Some constructive ECSO criticism

[Birdofredum Sawin]

As I was playing the ECSO, I was reminded of a few points I wanted to make about question writing. So here goes, with concrete examples (props to Jerry for making the set immediately available, which is very encouraging to potential commentators).

These remarks shouldn't be construed as criticism of particular individuals: I have no idea whether the questions discussed below were written by the editors, the ostensible authors of the packets, or super-intelligent squirrels.

POOR "CREATOR FROM DESCRIPTION OF WORKS" TOSSUPS:

Recently, we've seen a vogue for tossups on authors from descriptions of increasingly less-obscure works. First, let me say that I am wholly in favor of this kind of tossup; in fact, I've written more than my fair share of them over the last year or two. That said, it's very easy to write this kind of question badly. Consider this example:

A disputed will formed the basis of his Orley Farm, and obsession with an innocent wife was behind his He Knew He Was Right. The financier Melmotte stars in The Way We Live Now, and duke of Omnium began appearing in the political Can You Forgive Her? Withdrawing to Sussex in his final years, he began as a post office clerk and wrote two sets of novels, one featuring characters like Crawley, Thorne, and Mrs. Proudie, and the other starring Plantagenet Palliser. FTP name this English novelist of the Parliamentary Novels and the Chronicles of Barsetshire, including Barchester Towers.
Answer: Anthony Trollope

For all intents and purposes, this question might as well begin "This author wrote Orley Farm," since nobody could possibly buzz on the first "description." Disputed wills from the basis of lots of books, after all. This is really no better than tossups that begin "This artist favored dark and tempestuous scenes," or similar "impressionistic" lead-ins of yore.

Also, note that one of the best uses of this kind of tossup is to provide new ways of asking about well-known authors. You have to be more careful when using them to ask about lesser-known figures, as the following question illustrates:

In one of his short stories, an old woman sees a mysterious man outside of her cave before having a dream in which she swallows the title animal. In another, a linguistics professor tries to procure camel-udder boxes but ends up kidnapped, while in another, the murderer of Driss is caught when he tries to sell Filala leather. In a novel by this author, the distinction between a traveler and a tourist is explained and Captain Broussard offers advice on typhoid, after Eric Lyle’s car is shunned and Tunner attempts to initiate an affair. For 10 points, name this author of “The Scorpion,â€

[Chris Frankel]

I didn't play the tournament, but I did look at the packets.

One annoying trend I'd like a moratorium on is "common link" tossups. By that I refer either to "the name's the same" list tossups or "everything mentioned fits in category X."

I don't think they're inherently bad per se, but I'm getting sick of seeing them proliferate to the point where you can expect them in every round (and I think some categories, like myth and music, are being seriously messed up by having common link questions force out the traditional "write a detailed tossup on a single topic to reward players with deep knowledge of the topic" style). And frankly, most of them are usually poorly and lazily written.

By poorly written, I mean that the answer selection is often questionable, and can tend to emphasize rather silly and esoteric subjects. A good common link tossup combines a series of subjects that all have some importance, canonical value, and accessibility, and weaves deeper and more obvious bits of information about each in a fluid fashion. A tossup on French kings named Philip might be one example, or to use a real example, I enjoyed the "rulers named Lothair" tossup from ACF Nationals, since there is more than one ruler named Lothair that one with a general background knowledge of the Holy Roman Empire would be expected to recognize.

It would also be against the spirit of learning to say outright that some subjects are worthless and should never come up, but I think there many topics out there that one can look at and objectively say, "there isn't a compelling reason to force this answer into the canon; it's not significant or interesting enough to merit tossup status."

So as determined as one might be to impose Hottentot myth on the canon, just because one happens to Google around and find out that, say, the Hottentot had a forest god who took the form of a dog (I'm just making this up), and say, "Well, hey... they had a dog god, and Cerberus was a famous mythological character who was also a dog... LET'S WRITE A TOSSUP ON DOGS!" It's not too far off to realize that the leadin clue is going to go over the heads of pretty much everyone who doesn't study Hottentot myth in depth, and you just end up wasting 1-2 lines when you could just take a few extra minutes and look for a good leadin for Cerberus in its place.

