Science (DEES)

[Victor Prieto]

This thread is reserved for the discussion of the science distribution.

For clarification, I edited biology, chemistry, and earth science, while Brian edited physics, math, astronomy and computer science.

This set was originally intended to have 15/15 physics, 15/15 biology, 10/10 chemistry, 10/10 math, 4/3 earth science, 3/4 astronomy and some computer science. While the final set definitely had 2/2 physics/biology in every packet, the rest of the distribution sort of went to hell. I know for a fact that there ended up being extra chemistry and earth science. Once we realized that this was going to be a race to the finish, three weeks ago, our concerns beyond the basic distribution went out the window, and we just focused on getting every packet with 4/4 playable questions.

Personally, my science editing extremely suffered from the sudden head-editing responsibilities thrust upon me. I had a whole bunch of time earmarked in the past three weeks that basically disappeared because of all the organizational/administrational stuff I was doing. I'm really sorry that despite the fact that I was working on this tournament for the past couple months, ironically, my science may have been one of the subpar categories.

Thanks to Adam Silverman, Cody Voight, Auroni Gupta, Andrew Wang, and Joelle Smart for playtesting, and the other non-science people that were there when I playtested each of the biology easy parts (science editors: highly recommend doing this). I'd also like to give super-praise to Cody Voight for saving us with a handful of computer science tossups to stick in as the fourth tossup in a packet here and there. Absolutely invaluable.

[Excelsior (smack)]

The science in this tournament was generally quite pleasant. I have no particular comments on the chemistry (I mostly zoned it out), but the rest of the distribution was mostly enjoyable. I understand that this tournament was written under ridiculous time constraints; as such, I of course don't fault you for having some clunkers in the set. The below is more for future editors to take note of.

― As I have probably posted about before, tossups on "WORD" are generally not a great idea when you're going for a common link - since the various clues are in a "WORD" tossup are unrelated to one another, it is difficult to build a mental model of what the question is talking about. These questions lack thematic unity, which, I am led to believe, is a feature generally desirable in common link questions. "WORD" tossups seem to show up most frequently in math, but this time, it was the CS tossup on "vector" that was a real bear.
― The tossup on "odd" claimed something along the lines of "all groups with this property are solvable according to the Feit-Thompson theorem". In retrospect, I obviously see what the question was going for, but the wording is confusing: there is no such thing as an "odd group" (Max Schindler claims that this is valid nomenclature for a "group of odd order", and I guess he's technically right, since I found a paper that calls it that, but I'm not finding very many references to it - the usage is, at the least, rare enough to be confusing).
― You should be very wary of science questions where the answer is a symbol or letter. I have written these questions myself, but only for high school tournaments, where there is no need to dig into things that are uncommon enough that there might be competing notations. It is also important to be wary of the fact that it is entirely possible to know things without knowing the corresponding notations for them. Some notation is certainly important and the kind of thing you would want to ask about at the appropriate point in a question, e.g. chemical structure diagrams, basic mathematical notation, Einstein tensor notation, and so forth. On the other hand, the fact that "Q" is the letter used to denote the heat output of a nuclear reaction is not intrinsically important. Yes, it has a Wikipedia article, and if you were a nuclear physicist, you would surely have encountered a "Q value" at some point in your life, but this is a case where knowledge of the notation is a poor proxy for the underlying concepts denoted by the notation (this is again less of an issue in high school, where clues about important notations can make good middle clues sometimes). I believe that the "Q" and "gamma" tossups suffered from this in a few clues, but I would have to look at them again to be sure.
― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

There is also a correction that should be made for future mirrors: the bonus on random walk / Wiener process / binomial distribution mentions Wiener processes in the first part.

But, again, the science was generally pretty good. Good show, folks.

[RexSueciae]

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

The Virginia packet did not contain a tossup on normal forms when it was submitted, so it'd seem that the question was added by one of the editors. (The tossup went dead in the St. Lukes-Penn A game, and both teams started raging when the answerline was read.)

[Brian McPeak]

Thanks to Victor for making this tournament happen. Thanks to Stevejon Guth, Billy Busse, and Cody Voight for giving me a lot of feedback on my questions. Also thanks to Cody for writing the CS tossups (Vectors, Normal Form, Knuth) and basically rewriting some of the CS bonuses that I had messed up.

I appreciate the specific feedback Ashvin. To address some of what you said:

The odd question was my fault. It was submitted with the answer "odd order [or odd degree] ". In the original question, every clue pointed to order or degree. However, it was too hard so I put some things in there like the Fourier series stuff that didn't point to degree or order, and it seems like it confused a lot of people.

