On Tossup Writing Considered as one of the Fine Arts

Old college threads.
Locked
User avatar
Galadedrid Damodred
Wakka
Posts: 157
Joined: Sat May 25, 2013 4:58 pm

On Tossup Writing Considered as one of the Fine Arts

Post by Galadedrid Damodred »

(a sample of rejected titles: The Tossup with a Thousand Faces; The Tossup in the Age of Mechanical Reproduction; How to Explain Tossups to a Dead Hare)

This post is my attempt to reproduce my thought process during the writing of a single tossup for 2017 ACF Nationals. I did not take notes while writing the tossup in question 8ish months ago, so it will necessarily be an imperfect reconstruction. As far as I know, no one has ever gone into this much detail about how a tossup gets written, so I hope people will find it interesting.

I’ll begin with the final text of the question. This is the physics tossup from the Editors 5 packet:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. The Conwell-Weisskopf and Brooks-Herring models describe a phenomenon named for these entities that, at low temperatures, is the dominant term in Matthiessen’s formula for charge carrier mobility. Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. For 10 points, name these crystal defects that result from the substitution of foreign atoms.
ANSWER: impurities [or dopants; or doping agents; accept ionized impurity scattering; accept substitutional defects before “substitution” and prompt thereafter; prompt on crystal defects or lattice defects]
The genesis of this tossup lay in my desire to include more condensed matter physics, and solid-state physics in particular, than I was used to seeing in the typical high-level quizbowl tournament. I felt that these subfields were generally under-represented relative to the amount of research they attract in contemporary academia. Also, as an electrical engineering major, I took a lot of classes on semiconductor physics, so I wanted to leverage that knowledge as much as possible to help make the set the best it could be (operating under the assumption that I could write better questions in a shorter amount of time on topics where I had “real” knowledge versus self-taught knowledge).

On August 3, 2016, Matt Bollinger sent me an email asking if I could step in as the physics editor for 2017 ACF Nationals. At the time, I thought I had managed to successfully retire from quizbowl for good, so nothing like this offer was on my radar at all – but I won’t change my mind twice, so when I say now that I’m through, I really mean it! Anyway, after getting over the initial shock, I did a lengthy brainstorming session to see if I had enough creativity left in the tank to come up with a sufficient number of tossup ideas (answerlines plus outlines of the clue pyramids, especially the early-middle to middle clues or roughly the second through fourth sentences). As part of the brainstorming process, I went through my personal question archive looking for what topics hadn’t come up over the past 5 years with either sufficient exposure or, more importantly, sufficient depth (meaning that the same set of clues tended to be recycled without a more thorough exploration of the underlying physics). As you might guess at this point, one of my ideas was to write a tossup on dopants in semiconductors, or more broadly speaking, impurities in solids. I found two tossups with this answerline: one from 2011 Minnesota Open and the other from Cody Voight’s electrical engineering side tournament that I played in 2015. I have included these tossups below for reference:
2011 Minnesota Open wrote:18. The number of electrons added to a system by introducing one of these is proportional to the sum of the product of the Fermi wave vector times the phase shift of each angular momentum channel in the Friedel sum rule. They create localized magnetic moments by electrons tunneling out of them into an intermediate state, followed by an electron tunneling back into them; that phenomenon is described by a model in which they create virtual bound states with the surrounding Fermi sea, known as the Anderson model. In a superconductor, as the temperature approaches absolute zero, the conduction electrons begin to strongly couple to these elements, resulting in the Kondo effect. In a superconductor, vortices can form around these elements, which generally serve as nucleation points for a phase transition. They are added to a semiconductor in the process of doping. For 10 points, name these elements, which are foreign substances with a magnetic dipole moment in an otherwise pure material.
ANSWER: magnetic impurities [prompt on semiconductors]
Claude Shannon Memorial Tournament wrote:Though very hazardous, a solution of ethylene diamine and pyrochatechol, or EDP, is useful as an etch that depends on orientation and these things. The y-axis of an Irvin’s curves diagram is for the surface concentration of these things. The van der Pauw method is useful for determining the concentration and type of these things. The redistribution of these things during oxidation depends on their segregation coefficient. The solid solubility limit determines the electrically (*) active concentration of these things. Spin-on glasses are used in low-rent processes for including these things, but the state of the art process is ion implantation. Their addition shifts the Fermi level closer to the conduction or valence band. For 10 points, name these things added to semiconductors to increase the amount of n– or p–type carriers.
ANSWER: dopants [or: doping agents, ionized impurities]
I then made a list of possible clues, ordered from hardest to easiest. It looked something like this:
-leadin related to something I learned in school about dopants
-deep description of Anderson model or Kondo effect (may swap with ionized impurity scattering)
-ionized impurity scattering (probably hard to describe non-transparently; EEs should know this)
-basic definition of Anderson model, with namedrop – flashcarders should be starting to buzz now
-basic definition of Kondo effect (if no namedrop, is it still easier than reflex buzz on Anderson?)
-basic stuff about dopants (n-type/p-type; acceptors and donors)
-giveaway is definition of impurities as lattice defects

