question" sounds weird--we don't pretend that, even locally, that's supposed to be a clue.

[grapesmoker]

This is not the same as recognizing a specific clue as pointing to a very specific thing ("this clue is clearly a description of Carnation, Lily, Lily, Rose") and trying to figure out what link this very specific, known, thing has to do with the other not-known clues in the tossup.

You mean, "this is clearly a description of some people on this painting, and the question is asking for a certain type of person and it's telling you that this work contains some female ones, so perhaps I should think about what kind of female persons are in this painting."

[theMoMA]

I don't know why Jerry has decided to just ignore everything I wrote and instead assume that I'm a stupid player whom he shouldn't even have to listen to because of a misguided discussion from a year ago. I didn't even neg the tossup! I was trying to buzz with the correct answer on the Breughel clue! Go back to what I said, and read the part about starting out with the assumption that a question is good on paper, and trying to explain the reality of the question-as-played to fit the assumption. That's exactly what you continue to do, and your logic is comically distorted. Clearly because one time someone buzzed on "Lucy" with "Wordsworth" when it was actually Lovelace (on a tossup in which every clue pointed specifically to Lovelace), people should be able to figure out exactly what class of people that you decided to ask about from individual clues that point to at least a half-dozen equally likely answers. If they can't see the overlap, they're clearly bad players!

No one's trying to take you down a peg as a writer, or to insist that you fall on your sword over this one question. All you have to do is say "I thought this would work, but clearly it just didn't play well, and I'm sorry for that."

I think there's plenty of value, contrary to what Marnold says (though I do agree with his assessment, just not the conclusion that we should never try to talk about common problems with questions), in classifying problems with tossups so that people don't repeat them. Jerry still seems to think that asking about a class of things that has uneasy relationships to other promptable answers is acceptable, when we can see that it creates all sorts of practical issues with playability. It's also pretty useful to talk about common links in a way that precludes discussion in the "destroy all common links because one was bad" vein.

[grapesmoker]

I don't know why Jerry has decided to just ignore everything I wrote and instead assume that I'm a stupid player whom he shouldn't even have to listen to because of a misguided discussion from a year ago. I didn't even neg the tossup! I was trying to buzz with the correct answer on the Breughel clue! Go back to what I said, and read the part about starting out with the assumption that a question is good on paper, and trying to explain the reality of the question-as-played to fit the assumption. That's exactly what you continue to do, and your logic is comically distorted. Clearly because one time someone buzzed on "Lucy" with "Wordsworth" when it was actually Lovelace (on a tossup in which every clue pointed specifically to Lovelace), people should be able to figure out exactly what class of people that you decided to ask about from individual clues that point to at least a half-dozen equally likely answers. If they can't see the overlap, they're clearly bad players!

You're not stupid, but I see you committing what appears to me to be essentially the same mistake, where you argue from your particular experience about how the question played out to some generalization of categories of common link questions that doesn't make a whole lot of sense to me. I don't understand how in this question at hand, which I grow tired of defending, you could "try" to buzz with the correct answer on the Breughel clue. It's got a proper name, that name is "Chidlren's Games," you either know that or not. If you know the answer, then I can't imagine what possible barrier there could be for you to give it. I find your arguments about how this question played out unpersuasive because I saw two different incidents of this question playing out without any problems, and also because I provided what I think is sufficient evidence of the fact that really no other answer was acceptable for this question. I think it is rather you who are ignoring these facts and placing a premium on your subjective experience and I see no reason why I should do the same. I just don't accept your reasoning that this question not playing well for you is indicative of a flaw in the question.

No one's trying to take you down a peg as a writer, or to insist that you fall on your sword over this one question. All you have to do is say "I thought this would work, but clearly it just didn't play well, and I'm sorry for that."

No dice. I see no reason to apologize for a question which did in fact play out just fine in my presence, twice.

Jerry still seems to think that asking about a class of things that has uneasy relationships to other promptable answers is acceptable, when we can see that it creates all sorts of practical issues with playability.

You sure are intent on ignoring those proper titles I provided you with. How unfortunate for this discussion indeed.

[Auks Ran Ova]

I don't understand how in this question at hand, which I grow tired of defending, you could "try" to buzz with the correct answer on the Breughel clue.

I'll clear this up then: he could only "try" to buzz with the correct answer because, at the time, I was busy negging with "peasants". Later, I realized I had conflated Children's Games with Peasant Wedding, and my anger at the tossup diminished. Incidentally, I still think "children" was sort of a silly thing to write on, but I feel a lot less personally hosed by it, since the mistake I made was my own fault.