12. Every finite group with this property is solvable, which is equivalent to the statement that non-cyclic finite simple groups never have this property according to the Feit-Thompson theorem. If a subgroup of the symmetric group on n elements has this property, then it is also a subgroup of the alternating group on n elements. A graph with zero or two vertices with this property has at least one Eulerian circuit. The Fourier series of a periodic functions with this property has only sine terms. If a polynomials has this property, its graph has an end that goes to infinity and one that to minus infinity. Groups with this property cannot have a subgroup with order 2. For 10 points, what is this property of either a vertex or monomial that has its degree indivisible by 2?
ANSWER: odd order [or odd degree]

I wrote the gamma tossup from things I had done in class. I can maybe agree that the Breit-Wigner and gyromagnetic ratio clues weren't that satisfying-- you could probably have seen those formulas a few times and not remembered the letter.

14. Casimir introduced a trick that reduces the problem of finding S-Matrix elements to the problem of calculating traces of operators symbolized by this letter. This letter appears in the Breit-Wigner distribution where it stands for the resonance width, and the frequency of Larmor precession is equal to the magnetic field over two pi times a quantity symbolized by this letter. A quantity symbolized by this letter equals 5/3 for monatomic ideal gases and is the ratio of the constant pressure heat capacity to the constant volume heat capacity. Another quantity symbolized by this letter is one over the root of one minus velocity squared and appears in the formulas for length contraction and time dilation. For 10 points, name this letter which labels the most energetic rays in the electromagnetic spectrum.
ANSWER: gamma

Cody wrote normal form and Victor wrote Q. Here are those questions.

A example of these things used in automated theorem proving is defined as being a conjunction of one or more clauses, where a clause is a disjunction of one or more literals; that example of these things is prominently used when considering the Boolean satisfiability problem. Half of this two-word term is sometimes dropped when used to describe the notation that includes angle brackets around category names, a vertical bar for or, and most importantly a double colon-equals sign. That notation is used to formally define the syntax of programming languages or other context-free grammars, and is named for Backus and sometimes Naur. The three and Boyce-Codd examples of these things are commonly used in relational database design. For 10 points, name these standard methods of representing something.
ANSWER: normal form

9. For a moving gas, one half density times velocity squared equals this letter, representing the dynamic pressure. For a nuclear reaction, the sum of the product masses minus the sum of the reactants equals a value titled for this letter. It is not V, but this letter represents volumetric flow rate. Glutamine is represented by this letter. In a calorimetry, heat capacity equals this letter over the change in temperature. Internal energy plus work equals a quantity symbolized by this letter according to the first law of thermodynamics. For 10 points, identify the letter representing heat transfer and the reaction quotient in kinetics.
ANSWER: Q [or q]

The Wiener process thing was my fault too-- I'll fix it for other mirrors.

[Ras superfamily]

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

The Virginia packet did not contain a tossup on normal forms when it was submitted, so it'd seem that the question was added by one of the editors. (The tossup went dead in the St. Lukes-Penn A game, and both teams started raging when the answerline was read.)

What actually happened is that I negged it in the first sentence because why would the answer line ever be "normal form" and then complained loudly when it was appropriate to do so, namely at the end of the question. Obviously, I agree with Ashvin's advice.

[Ike]

I disagree with Ashvin on pretty much anything related to CS in this thread.

I actually did not get to see any of Cody's questions until just now, though I was a bit surprised: I thought all of them were pretty inspired and well executed. I thought the vector tossup was a good way of asking about machine learning, computer graphics and the STL, and that the normal form tossup was a great way of asking about compilers and database design. If you have any specific criticism as to why these tossups sucked, I would like to hear them.

Also, I really don't appreciate you saying "you should flog anyone who submits a normal form tossup" because you're basically scaring anyone away who is reading your post from writing a tossup on something that can be done well. As much as I would like for Cody to become a flagellant, in this case I give Cody a solid A- or B+ for his normal form tossup - the only thing I would have changed would have been to talk about the Kuroda or Greibach normal forms in the lead-in.

[Cody]

― As I have probably posted about before, tossups on "WORD" are generally not a great idea when you're going for a common link - since the various clues are in a "WORD" tossup are unrelated to one another, it is difficult to build a mental model of what the question is talking about. These questions lack thematic unity, which, I am led to believe, is a feature generally desirable in common link questions. "WORD" tossups seem to show up most frequently in math, but this time, it was the CS tossup on "vector" that was a real bear.