By now, I was feeling pretty good about this idea. The MO tossup had a good description of the Anderson model that I could use as a starting point, but that topic wasn’t well-known in quizbowl (though I guess it will be now?), so I thought it would still be fresh even if I just used the basic definition and gave the name in the middle of the question. Most of Cody’s tossup was on very difficult and specialized material, so I didn’t plan to lift from it at all, discounting very easy clues that are inevitably repeated over and over. I did realize from reading these tossups that the hardest part about writing my own tossup would be including hard clues that weren’t transparent. Once it’s clear to players that I’m describing things that are inside solids/semiconductors and interacting with electrons or generally behaving like (quasi-)particles, the answer space narrows considerably.

The last thing to do before starting in earnest was to decide on a noun indicating the answer category. I didn’t want to use the descriptor “these elements” because it would imply that I’m looking for a chemical element or group of elements, and “these things” didn’t sound professional enough, so I decided to use “these entities” prior to the giveaway, at which point I wouldn’t be trying to hide anything, so I could use whatever sounded clearest in context.

I’m now going to switch to present tense and write in a style that’s more stream-of-consciousness, because I think that will be easier to follow as I wade through the gory details of writing the tossup.

I start as I often start, by searching the answerline on Wikipedia. Impurities – not much here, let’s click on zone refining because I vaguely remember what that is and it’s “economically important” – hmm, that’s actually a nice clue because it implies these are things you might want to get rid of but it isn’t too obvious – what if I put that at the beginning of the sentence before FTP when I talk about what dopants are used for? Well, I guess I'm going to write the next-to-last sentence before anything else. As in the rest of this post, keep in mind this is only as accurate as the extent to which I can remember what I did 8ish months ago. For the most part, I’m also ignoring how I was obsessing over my word choice and grammar, except when it mattered for the overall construction of the tossup. Anyway:
Austin Brownlow wrote:Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors.
Now I search “dopants” on Wikipedia. Note: add “doping agents” to the list of acceptable answers – ion implantation is cool, but it’s too obvious of a term (as I suspect is the case for many of my ideas about clues on doping semiconductors) – Wikipedia has a separate article on “doping (semiconductors)” – heh, some of these sections look like they could have come straight from my undergraduate Physical Electronics textbook – I’m starting to think I should focus on the physics side of things and ignore other engineering aspects outside of maybe the lead-in sentence.

Goodbye, Wikipedia. Hello, Google Scholar. Search “Anderson model” – scroll through the first page – wow, “Relation between the Anderson and Kondo Hamiltonians” has ~2000 citations – search “Schrieffer Wolff” on regular Google – Schrieffer-Wolff transformation has a Wikipedia page! Somebody wrote a paper in 2011 extending their approach to many-body systems! There’s a PDF called “important canonical transformations” that looks like it’s from a textbook or review paper! Guess I don’t need to pore through my old lecture notes looking for a lead-in, because this seems like great lead-in material. It’s very well-cited and relates to more important topics that I’m planning to discuss later in the question. It’s never come up before, so I can use the names:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. {…} Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors.
On that Google Scholar search page, the other paper with >1000 citations is “Renormalization-group approach to the Anderson model of dilute magnetic alloys.” Let’s read the abstract – “The calculations were performed using the numerical renormalization group originally developed by Wilson for the Kondo problem” – hmm, I know Kenneth Wilson won the Nobel for RG stuff and that he solved the Kondo problem, but I haven’t seen the term “numerical renormalization group” before – search on regular Google – it has a Wikipedia page! – “The numerical renormalization group (NRG) is a technique devised by Kenneth Wilson to solve certain many-body problems where quantum impurity physics plays a key role.” – looks like it’s a specific RG method that is used for systems with impurities, so it’s uniquely identifying – the Wikipedia article talks about the procedure itself, which looks like it could fit the “deep description of something that’s known in quizbowl at a shallow level” sentence that I want to put right after the lead-in – maybe I can find a review article or something else that’s a reliable academic source – search “numerical renormalization group” on Google Scholar – what do you know, the first result is a review article (“Numerical renormalization group method for quantum impurity systems”), and it’s available in PDF on the arXiv! Need to find a section that gives an overview of how to perform the computation (I make no claim that I read all 55 pages) – how can I turn this into one sentence – here goes:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. {…} Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors.
That’s about as “real” as a physics clue can get. To the hypothetical (and as of last weekend, I hope real?) people who buzz during the first two lines of this tossup: you rock. Actually, if you buzz during the ionized impurity scattering clues, that’s very impressive as well, but I’m getting ahead of myself. Since I just researched the Anderson and Kondo models, I should write the easier clues for those topics while they’re fresh in my mind. First, I’ll paraphrase and condense the clues about the Anderson model from the 2011 Minnesota Open tossup (using other sources to check its accuracy, of course):
Austin Brownlow wrote:These entities create virtual bound states that facilitate electron tunneling in the Anderson model.
Saying that impurities “facilitate” electron tunneling, while still factually correct, is not nearly as transparent as saying that electrons are tunneling into and out of impurities like the MO tossup did. (To whoever edited that question: I’m not slagging off your tossup, just explaining my own approach.)