[theMoMA]

In the EFT thread, I believe we heard that the question played poorly at least three times. How is that acceptable? People were confused and negging with pseudo-right answers on the clues after the Breughel clue, which are not specific titles. Regardless of what you think, this is not a discussion about my experience playing the question, or even something spurred on by the fact that my teammate negged it. It struck me as a poor execution of a dubious idea, and that seems to be born out by the fact that people were negging the question. The clues in your tossup simply don't stand alone to point to one answer, and the relationship between the possible answers is such that even the notion of where or what to prompt on is muddled. And those issues played out in games. But you've seen the tossup played well in person twice, and one time I had a misguided discussion with you, and people shouldn't have been so stupid and figured out what you wanted, so clearly the tossup was good.

[Cheynem]

I looked through the EFT thread, and I'm confused as to Andrew's claim that at least three people mentioned the children tossup not playing well.

The first complaint about not the "children" tossup not playing well came from Andrew. Carsten also incorrectly answered it, but he was in the exact same room as Andrew, so I'm not sure if this really counts as as different manifestation of it. Charles Hang says he was just confused what the answer actually was; I'm not sure if you can really determine this to be him saying "I don't think it played well."

Things then move into a general discussion of common links and it's tough to figure out who's actually talking about the children tossup. It seems to me no one really is talking about that tossup but railing against unclear common links in general (at least that's what I was doing, I never played the children tossup so I have no idea if it would have confused me or not).

[theMoMA]

I believe Mike Sorice's post talked about the pseudo-right answers that were given in his room.

[Cheynem]

Ah yes, just noticed it. Thanks.

[Megalomaniacal Panda on Absinthe]

I don't understand how in this question at hand, which I grow tired of defending, you could "try" to buzz with the correct answer on the Breughel clue.

I'll clear this up then: he could only "try" to buzz with the correct answer because, at the time, I was busy negging with "peasants". Later, I realized I had conflated Children's Games with Peasant Wedding, and my anger at the tossup diminished. Incidentally, I still think "children" was sort of a silly thing to write on, but I feel a lot less personally hosed by it, since the mistake I made was my own fault.

I'll point out the exact same thing occurred when I played the question. I was about to buzz with "children" at the same point and someone on the opposing team negged with "peasants." I was sure the answer was "children" after the mention of the Runge and Brueghel paintings, but I was completely confused by the end of the tossup and fumbled around by bemusedly saying "Girls!? Daughters!? Sons!? What do you want from me?" The description of the two Sargent works and the Cassat actually led me away from the correct answer; I'm not sure I would have reverted to my original answer of "children" even had Mike prompted me.

My objection to this tossup isn't that I feel hosed or cheated of points; the illative failure is ultimately mine. That said, I think "children in art" seems an unfortunate and easily confused answer line for a tossup.

[Auroni]

Yeah, I think that once you start talking about "female" and "male" ones of these, it throws several people with the knowledge off track. Which is probably why this tossup might not have been the best idea, since there aren't giveaway-level paintings whose title has the word "children"

[Matt Weiner]

I'm sympathetic to any sort of empirical argument; "there might be something wrong with a question that produces way more negs than other questions in the packet" is a valid point. However, this just means we should look into the question. If, upon looking into it, we find that "people have insufficient knowledge about the clues and make wrong buzzes, just as they might neg Poltava for Narva for the same reason," "people were playing the questions online, where things are not taken as seriously and negs have no meaning, and just throwing out random words on a lark," and "people have already decided that common link questions are impossible to play on and can only be answered by guessing and hoping to get lucky" are possible scenarios, then we don't have to conclude that a question was faulty after performing our investigation.

[Captain Sinico]

Right, but, in addition to having played poorly, this question had major issues specifying a unique answer, which, to every appearance, is why it played poorly. To reiterate a point I made in the other thread, that is a problem endemic to common link questions that wouldn't be tolerated from other types of questions and oughtn't be tolerated from common link questions, either. Questions must contain a number of substantive clues, as many of which as possible are uniquely identifying.

MaS

[No Rules Westbrook]

So, I don't much like this question, but hey...it was trying to be creative in an easier tournament and it's not terrible. On first blush, it seems too easy for people to "sniff out" the answer without any real knowledge - but that's easy for me to say just looking at in text and not having played it. After all, some people did say "peasants" so there is at least one semi-viable other avenue of thought.