You're wrong. "Thematic unity" is nice but by no means necessary. Many, many common link questions outside of science have little thematic unity except for, you know, the answerline. There's nothing wrong with this, and it's a perfectly fine way to write a common link as long as the clues support it. Your "mental model of what the question is talking about" is given by the referent to the answerline; not every question needs to narrow down the answer space in such a way as you are claiming. There's something to be said for testing a breadth of knowledge in different topics, which is what the vector question did.

― The tossup on "odd" claimed something along the lines of "all groups with this property are solvable according to the Feit-Thompson theorem". In retrospect, I obviously see what the question was going for, but the wording is confusing: there is no such thing as an "odd group" (Max Schindler claims that this is valid nomenclature for a "group of odd order", and I guess he's technically right, since I found a paper that calls it that, but I'm not finding very many references to it - the usage is, at the least, rare enough to be confusing).

You are correct that it would be better if the word "order" is stated. However, the phrasing does not interfere with buzzing on the clue if you know the Feit-Thompson theorem -- it is not really that confusing.

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

[Ras superfamily]

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

The first sentence reads: "A [sic] example of these things used in automated theorem proving is defined as being a conjunction of one or more clauses, where a clause is a disjunction of one or more literals; that example of these things is prominently used when considering the Boolean satisfiability problem."

Any logic formula in CNF is still a logic formula. I should be able to answer logic formula, Boolean expression, or possibly other things.

[Cody]

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

The first sentence reads: "A [sic] example of these things used in automated theorem proving is defined as being a conjunction of one or more clauses, where a clause is a disjunction of one or more literals; that example of these things is prominently used when considering the Boolean satisfiability problem."

Any logic formula in CNF is still a logic formula. I should be able to answer logic formula, Boolean expression, or possibly other things.

This is a fair criticism of the question. (It, of course, is not inherent to the answerline though -- one could rephrase the first sentence and have a perfectly fine tossup on normal forms)

[Excelsior (smack)]

I wrote the gamma tossup from things I had done in class. I can maybe agree that the Breit-Wigner and gyromagnetic ratio clues weren't that satisfying-- you could probably have seen those formulas a few times and not remembered the letter.

Right, that's what happened to me with the Larmor precession clue. On the other hand, the Lorentz factor and gamma rays clues are perfect examples of clues that work with this type of question, since "gamma" is virtually universal for the Lorentz factor, and gamma rays are literally called gamma rays.
(Unrelatedly, I'm curious what the lead-in is referring to. It seems interesting.)

You are correct that it would be better if the word "order" is stated. However, the phrasing does not interfere with buzzing on the clue if you know the Feit-Thompson theorem -- it is not really that confusing.

I agree that if you buzz on the first clue, there is nothing confusing. For whatever reason, I forgot what the Feit-Thompson theorem was and let the first clue pass me by, at which point clues about things that were not odd order/degree started showing up. By then, I had remembered that the Feit-Thompson theorem was a theorem about groups whose order had some particular property, but this was incompatible with the clues I had received later, e.g. about odd functions. I don't think this makes the question terrible or anything, but it is something that should be avoided to whatever extent possible.

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

In addition to what Saajid said, BNF is not even really a "normal form". Now that I see the question, the later clue reading "the three [...] [example] of these things" is confusing. With the answerline in front of me, I can see that this is talking about 3NF, but that is pronounced "third normal form", not "three normal form". In any case, relational data modelling is, I'm pretty sure, harder than BNF (at least for a quizbowl audience). I suppose it is possible to write this question well by focusing exclusively on normal forms in the context of relational data or in the context of formal language theory (or possibly both combined), so I guess I should withdraw my demand for lashings; however, this question did not focus on either of those things, and was just not really good. (Also, it's really hard for a regular-difficulty tournament, but that is neither here nor there.)

You're wrong. "Thematic unity" is nice but by no means necessary. Many, many common link questions outside of science have little thematic unity except for, you know, the answerline. There's nothing wrong with this, and it's a perfectly fine way to write a common link as long as the clues support it. Your "mental model of what the question is talking about" is given by the referent to the answerline; not every question needs to narrow down the answer space in such a way as you are claiming. There's something to be said for testing a breadth of knowledge in different topics, which is what the vector question did.