Now I need to find a way to include numerical RG as a clue. Just saying “the numerical RG method is used on systems containing these entities” is bad because it doesn’t provide any context. I decide it might help if I write a simple definition of the Kondo effect and then see if I can relate the two:
Austin Brownlow wrote:Systems containing these entities exhibit finite minimum resistivity at a temperature above absolute zero in the Kondo effect.
The obvious link is Kenneth Wilson, since he developed numerical RG to explain the Kondo effect. But then I’m going to need to use his name at the beginning of the sentence (to avoid torturing the syntax), and that might lead people to neg with “phase transitions” because that’s what is cited in his Nobel Prize award and what he’s more well-known for in quizbowl. I decide that if you want to buzz at the beginning of a sentence based solely on hearing the name of a physicist who is not just known for one thing (let’s not forget Wilson loops and lattice QCD – the guy got around!), then you have to accept the risk because you haven’t waited for a complete phrase. As it turns out, Seth Teitler made this exact neg in playtesting and immediately admitted that he was just guessing without really understanding what was going on. Sorry for bringing this up in a public forum, Seth!

Furthermore, I decide that if I don’t namedrop the Kondo effect, that clue becomes harder than the Anderson one because the latter makes it fairly obvious what type of “entities” we’re talking about. I now realize that I can mash all of this together into one sentence that, although long, sounds fine when I read it out loud. As a writer, not just in quizbowl but in general, I prefer longer sentences that have a good “flow” to shorter ones that sound “choppy,” as you are no doubt aware by now if you’ve read this far. The tossup now contains all but two of the sentences from the original outline:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. {ionized impurity scattering} Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. {giveaway}
I like to leave giveaways until last because it’s often easier to adjust their lengths than it is for the harder clues. Therefore, I’m going to write the sentence on ionized impurity scattering next. Since this is a topic I learned in a classroom setting rather than for quizbowl, I want to write the clues in such a way that if I were playing, I’d have no problem understanding and I’d buzz with confidence. I start with what I remember having to memorize for a test in that same undergraduate Physical Electronics class I mentioned earlier, which is that ionized impurity scattering dominates charge carrier mobility at…is it high or low temperatures? Phonon scattering is a thermal process, so that should dominate when T is high, and that means IIS dominates when T is low. Yeah, that’s right.

