https://docs.google.com/spreadsheets/d/1BBe02UrxndbKHJ5xx3cyvI7c0Jg5RBb0-ziTJdch0Fs/edit?gid=494323376#gid=494323376

I made some adjustments to the data that are specific to my model (and sometimes incorrect). Lead scores are adjusted so that the top is 0 (and 1 hold from the top is -1, etc.). If I couldn’t find the number of holds in the route, I assumed that the highest non-top was a -2.

For boulder, the scores are listed as (attempts to top 1, attempts to zone 1, attempts to top 2, etc.). For B&L, it’s (attempts to top 1, attempts to 10pt zone 1, attempts to 5pt zone1, etc.). Inf means it wasn’t reached. In some cases, it was impossible to get data on attempts, so 0’s were used for successful tops/attempts with no attempt data. There were also some cases where I had access to the attempts to the top/high zone, but not the number of attempts to lower zones. In these cases, I just listed the highest number of attempts for everything. (eg it took 3 attempts to reach the 5pt zone, 5 attempts to reach the 10pt zone, and 7 attempts to reach the top, but I just had to fill in 7’s for everything). These didn’t affect my model, but might matter for other potential uses of the data.

I got most of my data from github: https://github.com/DavidBreuer/ifsc-analysis. The data goes up to mid-2022. I manually added the last few years of data, but forgot to keep the old versions before I reformatted everything. I’m hoping to do some scraping of my own before next season, so I’ll update if I have something better.

LikeLike

Do you have a link to see the dataset you compiled?

LikeLike