This is true; however, I think that these questions work slightly more poorly in science than in other categories. Here's why: if there is a tossup on, say, "dogs in various mythologies" or "sportsmen in American literature", the player will know from the outset ("In Incan myth, one of these animals..." / "In one work by So-and-So, a character of this profession...") that there is no thematic unity to be expected from the question, and that it will draw its clues piecemeal from unconnected sources. This is usually not obvious in science questions, and so it leads to an attempt at constructing an incorrect mental model of the answer, which impedes later buzzing. I would have to see the "vector" tossup to identify precisely how this played out there.

A graph with zero or two vertices with this property has at least one Eulerian circuit.

A-ha! I now see why this was particularly confusing to me - I parsed this as "(A graph with zero or two vertices) (with this property)" rather than as "A graph with zero or two (vertices with this property)". My parse is obviously the stupider of the two options, but it might be worth changing this to read "A graph that has zero or two vertices with this property..." so that it is fully unambiguous.

[Cody]

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

In addition to what Saajid said, BNF is not even really a "normal form". Now that I see the question, the later clue reading "the three [...] [example] of these things" is confusing. With the answerline in front of me, I can see that this is talking about 3NF, but that is pronounced "third normal form", not "three normal form". In any case, relational data modelling is, I'm pretty sure, harder than BNF (at least for a quizbowl audience). I suppose it is possible to write this question well by focusing exclusively on normal forms in the context of relational data or in the context of formal language theory (or possibly both combined), so I guess I should withdraw my demand for lashings; however, this question did not focus on either of those things, and was just not really good. (Also, it's really hard for a regular-difficulty tournament, but that is neither here nor there.)

Yes, BNF is not technically a normal form, but it IS called that -- and hence why half of the name is sometimes dropped and the acronym adjusted to put Naur in its place. The clue is still accurate and gettable for anyone who's learned about BNF. You're right about "three" vs. "third" -- wasn't thinking when I wrote that clue. I don't think the relational normal forms are harder than BNF -- not in the real world or quizbowl.

All the normal forms I choose for the question were just fine -- they are very much in the top 5 most famous CS things to which the term "normal form" is applied. A good tossup on "normal form" need not rely just on clues from relational database or formal language or both (especially given how important CNF is!). Focusing on other NFs would've made this question much harder than it was.

This is true; however, I think that these questions work slightly more poorly in science than in other categories. Here's why: if there is a tossup on, say, "dogs in various mythologies" or "sportsmen in American literature", the player will know from the outset ("In Incan myth, one of these animals..." / "In one work by So-and-So, a character of this profession...") that there is no thematic unity to be expected from the question, and that it will draw its clues piecemeal from unconnected sources. This is usually not obvious in science questions, and so it leads to an attempt at constructing an incorrect mental model of the answer, which impedes later buzzing.

I really do not understand this; you're going to have to explain in more detail why it "impedes buzzing".

The starting address of an interrupt service routine is the interrupt type of these things and is stored in a namesake table. A type of quantization named for these things divides an input space into a Voronoi diagram and is called "learning" when supervised. These objects name a class of learning algorithms that apply to non-linearly-separable patterns by transforming data into a new feature space using a kernel function. That class is so-named for data points that lie closest to the decision surface, which are the “support” type of these objects. Raster graphics are contrasted with graphics denoted by this term, which are inherently far more scalable. This term is often applied to 1-dimensional arrays. For 10 points, name this term that, in computer graphics, denotes a quantity with length and direction.
ANSWER: _vector_s [or support _vector_s]

(I guess I don't know if this was the question as it appeared in a packet -- it's the version I sent Brian, so it may have been edited or proofread between me and the packet.)

[Lagotto Romagnolo]

I seem to remember the Dirichlet question saying that his namesake boundary conditions give the value of a function's derivatives at boundaries, when in fact it's Neumann conditions that give values of the derivative and Dirichlet conditions that give values of the original function. Or maybe I'm mis-remembering - could someone post the text?

[Victor Prieto]

This mathematician’s principle states that solutions to Poisson’s equation can be obtained by minimizing his namesake energy, a functional measure of variance. The sum over integers k from minus n to n of e to the power ikx is called his kernel and the indicator function on the rationals is sometimes named for him. He proved that for two positive coprime integers a and b, there are infinitely many primes congruent to a mod b. Specifying the values of a differential equation’s solutions at the boundary is the mathematician’s boundary conditions; specifying the derivatives are Von Neumann’s. For 10 points, name this German mathematician who formulated the pigeonhole principle.