Or take the awful Alexander II tossup. There were ~260 popes; I don't think it's a shock to anyone to realize that not all of them were profoundly significant, not all of them are studied in detail, and that, therefore, it would be foolish to think that all of them would be fair game for a tossup. Just because Pope Alexander II had the same name as a very important Russian czar doesn't mean the former ranks up there with the latter in importance. Again, you're wasting 1-2 lines that won't mean anything to nearly anybody, and that could be used to research deep clues that could reward the people who have studied Czar Alexander II (which is a likely-to-exist group because of his historical significance). I almost get a feeling like people are writing common links questions out of laziness to avoid having to find new leadins and deep clues on a single major topic in lieu of just going through Wikipedia's disambiguation page or an equivalent and just loading up on lesser names that are shared.

Anyways, I don't outright hate common links questions, and have even been known to write the _occasional_ one for every few packets I write. But I think their execution is usually poor, and people seriously need to tone it down, as will hopefully be the case for CO. I just dread the day when I hear the 10 line ACF version of "Symmachus, Anterus, Conon, Marinus I... Urban II, Alexander VI, and John Paul II" by someone whose train of logic goes "hey they're all names of popes, and popes tend to be important, so let's write a tossup on 'popes' so I can force all the former names in the canon!"

[Matt Weiner]

The problem that Frankel alludes to with people using common-link questions to ignore pyramidality is real and should be addressed. If you write, for example, a people-named-Alexander-II tossup (one where all the Russian clues are actually about him and not about Alexander III or Nicholas II) then you need to start with a proper 4-5 line tossup on the Russian tsar and then put your Pope and Scottish king clues in front of that. You can't go from the clues about obscure Alexander IIs right into the easiest material about the tsar because then you are causing a buzzer-race on what you admit, by using it as the last clue, is the thing most people will know in your list of people named Alexander II.

(This particular tossup was not actually guilty of the above, though it had other problems; it's just a convenient example.)

However, I disagree with the conceptual opposition that Frankel expresses. The twofold point of writing questions like this is:

1) to put things that most people do not in fact study, like "Hottentot mythology" into questions in a way that rewards those with such rare knowledge but still lets people with more common knowledge actually get the question
2) to continue keeping difficulty down and asking about important subjects like Greco-Roman myth in every round even though it is very difficult to come up with good leadins for figures such as Hera or Mercury

Myth enthusiasts such as Seth Teitler or very talented question writers such as Frankel can probably write new questions on major Greco-Roman gods all the time that have leadins challenging and original enough for a game between this past weekend's Ferrari-Samelson-Yaphe-Turner team and Wolpert-Lafer-Westbrook-Kemezis team. However, expecting every writer in a tournament with 14 submitted packets to be able to do that, week in and week out, is completely unrealistic. It's much better and easier to show the bulk of writers, who are competent but not great, how to write these "dogs" tossups properly. And it does get esoteric knowledge into the game without crowding out more mainstream clues.

[Nathan]

"This isn't a dreadful question, but unless you are a real Bowles expert it's hard to see how you can buzz before a point roughly halfway through the final sentence."

Although the question could have benefited from some more clues (he was a notable composer etc.) anyone who's actually read The Sheltering Sky would have a shot at getting the tossup after Eric Lyle and certainly after the mention of Tunner and the affair (throw in that it occurred on a train trip and it should be pretty gettable there)...

agree with your general points though.

[cvdwightw]

Consider this stretch of three consecutive bonuses from one of the ECSO packets:

[difficult art bonus]
[difficult history bonus]
[find-your-ass art bonus]

I think this was a result of merging two partial(?) packets of widely varying difficulty. I skewed most of my questions to the easy side, leading to several bonuses where most good teams should have gotten at least twenty. The ass-easy bonus (and, probably, all such bonuses in that packet) was mine. I would like to think that I wrote my bonuses at a fairly consistent difficulty. Most of the harder questions in the packet were not mine, and also seemed written at a fairly consistent difficulty. I think the "widely varying bonus difficulty" would have been less apparent in that packet if the packets had been combined differently.