Time to check the textbook (Semiconductor Device Fundamentals by Robert F. Pierret) to make sure I’m right – hooray, I remembered something from a class that was over 4 years ago! – I know the mobility formula is named for a dude whose name is pronounced muh-THEE-sen, but I must have learned that in a later class because it’s not in this textbook – maybe Wikipedia will know – search “charge carrier mobility” – the article is “electron mobility” because Wikipedia apparently hates holes – lots of good stuff in this article – ah, that’s how you spell it – very few hits for that name on aseemsdb, in physics questions anyway, since apparently there are authors named Peter Matthiessen and F.O. Matthiessen, the former of whom was also a CIA agent! – since there’s some space left in the tossup, let’s do a Google search for “ionized impurity scattering” to see if there’s anything else I can use – the first hit is Wikipedia, which has a bare-bones article – the second hit seems to be about modeling software, which is too specialized – the third and fourth hits are papers that I can’t access, but one of them refers to the Brooks-Herring formula, as do 3 other hits on the first page, so that’s clearly a thing – the fifth hit is a researcher’s personal website, and it goes into depth about Conwell-Weisskopf and Brooks-Herring being the two big models of how to describe ionized impurity scattering – Google Scholar search “Conwell Weisskopf” since I’ve already seen Brooks-Herring several places – Conwell and Weisskopf’s 1950 paper has ~900 citations, and a bunch of hits on the first page mention the two models together, so they’re probably worthwhile to namedrop – which one to put first? Conwell-Weisskopf has way fewer hits on Google Scholar, so it goes first – I should refer to ionized impurity scattering as a “phenomenon” because to mention the word “scattering” makes it too easy this early in the question – instead of something vague like “a phenomenon involving these entities,” I should use “a phenomenon named for these entities” so people in the know who buzz here don’t second-guess what they’re supposed to say as the answer – I’m going to call it “Matthiessen’s formula” even though “Matthiessen’s rule” is more commonly used, because I’m talking about relative weights of quantities in an equation and “rule” sounds more conceptual, so it’s cleaner to state explicitly that it’s a formula – now I can type it up:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. The Conwell-Weisskopf and Brooks-Herring models describe a phenomenon named for these entities that provides the dominant low-temperature contribution to Matthiessen’s formula for charge carrier mobility. Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. {giveaway}
For the giveaway, I’ll just use the most straightforward and concise definition that I can come up with, since I’m supposed to stay below 8 lines and there’s no real gradation of clues at this point:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. The Conwell-Weisskopf and Brooks-Herring models describe a phenomenon named for these entities that provides the dominant low-temperature contribution to Matthiessen’s formula for charge carrier mobility. Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. For 10 points, name these crystal defects that result from the substitution of foreign atoms.
I’ll put impurities as the main answerline and dopants/doping agents as the alternates, as well as ionized impurity scattering because of the phrase “A phenomenon named for these entities.” I also need to include a prompt on “defects,” and I can think of 2 ways to say that answer. Finally, since impurities are substitutional defects, I need to accept that, but only before I say it in the giveaway:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. The Conwell-Weisskopf and Brooks-Herring models describe a phenomenon named for these entities that provides the dominant low-temperature contribution to Matthiessen’s formula for charge carrier mobility. Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. For 10 points, name these crystal defects that result from the substitution of foreign atoms.
ANSWER: impurities [or dopants; or doping agents; accept ionized impurity scattering; accept substitutional defects before “substitution” and prompt thereafter; prompt on crystal defects or lattice defects]
This is what I read during playtesting in early September, unless I’m misremembering how I initially worded the sentence on ionized impurity scattering, specifically the phrase “provides the dominant low-temperature contribution.” Based either on feedback from Billy and Seth or my own second thoughts after I read the whole tossup in a game situation (I don’t remember which it was), I changed that phrase to make it flow better and emphasize the part about being at low temperature:
Austin Brownlow wrote:Two different models of these entities are related by the Schrieffer-Wolff transformation. A procedure for computing these entities’ thermodynamic properties consists of iteratively adding degrees of freedom along a logarithmically discretized, semi-infinite chain. The Conwell-Weisskopf and Brooks-Herring models describe a phenomenon named for these entities that, at low temperatures, is the dominant term in Matthiessen’s formula for charge carrier mobility. Kenneth Wilson developed the numerical renormalization group method to explain the existence of finite minimum resistivity in systems containing these entities, which create virtual bound states that facilitate electron tunneling in the Anderson model. Zone refining decreases the concentration of these entities in semiconductors, where they act as acceptors or donors. For 10 points, name these crystal defects that result from the substitution of foreign atoms.
ANSWER: impurities [or dopants; or doping agents; accept ionized impurity scattering; accept substitutional defects before “substitution” and prompt thereafter; prompt on crystal defects or lattice defects]
This version of the tossup is what went into the final set. It didn’t change at all between September 9, 2016 and April 23, 2017. I don’t think it’s the best tossup I wrote for this set – the first half to two-thirds is extremely hard, and there are quite a few other physics tossups that have a smoother difficulty gradient along with more buzzable early clues for people who aren’t experts. However, I think it serves as a good illustration of what goes into writing an ACF Nationals tossup.
Last edited by Galadedrid Damodred on Mon Aug 30, 2021 11:15 pm, edited 1 time in total.
Austin Brownlow
Louisville '14, Stanford '16
User avatar
Cheynem
Sin
Posts: 7220
Joined: Tue May 11, 2004 11:19 am
Location: Grand Rapids, Michigan

Re: On Tossup Writing Considered as one of the Fine Arts

Post by Cheynem »

I don't know anything about physics, so the specifics in this post sometimes eluded me, but it was an interesting way to look at how tossups are conceived.
Mike Cheyne
Formerly U of Minnesota

"You killed HSAPQ"--Matt Bollinger
Locked