ANSWER: Johann Peter Gustav Lejeune Dirichlet

[Lagotto Romagnolo]

Ah, I guess I confused "differential equations" with values of the differentiated function. Sorry about that - this question is fine as is.

[Brian McPeak]

I'd also like to hear what people thought about having so much math in the set. I enjoyed having a more than normal because I got to ask about some things I normally wouldn't have. This amount (9/8 out of 15/15) felt like too much though.

Another reason that it was nice to have more math is that chemistry submissions are usually terrible-- Victor can comment on this more, but while there only 2/3 as much chem as there was bio in the set, he had write way more chem from scratch. I didn't like having to sacrifice astronomy, CS, and earth science though-- it felt like there weren't enough of these to me.

[Cody]

I didn't like having to sacrifice astronomy, CS, and earth science though-- it felt like there weren't enough of these to me.

Did you? If you have 10/10 Chem and 10/10 Math in a 15 packet set, then you really only have 5/5 Math in the "Misc. Science" distribution, which is not that out of normal and leaves 10/10 for Misc. Science. I don't know what the actual breakdown here was, but normally you aren't overriding the Misc. Science by many questions.

[Ike]

― If you are an editor, and you receive a common-link tossup on "normal forms", you should have the author of the tossup flogged. You don't necessarily need to flog the person who wrote this tossup for DEES, but, you know, if it ever happens again.

Do explicate. I accept that this tossup is definitely on the harder end (possibly the hardest end), but all of the clues are both specific and very important in CS, so I very much disagree.

The first sentence reads: "A [sic] example of these things used in automated theorem proving is defined as being a conjunction of one or more clauses, where a clause is a disjunction of one or more literals; that example of these things is prominently used when considering the Boolean satisfiability problem."

Any logic formula in CNF is still a logic formula. I should be able to answer logic formula, Boolean expression, or possibly other things.

For what it's worth, I would not accept your answer as correct. This would be the equivalent of a question that reads 'A 19th century author from this place wrote about "Jean des Essientes in Against the Grain"' buzzing in and saying "the planet Earth" and demanding that your answer to be taken, when the correct answer is France. Yes, technically, your answer is right, but by answering "boolean expression" you completely trivialize the question, especially since the question is talking about conjunctions, disjunctions and literals.

You might convince me to throw the question out, but even then I can't say you have a compelling enough answer to make me do so. In the context of automated theorem proving, an answer of Boolean expressions just doesn't make any sense.

[Ras superfamily]

Any logic formula in CNF is still a logic formula. I should be able to answer logic formula, Boolean expression, or possibly other things.

For what it's worth, I would not accept your answer as correct. This would be the equivalent of a question that reads 'A 19th century author from this place wrote about "Jean des Essientes in Against the Grain"' buzzing in and saying "the planet Earth" and demanding that your answer to be taken, when the correct answer is France.

Are you trying to defend badly written questions? Your hypothetical lit question is bad because the pronoun "place" is too vague (and I will add that your example is rather unreasonable since saying "the planet Earth" is technically correct for all questions asking for the origin of any real person whereas the correctness of saying "Boolean expression" depends on having understood things in the first clue of the normal forms tossup, making it significantly less trivializing unless the fact that Boolean expressions use things called "literals" is as well known as the fact that all humans are from the planet Earth). There are multiple correct answers up to a significantly deep point in the question; are you sure the "correct answer is France" when Paris, Europe, the eastern hemisphere, and Earth are all true responses (though this could be solved easily using prompts as appropriate)? Similarly, the tossup on normal forms was bad because the first sentence, which took up more than 2.5 lines of an 8 line tossup, failed to specifically point to the answer and the answer line had no prompts.

Perhaps that answer line of just "_normal form_s" could be done well, but it was not done well in this instance and I would hesitate to say it's difficulty appropriate and would guess that the conversion stats were abysmal. If the editors insist on keeping it anyway, either the first 2.5 lines of the question or the answer line needs to be changed.

[Cody]

Perhaps that answer line of just "_normal form_s" could be done well, but it was not done well in this instance and I would hesitate to say it's difficulty appropriate and would guess that the conversion stats were abysmal. If the editors insist on keeping it anyway, either the first 2.5 lines of the question or the answer line needs to be changed.

Normally, one would fix the question, yes. But ... you, yourself, spoiled it on IRC, so you should know that I'm writing a replacement!