[canaanbananarama]

While I made the same logical leap on the Diwali tossup (sounds like an Indian religious festival, can anything else really come up at this level of difficulty?), your claim on the "mahdi" tossup, which I wrote, is misinformed. There isn't a large answer space for what it could be, I'll grant you, but making a logical leap "this is a title that Islamic people take" and jumping to "mahdi" is ill-advised, as I could have wrote an adequate tossup on "imam" and a less good tossup on "caliph" and I don't believe any of those topics would have been uncalled for within the context of the stated tournament's difficulty. So, there was, granted, a limited answer space, which will naturally happen as there isn't terribly much you can ask about in terms of Islam that would be a worthy subject for a tossup in a tournament such as this, but your buzz reflects a guess and not the only possible answer in that situation. Congratulations on that, but I don't see your logic in claiming that that was the only possible answer or that it was even reasonable for you to buzz in that situation.

Charles Meigs

[recfreq]

I wrote the Trollope. I agree that it's not that great, but depending on what you mean by "poor," I may or may not agree. You could definitely write a longer, more detailed summary of Orley Farm in lieu of what is in that question, but I don't think the "disputed will" is akin to "tempestuous tones" in painting, because at least it is something specific from a specific work, even if the clue itself can be general. If you're looking to narrow down the answer space, you don't always want the most specific clues, but only clues that get you closer to what you want at the beginning of the question.

I favor going from a bit more general to a bit more specific clues in a TU, b/c it gives you more to think about initially before reflexes take over towards the end. That's one particular style, and not one that is favored by everyone (or possibly anyone). I understand the need for specificity, but I feel that sometimes that leads to longer questions. So in the Trollope case, you could write a description of Orley Farm as well as of He Knew He Was Right, and make the sentence roughly twice the length. There's a tradeoff b/t the number of words you use and the content of the question, and I simply chose the former in this case, b/c personally, I felt that Jerry's questions in the past were a bit long. If you noticed from the formatting, I tried to keep all my questions at roughly the same length, so that the player can recognize more or less when certain types of clues should come up. I happen to think that uniform length across TUs is important, and perhaps this has forced me to make the Trollope question a bit shorter. But I agree with you that I could do better as a question writer, and with more time for that question, I probly could. Sorry, I was pressed for time. I would have written a good question if I knew ahead of time that it was guaranteed to be used.

I don't know if the pain BO was harder than the Pope BO. If you're a scientist, the pain and placebo are very gettable 20s, and several of my teammates knew about allodynia. I thought Arbuthnot was sufficiently difficult, but perhaps I'm wrong. I recall ACF Nats having Jacob-Monod, operon, and plasmid as a BO answer set. That seems to be almost trivial, and I could have complained analogously that that belongs in a HS set, not ACF Nats, by comparison with, say the lit questions in that round. My main point is that it's hard to compare difficulty across categories. On the other hand, your argument about the two art bonuses, one beginning with Honhorst, and the other three famous paintings, is valid indeed. It seems to me that the most important and easily accomplished thing to do to the bonuses is to make the questions for each category have the same level of difficulty (of course you want to shoot for global uniformity as well, but the local one for each category has got to be achieved). Again though, neither of those bonuses were very good.

[No Rules Westbrook]

I don't agree. It really doesn't matter to me whether all the lit bonuses in a given set are of equal difficulty, just whether all of the bonuses in a given packet are of roughly equal difficulty...for obvious reasons, so that teams don't hit the proverbial jackpot on luck of bonus draw. Any packet that combines questions from different people - whether players on the same team who submitted the packet or different editors tending to different subjects in the packet - will probably have these types of difficulties if they're not globally edited. We all have pretty varying conceptions of appropriate bonus difficulty, even when you specify a set difficulty level for a tournament. For example, my default when possible for almost any lower-level tourney is - 10 points that are almost a giveaway, 20 points for the handful of experienced teams there i.e. something generally canonical, 30 for either noone or one or two specialists. For a harder-level tournament, I drop the easy 10 and make that something equivalent to the second level. I'm not saying this is the superior way to conceive of appropriate bonus difficulty, I'm just giving an example. When you combine questions into a packet, though, you're probably not going to get this kind of consistency.

