A friend asked whether anyone else had noticed a pattern where big contra dance events were generally booking more established callers since restarting. This could make a lot of sense: the established callers will be less "overplayed" than they had been, and many events will be less robust financially and so more risk averse. Can we use the trycontra.com/events data to see if this is happening?
I have the caller listings for 2016, 2017, 2018, 2019, and 2023, plus part of 2024 for dance weekends, camps, long dances, and festivals. And you can see the raw data in this sheet if you think I'm missing any!
A reasonable measure for whether someone is "established" is how many events they've previously been booked for. But where to draw the line? Someone calling their first is clearly new, but so probably is someone on their third? I decided to graph several (code):
| threshold | 2018 | 2019 | 2023 | 2024 |
|---|---|---|---|---|
| 1 | 10% | 8% | 9% | 13% |
| 2 | 16% | 15% | 11% | 18% |
| 3 | 21% | 22% | 15% | 22% |
| 4 | 24% | 27% | 20% | 23% |
| 5 | 28% | 29% | 22% | 27% |
Overall it looks to me like this was a real effect for 2023, but with 2024 it's going back to normal.
You can see I've left off the first two years: I don't want to count someone as "new" just because it's 2016 and that's the first year I have any data. One limitation with this approach is that someone who called a lot pre-2016 and then took a few years off will (briefly) show up as new in year they returned. Another is that someone who's been calling for years but doesn't take many gigs will show up as established, and be more likely to show up in the first few years.
These effects will be more pronounced at higher thresholds, because it's less likely someone called enough in 2016 and 2017 to pass the threshold, so I trust threshold=1 ("first time") a lot more than threshold=5 ("fifth time"). Even that is still not perfect: this categorization puts Dudley Laufman as a new caller in 2024, since it didn't see him in earlier years. Still, skimming the categorizations of callers, it looks pretty good to me.