[Ras superfamily]

Yeah, that's definitely my fault, though I hope that the outcome would have been the same in either case (the question being rewritten)

[Adventure Temple Trail]

spoiled it on IRC

Don't do this, people! I presume this particular instance was a careless accident rather than a premeditated act of quizbowl terrorism, but still: take care not to do this! If you just have to get your vent out and these forums don't serve the function well, ask a set editor to start up an IRC channel that only people who played the set can enter or something.

[Ike]

Any logic formula in CNF is still a logic formula. I should be able to answer logic formula, Boolean expression, or possibly other things.

For what it's worth, I would not accept your answer as correct. This would be the equivalent of a question that reads 'A 19th century author from this place wrote about "Jean des Essientes in Against the Grain"' buzzing in and saying "the planet Earth" and demanding that your answer to be taken, when the correct answer is France.

Are you trying to defend badly written questions? Your hypothetical lit question is bad because the pronoun "place" is too vague (and I will add that your example is rather unreasonable since saying "the planet Earth" is technically correct for all questions asking for the origin of any real person whereas the correctness of saying "Boolean expression" depends on having understood things in the first clue of the normal forms tossup, making it significantly less trivializing unless the fact that Boolean expressions use things called "literals" is as well known as the fact that all humans are from the planet Earth). There are multiple correct answers up to a significantly deep point in the question; are you sure the "correct answer is France" when Paris, Europe, the eastern hemisphere, and Earth are all true responses (though this could be solved easily using prompts as appropriate)? Similarly, the tossup on normal forms was bad because the first sentence, which took up more than 2.5 lines of an 8 line tossup, failed to specifically point to the answer and the answer line had no prompts.

Perhaps that answer line of just "_normal form_s" could be done well, but it was not done well in this instance and I would hesitate to say it's difficulty appropriate and would guess that the conversion stats were abysmal. If the editors insist on keeping it anyway, either the first 2.5 lines of the question or the answer line needs to be changed.

No, I am not trying to defend bad questions, but just because a question is bad doesn't mean you get the right to neg it with an answer like "the planet Earth" protest and collect points, or you get to buzz in, say something that makes sense trivially and get the question thrown out. I'm not saying I would have chosen it as a lead-in, but if you think about the context of automated theorem proving on Bool-SAT, you have to know that you must convert the Boolean expression to some kind of "normalized form" in order to do any type of automated work on it. It's not that TERRIBLE of a lead-in.

[Ras superfamily]

I will again point out that the comparison is kind of unfair because "logic formula" or "Boolean expression" isn't unreasonable to expect as an answer to a tossup whereas "Earth" would be totally unreasonable as an answer to a question about authors' origins. You also seem to have glossed over my point that it's not trivial, despite your claims, to get from the first clue (say, before the semicolon) to "Boolean expression" without knowing things, unlike the lit example.

just because a question is bad doesn't mean you get the right to neg it with an answer like "the planet Earth" protest and collect points, or you get to buzz in, say something that makes sense trivially and get the question thrown out.

I thought that's exactly how the rules worked, but could someone better informed than me explain what would happen in this situation?

[Ike]

I will again point out that the comparison is kind of unfair because "logic formula" or "Boolean expression" isn't unreasonable to expect as an answer to a tossup whereas "Earth" would be totally unreasonable as an answer to a question about authors' origins.

Rob Carson once wrote a tossup on earth using Lucian of Samosata clues and others in literature. Yes, it doesn't happen all the time, and yes, there were circumstances that allowed for it - since Lucian is whacky, and yes It wasn't an "origin" tossup per se, but it certainly isn't a "totally unreasonable answer."

You also seem to have glossed over my point that it's not trivial, despite your claims, to get from the first clue (say, before the semicolon) to "Boolean expression" without knowing things, unlike the lit example.

Again, I'm not really defending the clue or the question, although I still insist isn't that either isn't that terrible. I'm just saying I wouldn't take your answer. The problem with answering "Boolean expressions" is that it takes a trivial amount of knowledge to generate that answer--why on earth would you ever begin a tossup on Boolean expressions with the word "literals?" I will offer a more apt analogy in the realm of other science: consider a tossup that begins "An example of things with this property used in Galois theory take the form of "x^2 + y^2 - 10"." The answer to this tossup would be "symmetric" polynomials, but someone might buzz in here and say "quadratic" or "polynomial expression" because that polynomial is quadratic. However, that answer should not be taken because it ignores the context of Galois theory and is so trivial and easy to understand relative to what level of knowledge the question is asking for (if the answer is really quadratic polynomial (boolean expression) why mention Galois theory (automated theorem proving)?