[Birdofredum Sawin]

I wrote the Trollope. I agree that it's not that great, but depending on what you mean by "poor," I may or may not agree. You could definitely write a longer, more detailed summary of Orley Farm in lieu of what is in that question, but I don't think the "disputed will" is akin to "tempestuous tones" in painting, because at least it is something specific from a specific work, even if the clue itself can be general. If you're looking to narrow down the answer space, you don't always want the most specific clues, but only clues that get you closer to what you want at the beginning of the question.

I favor going from a bit more general to a bit more specific clues in a TU, b/c it gives you more to think about initially before reflexes take over towards the end. That's one particular style, and not one that is favored by everyone (or possibly anyone). I understand the need for specificity, but I feel that sometimes that leads to longer questions. So in the Trollope case, you could write a description of Orley Farm as well as of He Knew He Was Right, and make the sentence roughly twice the length. There's a tradeoff b/t the number of words you use and the content of the question, and I simply chose the former in this case, b/c personally, I felt that Jerry's questions in the past were a bit long. If you noticed from the formatting, I tried to keep all my questions at roughly the same length, so that the player can recognize more or less when certain types of clues should come up. I happen to think that uniform length across TUs is important, and perhaps this has forced me to make the Trollope question a bit shorter. But I agree with you that I could do better as a question writer, and with more time for that question, I probly could. Sorry, I was pressed for time. I would have written a good question if I knew ahead of time that it was guaranteed to be used.

This is just completely wrong. The problem with clues like "he painted dark, tempetuous scenes" was that they were impossible to buzz on. A player simply couldn't know what the question writer had in mind, because the description, vague as it is, is applicable to so many different people. A clue like "a disputed will is the basis of one of his books" is faulty in exactly the same way. Lots of people, including most Victorian novelists, wrote books "based on disputed wills." I've actually read Orley Farm within the last two years, and I'm telling you that it's simply not possible to buzz on this question until you hear the title. That means that you might as well have written "This author of Orley Farm" as your first clue, which means that this question only pretends to be an "author from descriptions" tossup and is in fact effectively an "author from list of works" tossup.

If you're worried about length, then only describe 3-4 works rather than 6-7. The point is, you have to actually describe those works. Tossing out a hopelessly unspecific nugget about them does not count as "description."

Andrew

[recfreq]

If you're worried about length, then only describe 3-4 works rather than 6-7. The point is, you have to actually describe those works. Tossing out a hopelessly unspecific nugget about them does not count as "description."

OK, I see now that 1st clue is really terrible; this has been very helpful, I just assumed it was passable before. I see also from the lit people that the preference is for specifics on a few works as opposed to generalities on a list of works. I agree completely for a well known author, but I don't know if that's necessarily preferable for a lesser known author (as in Bowles case), but in this case, writing lit as a scientist probly got in the way, and I'd probly be similarly upset on a bio question. So I guess we should have specificity upfront then move on to generalities; or to say it better, descriptions of specific works moving on to titles of some other works. So how about this rewriting of the TU (in an effort to keep it at roughly the same length as well, excuse my unchecked grammar)?

His character Mary Snow is groomed to be the wife of Felix Graham in his novel about Lady Mason's forgery trial, Orley Farm. Accusation of the innocent Emily Trevelyan's extramarital affair forms the basis of his He Knew He Was Right. The financier Melmotte and the duke of Omnium are characters in his The Way We Live Now and Can You Forgive Her? An ex-post office clerk who withdrew to Sussex, he wrote two sets of novels, one about Crawley, Thorne, and Mrs. Proudie, and the other centering on Plantagenet Palliser. FTP name this author of the Parliamentary Novels and the Chronicles of Barsetshire, including Barchester Towers.
Answer: Anthony Trollope

Yeah, I'll definitely keep in mind to avoid sweeping generalities like "disputed will is the basis..." in the future. This is great feedback. Thanks.

