I started this week's practice with a conundrum. I accomplished everything I wanted to ahead of the practice. I put together two study guides, and wrote a couple questions on Heroes in Greek mythology, so I could salt them into the practice packets. But even though I felt I had prepared for the practice very well, I had a gut feeling something was off, a trepidation ensued. And I stumbled into the problem during a tossup on Carnival of the Animals, when I told them that it was a one-hit wonder for Camille Saint-Saëns, and they had no idea what I meant by that. I ended up needing a long explanation of my shorthand reference, and I knew what my gut was telling me.
What you'll need to teach this lesson to high school kids today: "A one-hit wonder in quiz bowl is someone's work that is the only thing the creator(s) are commonly known for today. It comes from the one-hit wonder idea in Billboard charts: "an act that has won a position on [the] national, pop, Top 40 record chart just once” is the common description They may also refer to it as 'onlies', "the only work by X you need to know for quiz bowl," shortened to "the only work by X." Onlies are slightly broader, can refer to more categories than works, and are more subject to change; "the only North Korean" was a joking way to refer to Kim Il-Sung at the time, and then Kim Jong-Il replaced him. But it lays a perfectly plausible way to place a last clue which isn't uniquely identifying in fact, but is uniquely identifying in practice. In music a one-hit wonder made the charts, and so it's part of the historical record, and is the thing that would go in the obituary first, and so it would go in the tossup last."
What does a one-hit wonder even mean today?
I mused on this on the way home because I felt like I had had something knocked out from under me. It's true that a lot of the things that formed weekly writing prompts for question writers, like the New York Times Best Seller List, the weekly box office, and the Billboard Top 40, are not being pushed in the same way to the consumer. (I even had to explain that Billboard charts don’t have anything to do with things on the side of the highway.) The consumer is getting a personalized list relevant to them, over the aggregate data, or they're getting a source-specific list. And since things like one-hit wonders rely on seeing the weekly data over time, they're not discussed in the same way. It's difficult to point out statistical anomalies when the public doesn't see the advanced stats, and really doesn't even see the scoreboard. The weekly data is also being deemphasized in favor of the year end notes, as we discussed last week.
And of course, during the course of this I remembered Danse Macabre exists and that even in high school Camille Saint-Saëns is either a no-hit or two-hit wonder, completely detonating my point.
A one-hit wonder is nonetheless a handy model of a pattern in quiz bowl, though you might have to explain it now before you introduce the concept. Even if it's not an actually bijective relation between last clue and answer, it's enough of a one-to-one correspondence that both writer and editor can feel confident in using it with other uniquely identifying clues, and a player can feel confident in using it, even if they don't know anything about the answer, other than it is a possible answer. So collecting one-hit wonders, noting them to the team when they appear, and even collecting lists of them as study guides, is a plausible strategy to improve your team. So how do we get there?
How to tackle the one-hit wonder
The problem you're going to face as a new coach is that while it's an easy paradigm to teach when you're familiar with the concept and find examples, you won't know the examples without experience. As seen above, I have some troubled recognizing it at times. An objective method of identifying a one-hit wonder in a field can be done by either by searching a publicly available frequency list for that field, or by using a question database search to test your hypothesis that something is a one-hit wonder.
Method one: Frequency Lists
You'll need a frequency list, and you'll need to know how many mentions need to have happened to make the list. We'll call that number n for now. You can start by looking down the list, and seeing any creator's first appearance in the list. You then search the file for other mentions of the same creator. If you have a creator with one creation in the list, and there being only that mention, AND they're getting something like 5n mentions, they're likely a one-hit wonder. If their first appearance in the list is an X mentions, and their subsequent mentions are X/3 or less, they're likely a one-hit wonder, where those mentions further down the tail are the product of writers trying to find other clues than their one hit. Those works further down the tail are possible clues like The Pearl Fishers is for questions about George Bizet, something that gets mentioned in early clues. As I mentioned in a post many years ago on Facebook, details about those deep tail items are useful for tossups, as they are ideas for early clues where the answer is the creator, but details about the top one-hit wonder creation are useful for bonuses.
Method two: Question Databases
This is a less effective method than the previous, and it requires you to sort of collect ideas to test as you read through packets, but it is more applicable to a larger number of fields.
You have to train yourself to suspect that certain clues must lead in only one direction, and then make notes of those possibilities. After you have a list of possible answers to try, you feed each answer into a text search of the database. As you look at the results, you do the same sort of evaluation of peak clues (the one that appears most often) and the other clues that appear in some of the questions. If the peak is 3-5 times the appearances of other clues, it's likely a one-hit wonder for that answer.
You then have to check the results for false positives, where you then take the smallest part of the clue and perform the same search. If the clue then points to multiple answers, it's probably not as useful as you would have hoped.
The Red Queen of frequency lists
The problem I've mentioned with frequency list is that they don't respond well to the new hotness. That is for a new work to break into a frequency list it has to hit quickly and repeatedly in a small window of time. It then has to collect all of those and make its mark before the next iteration of the frequency list is calculated. Typically writers use the top of the frequency list as equivalent to ease of clue, and this means those clues that have been around and were frequent at the beginning stay frequent in the current list, and writers then keep that going by writing about the items at the top of the list. It has the feel of the Red Queen effect, where something must keep appearing to not fall behind.