As I pointed out in some other thread, you can't make every clue uniquely identifying, especially in science. At some point you just have to accept that you've used enough precision to specify the answer. I'm not sure if Cody used enough precision to point people towards the answer of CNF, (when I read through it on Saturday, I was 50% sure of the answer on the first clue, but probably wouldn't buzz there) but he certainly did use enough precision to rule out answers of "Boolean expressions."

Just because a question is bad doesn't mean you get the right to neg it with an answer like "the planet Earth" protest and collect points, or you get to buzz in, say something that makes sense trivially and get the question thrown out.

I thought that's exactly how the rules worked, but could someone better informed than me explain what would happen in this situation?

I don't know why you insist that I'm not one of those knowledgeable people as I've been around long enough to resolve a lot of protests, but I will point you toward Matt Weiner's example about Aius Locutius. This answer really does feel tendentious to me, according to the reasons why I outlined above.

[Auks Ran Ova]

Rob Carson once wrote a tossup on earth using Lucian of Samosata clues and others in literature. Yes, it doesn't happen all the time, and yes, there were circumstances that allowed for it - since Lucian is whacky, and yes It wasn't an "origin" tossup per se, but it certainly isn't a "totally unreasonable answer."

Yeah, but as you can see below I also used the pronoun phrase "this planet", not only in the first sentence but in every sentence. It's a lot more specific than "this place" and circumvents any of the issues under discussion.

A culture on this planet uses a calendar that includes the month of Minous, during which trees bear fruit twice. An island on this planet inhabited by a race of hooved women is called Cabbalussa. Another island on this planet, Taprobane, is home to a snake with a head at each end of its body. A culture from this planet dresses in purple spiderwebs to cover their Observer-like lack of bodies and is ruled by a king named (**) Radamantus. A polity on this planet communicates with its neighbors using messengers called Perpherees. Those inhabitants of this planet leap from a rock into the sea when they become tired of life and experience only one sunrise and one sunset per year. An expedition launched from this planet becomes entangled in a war over colonization of the (*) Morning Star between the armies of Hyperion and Phaethon. For 10 points, name this planet that is home to the Island of the Blessed and Hyperborea, at least according to Lucian of Samosata and Herodotus, two authors from it.
ANSWER: Earth [or Gaia I guess?] <classical lit>

[Ras superfamily]

I will point you toward Matt Weiner's example about Aius Locutius. This answer really does feel tendentious to me, according to the reasons why I outlined above.

From the same post:

It's on the player to be a reasonable human being (not guess after hearing "This ruler") but not to police every question for quality or difficulty-appropriateness (make judgments on the fly as to whether the 'ladder' clue would or would not be the leadin for Wittgenstein at ACF Nationals).

Are players on the hook to make a difficulty judgment on this question when they hear the leadin? That's what you are suggesting ("that answer should not be taken because it ignores the context of Galois theory and is so trivial and easy to understand relative to what level of knowledge the question is asking for (if the answer is really quadratic polynomial (boolean expression) why mention Galois theory (automated theorem proving)?").

[Excelsior (smack)]

I don't think the relational normal forms are harder than BNF -- not in the real world or quizbowl.

I guess this is a matter on which reasonable minds can differ (as can we) but in a typical sequence of CS courses, I imagine one is unlikely to encounter relational normal forms until a course on database theory (which is typically a late-stage elective or somesuch), whereas CFGs are the bread and butter of any course on the theory of computation, which is usually taught earlier, and if you're going to teach CFGs, one would surely also mention the standard way of writing them down. In the real world, I am quite certain that - unless you are polling a room full of DB admins - you will find more people familiar with BNF than with relational normal forms.

I really do not understand this; you're going to have to explain in more detail why it "impedes buzzing".

Okay, I was probably not being clear earlier. Here is my problem with the "vector" tossup. Allow me to illustrate my thought process as the question plays out.

>> The starting address of an interrupt service routine is the interrupt type of these things and is stored in a namesake table.
Hm, okay, this is a thing that is an address and can be stored in tables.
>> A type of quantization named for these things divides an input space into a Voronoi diagram and is called "learning" when supervised.
Alright, I want to fill in the blank in "learning ______ quantization", and the blank is something that is an address and can be stored in tables.
>> These objects name a class of learning algorithms that apply to non-linearly-separable patterns by transforming data into a new feature space using a kernel function.
Wait, now I'm confused. The answer is apparently a thing that names a class of learning algorithms - but what do machine-level addresses have to do with that? Learning algorithms operate at a much higher level than that! Let me listen to the next clue to figure out what is going on.
>> That class is so-named for data points that lie closest to the decision surface...
So the answerline is simultaneously a memory address and also a data point? I'm completely lost.
>> ...which are the “support” type of these objects
Oh so this is where I buzz and say "vector".