[Matt Weiner]

His character Mary Snow is groomed to be the wife of Felix Graham in his novel about Lady Mason's forgery trial, Orley Farm. Accusation of the innocent Emily Trevelyan's extramarital affair forms the basis of his He Knew He Was Right. The financier Melmotte and the duke of Omnium are characters in his The Way We Live Now and Can You Forgive Her? An ex-post office clerk who withdrew to Sussex, he wrote two sets of novels, one about Crawley, Thorne, and Mrs. Proudie, and the other centering on Plantagenet Palliser. FTP name this author of the Parliamentary Novels and the Chronicles of Barsetshire, including Barchester Towers.
Answer: Anthony Trollope

His character Mary Snow is groomed to be the wife of Felix Graham in his novel about Lady Mason's forgery trial. Accusation of the innocent Emily Trevelyan's extramarital affair forms the basis of another novel by him. A third novel features the financier Melmotte, while the duke of Omnium is in another. This author was an ex-post office clerk who withdrew to Sussex, and he wrote two sets of novels, one about Crawley, Thorne, and Mrs. Proudie, and the other centering on Plantagenet Palliser. FTP name this author of Orley Farm, He Knew He Was Right, The Way We Live Now, and Can You Forgive Her?, as well as the Parliamentary Novels and the Chronicles of Barsetshire, which includes Barchester Towers.
Answer: Anthony Trollope

Remember that plot/character knowledge is almost always > title knowledge.

[No Rules Westbrook]

Meh, I don't like stacking all the titles at the end. Obviously, you have to use your judgment in not throwing out a title that a bunch of people know in the first line. But, I don't like questions that seem to run away from titles just in case people might know them.

[ValenciaQBowl]

Andrew's point about providing uniquely identifying clues in a TU describing literary works is well taken, as, in general, is Matt's point about titles. But the purpose of TU clues is to differentiate levels of knowledge, obviously, and I wonder if in the specific Trollope question above if Ray's rewrite above is okay mentioning "Orley Farm" at the end of the first line. I mean, have we gotten to a point where after providing a one-line (15+-word) description (with characters), he couldn't just say "Orley Farm"? I've never heard of it, and though that doesn't prove anything, I am (in my own humble self-examination) a lit specialist, and I wonder how many players could buzz on "Orley Farm." My guess is not nearly enough to make one have to leave it till the end, where, as Ryan noted, there is suddenly a crowded run of titles. (Another note is that for those trying to learn new clues at the time, it can be hard to connect the plot points to the title to be ready for the next tournament, though of course one can pore over the packets later).

One has already offered a chance for Andrew, Ray, and whatever handful of others who have actually read (or read a summary) of that obscure work a chance to be rewarded for deep knowledge, and I don't think it would be bad to give the question to someone who's worked hard enough to be familiar with all titles by Trollope that early. Otherwise, even in a masters-level tournament, whom are we differentiating?

I don't say this to nitpick this one question; since this trend of describing (rather than naming) lesser-known works is here to stay, we all need to be thinking about how long we're making the hoipolloi wait in what are often very long questions. Another example is the Bowles question above--a short story is described in which the woman "swallows the title animal" ("The Scorpion," apparently): if that title were given at the end of that first sentence, would it make the question too easy? Again, I would guess that a miniscule fraction of players could get Bowles from that story title.

Just a dinosaur's two cents. I presume Mr. Berdichevsky is even now endeavoring to make my (and Billy's) CO pack questions harder . . .

[Mr. Kwalter]

Andrew's point about providing uniquely identifying clues in a TU describing literary works is well taken, as, in general, is Matt's point about titles. But the purpose of TU clues is to differentiate levels of knowledge, obviously, and I wonder if in the specific Trollope question above if Ray's rewrite above is okay mentioning "Orley Farm" at the end of the first line. I mean, have we gotten to a point where after providing a one-line (15+-word) description (with characters), he couldn't just say "Orley Farm"? I've never heard of it, and though that doesn't prove anything, I am (in my own humble self-examination) a lit specialist, and I wonder how many players could buzz on "Orley Farm." My guess is not nearly enough to make one have to leave it till the end, where, as Ryan noted, there is suddenly a crowded run of titles. (Another note is that for those trying to learn new clues at the time, it can be hard to connect the plot points to the title to be ready for the next tournament, though of course one can pore over the packets later).

One has already offered a chance for Andrew, Ray, and whatever handful of others who have actually read (or read a summary) of that obscure work a chance to be rewarded for deep knowledge, and I don't think it would be bad to give the question to someone who's worked hard enough to be familiar with all titles by Trollope that early. Otherwise, even in a masters-level tournament, whom are we differentiating?