I don't like the idea that we should be callous to all negs based on "insufficient knowledge." There are lots of times when I've reworded a tossup because I thought it was encouraging people to buzz with a wrong answer - even though I could have said "if you buzz with that answer there, you have insufficient knoweldge - it's clearly not right, and I don't care whether you neg." I think it's a good practice to do that. When you read any tossup over, you have to look at it in the lens of a real match with real players - people don't hear clues the same way the do when they're read, people have to make quick decisions about whether it's worth it to buzz in and answer a question at certain points with something less than positive knowledge, people's memories of prior clues in a tossup is highly variable (some players are really good at keeping all given clues in their mind - others, like me, tend to focus more single-mindedly on the present clue).

All that said, we all know the perils of the common link and I don't find this question to be particularly odious in that regard. I am sympathetic to the notion of holding common links to the same standard as everything else. Like always, the thing I get most concerned about is "sniff out-ability" (coining this term in place of the often-attacked transparency) - I really hate it when qb forces people to deduce or speculate at answers instead of just consulting their knowledge banks in whatever subject the question is in.

[grapesmoker]

I'm pretty tired of defending this question so this will be my last contribution to this debate. So far, the empirical claims that this question played poorly seem to rely solely on a bunch of people negging with "peasants," which is a wildly incorrect answer that doesn't fit any of the clues. I don't know how this is my problem or why anyone is claiming they were misled; that answer was simply not correct even under the most generous interpretation.

So, I don't much like this question, but hey...it was trying to be creative in an easier tournament and it's not terrible. On first blush, it seems too easy for people to "sniff out" the answer without any real knowledge - but that's easy for me to say just looking at in text and not having played it. After all, some people did say "peasants" so there is at least one semi-viable other avenue of thought.

Semi-viable is not correct! Stop pretending that people should be entitled to points for kind-of guessing what the answer might be. Just because Breughel once painted some peasants doesn't mean that this painting contains peasants or is titled after peasants. An alternative avenue of thought that leads you to the wrong answer is a wrong avenue.

I don't like the idea that we should be callous to all negs based on "insufficient knowledge." There are lots of times when I've reworded a tossup because I thought it was encouraging people to buzz with a wrong answer - even though I could have said "if you buzz with that answer there, you have insufficient knoweldge - it's clearly not right, and I don't care whether you neg." I think it's a good practice to do that. When you read any tossup over, you have to look at it in the lens of a real match with real players - people don't hear clues the same way the do when they're read, people have to make quick decisions about whether it's worth it to buzz in and answer a question at certain points with something less than positive knowledge, people's memories of prior clues in a tossup is highly variable (some players are really good at keeping all given clues in their mind - others, like me, tend to focus more single-mindedly on the present clue).

Everyone tries to word their questions to avoid hosing people. But at some point, things like proper titles come into play, and then there is no longer any excuse for buzzing with an answer that literally does not match any of the clues.

I'm clearly not going to convince anyone whose argument is "but it played poorly," despite the fact that it played just fine in at least two instances and the times when it apparently played poorly were both instances of completely incorrect negs. You're welcome to ignore all the facts at hand and continue with the delusion that this question somehow screwed you; I don't really care enough to continue this argument.

[Megalomaniacal Panda on Absinthe]

Jerry, I don't think a single person in this thread is actively bemoaning the tossup because they buzzed with 'peasants.' In fact, both Andrew Hart and I have told you we object to the content of the tossup independent of the fact we assuredly could have buzzed correctly. I'll reiterate my problem: the second half of the tossup, dealing with two Sargent paintings and a Cassat, provided clues that in no way uniquely pointed towards an answer of "children." Again, I'm not claiming I was cheated of points; the inferential disaster was mine. But there's clearly some evidence that this tossup demanded an unreasonable set of inferences for a player to make, especially since we are discussing players who clearly had knowledge of one of the titles which ostensibly justifies the non-uniquness of subsequent clues (unless you are contending that Andrew and I don't have knowledge of, at bare minimum, the title of Children's Games).

I am not particularly invested in the fate of this tossup, but I think it exemplifies a class of unfortunate common link which Andrew adumbrated in the original post. Thus I think you are being strangely sanguine regarding a question which clearly proved confusing even for people who knew the right answer and would have attempted to give it.

[No Rules Westbrook]

Just to be clear, you're misreading my post Jerry - I meant "semi-viable" in a good way for you. I was looking at the tu and thinking children was too obviously the answer - but someone said peasants, which is probably another possible (not right) answer - this works against the "what else could it be!" transparency critique.