This is the problem: at different times, the pronouns refer to completely different things that happen to share the same name. This is like writing a tossup on "John Adams" using clues about the president and the composer - but rather than saying "one person of this name", you just say "this person". (This is not really an issue of "thematic unity", and I apologize for characterizing it as such.)

I'd also like to hear what people thought about having so much math in the set. I enjoyed having a more than normal because I got to ask about some things I normally wouldn't have. This amount (9/8 out of 15/15) felt like too much though.

I thought it worked great and didn't end up delving into uninteresting minutiae, as sometimes happens with large math distributions. I would not be opposed to seeing other tournaments try this variation on the science distribution, even if that meant eating into something other than chemistry.

"An example of things with this property used in Galois theory take the form of "x^2 + y^2 - 10"

This is a terrible leadin - It fails to specify in _any_ way what the actual property being sought is! Just because you say "in Galois theory" does not make "x^2 + y^2 - 10" not quadratic. You can't just sort of vaguely gesture in the direction of the answerline and expect players to figure out what you're talking about.

[Ike]

Are players on the hook to make a difficulty judgment on this question when they hear the leadin? That's what you are suggesting ("that answer should not be taken because it ignores the context of Galois theory and is so trivial and easy to understand relative to what level of knowledge the question is asking for (if the answer is really quadratic polynomial (boolean expression) why mention Galois theory (automated theorem proving)?").

This is something at the protest resolver's discretion. I, too, would rule the way Matt ruled on the Wittgenstein ladder clue. But if you can't see why I would accept an answer of Wittgenstein's Tractatus where the clue in question requires a pretty deep reading of Wittgenstein, even though it's a famous quote, and why I would not accept an answer of Boolean expressions given "literals" - which are the very thing that comprise and define a Boolean expression, I'm more than happy to discuss it with you through email -- unless other interested parties are really interested in hearing more about this. I guess I should say this: I'm handling the other science at ACF Nationals this year, so you can expect me to rule* based on what I said in this thread. If you have any doubts / are just confused as to my logic on why I can't take your answer as correct, please feel free to post here. You may convince me that the question has "no right answer," but I don't see how I can do that, given that Cody actually defined conjunctive normal form precisely and accurately using the textbook definition! The problem that I think you are having is that you don't know how to arrive at the answer of "normal form;" I sympathize with you as a player. However, that's not grounds for a protest.

*By rule, I mean provide whoever the powers that be are with my reasoning and recommendation.

[dxdtdemon]

I'm pretty sure that there was a bonus part that was looking for the official name for mu_0 and gave the official name for epsilon_0 instead.

[Mike Bentley]

I'd also like to hear what people thought about having so much math in the set. I enjoyed having a more than normal because I got to ask about some things I normally wouldn't have. This amount (9/8 out of 15/15) felt like too much though.

Another reason that it was nice to have more math is that chemistry submissions are usually terrible-- Victor can comment on this more, but while there only 2/3 as much chem as there was bio in the set, he had write way more chem from scratch. I didn't like having to sacrifice astronomy, CS, and earth science though-- it felt like there weren't enough of these to me.

I was a bit disappointed to hear 1 CS tossup and I think 2 CS bonuses in the 13 packets we played.

[Gautam]

That tossup on Meselson-Stahl experiment was not very good. There aren't too many things it could have been after the clue on centrifugation + CsCl.

The 9/8 math ended up making the "other sci" category substantially harder. I speak from a perspective of one who doesn't know much of the math that came up in this tournament - usually, even if you know very little math, you still have a shot at converting 60-70% of the Other sci questions which don't draw much from math concepts. But the situation was reversed here. It might be good to advertise this, I suppose.

Just my 2 cents.

[Victor Prieto]

That tossup on Meselson-Stahl experiment was not very good. There aren't too many things it could have been after the clue on centrifugation + CsCl.

I was going to swap that tossup with the lysosome bonus in the same packet, because I feel the latter would lend itself better to a tossup than Meselson-Stahl, but I simply didn't have enough time to make the switch.