I think to some extent you and Ryan are right, these days a lot of writers are leaving the middle part of the pyramid out by loading all the titles at the end. However, I don't think that the leadin should include a title clue, even if some of the subsequent "plot summary" clues should then mention the works in question. If you look at a list of trollope novels and memorize them, then you'll get that tossup off the mention of Orley Farm, which to me marks a violation of the pyramid that comes way too early in the tossup.

[Birdofredum Sawin]

I agree with Chris. One of my criticisms of that Bowles tossup is that all the titles are bunched at the end. I'm really not sure what the point of that backloading is supposed to be, as I find it hard to imagine that many players have memorized lists of Paul Bowles short story titles. In general, I prefer the form "In one of this author's works, some stuff happens. In another of his works, some other stuff happens. In addition to X and Y, he wrote ..." This tends to produce more of a pyramid, and also cuts down on confusion. I've heard too many "description of works" tossups this year in which no titles are revealed until the end; in such tossups, the descriptions tend to run together, causing me to lose track of what is purportedly being described ("are they still talking about that first story? Is another work being described now?") so that I have no choice but to wait for a trigger word. I don't think the game is so rife with title memorizers that we need to go to such extreme lengths to foil them.

Andrew

[Dan Greenstein]

I agree with Chris. One of my criticisms of that Bowles tossup is that all the titles are bunched at the end. I'm really not sure what the point of that backloading is supposed to be, as I find it hard to imagine that many players have memorized lists of Paul Bowles short story titles. In general, I prefer the form "In one of this author's works, some stuff happens. In another of his works, some other stuff happens. In addition to X and Y, he wrote ..." This tends to produce more of a pyramid, and also cuts down on confusion. I've heard too many "description of works" tossups this year in which no titles are revealed until the end; in such tossups, the descriptions tend to run together, causing me to lose track of what is purportedly being described ("are they still talking about that first story? Is another work being described now?") so that I have no choice but to wait for a trigger word. I don't think the game is so rife with title memorizers that we need to go to such extreme lengths to foil them.

Andrew

I prefer the following form:

This author wrote about the Action A of Character N and Character P in Place Z in Name of Work. He also wrote about Characters C, D, and E in a novella about Occupation B, Name of Work 2. In addition to Name of Work 3 and Name of Work 4, he also wrote about Action F of Character G in Event N. FTP, name this Nationality T author of Name of Work 5, Name of Work 6, and a Descriptor Q, Name of Work 7.

In the previous, Name of Work 7 is described in the sentence before "for ten points."

I like this form because it removes the "losing track of what is purportedly being described" lamented by Andrew. The description of each work is limited to one sentence and the description is directly tied to the work by having the work mentioned immediately thereafter. If you have two or three or more sentences describing one work, then there is enough material to write the tossup or a bonus part about that work.

[Nathan]

Andrew wrote:
"I don't think the game is so rife with title memorizers that we need to go to such extreme lengths to foil them. "

I would generally agree. I've never heard of Orley Farm and if you can get it off the plot or simply knowing the title...god bless ya.

On the other hand...I see why Matt wanted to move "He Knew He Was Right" further back...its a title that shows up in Trollope tossups pretty consistently and would generate a lot of reflex buzzes.

Specifically, like, say, "It Can't Happen Here", "He Knew He Was Right" is a memorable title and therefore conducive to list memorizers in a way that "Orley Farm" is not.

So, make the tossup flow, don't lump everything together confusingly, but also pay attention to what might be provoke unearned reflex buzzes too early. Use your judgment.

[No Rules Westbrook]

The two writing "formats" suggested work fine. There's no need to be formulaic or mechanical about them as long as the writer has some common sense about what clues go where. I think that in general, on lit and other questions, there is way too much running away from clues lately. Sometimes, it really does seem like people are afraid to drop "keywords" because they think someone may have memorized them (gasp, you mean people do that?). This is understandable especially if you're not sure where the clues in your question fit pyramidally...but, you may as well just try your best to put them in order using common sense, because it's almost impossible to produce a good question being purposely vague. And, if people know the title of the eighth most famous Paul Bowles short story...good for them, they win. It seems that when you're sitting there writing a question, you tend to give the memory of players way too much credit...and, then you remember "oh yeah, there are millions and billions of little nuggets of info to remember, that's pretty tough to do."