Fantastic. This was literally going to be a DS grad Capstone project for me, though I was surprised at how difficult it was to get a hold of this data (intended to use Mountain Project API). I intended to build a front end that would allow users to do something like "Show me the very best climbs per grade within a X mile radius, for Bouldering, Sport, Trad, (maybe ICE). Idea started from daughter and I discussing things like, if you could only do one V2 in the United States, what would it be?
Thanks, glad you liked it. Yes, the OpenBeta database is the foundation of everything here, it was built with a custom crawler. There is a lot of info that is not available through the MP API, like individual user ratings as opposed to just the mean, which is all you can get from the MP API. Your idea (more of a route-finder) would be a great contribution since my map is really geared toward area searches, not individual routes/boulders. You should try it with the OpenBeta data! Also, we don't have rating data for boulders yet, but will definitely add it in the future.
You mention "However, there is a problem with RQI: hard routes get fewer ascents and thus fewer votes". That is pretty intuitive. It also seems to me at times, "easier" routes are often rated lower simply because they are easy, and at times, stronger climbers can cause their ratings to be lower than they otherwise would be. Did you see this in your exploratory analysis. I am not sure how it would be corrected honestly, just curious. It would also be interesting to see a chart of the percentage of climbs that qualify as "Classic" by grade using your method.
"It also seems to me at times, "easier" routes are often rated lower simply because they are easy, and at times, stronger climbers can cause their ratings to be lower than they otherwise would be."
This could definitely be true, I did look at how average stars, RQI, and ARQI change with difficulty, and they all tend to increase with difficulty. The increase is least pronounced with RQI. I also quickly plotted the % of classics by grade just now, and it peaks in the 12+/13- area at around 10-12%. You can look at the additional figures (and plotting notebook) here:
Fantastic. This was literally going to be a DS grad Capstone project for me, though I was surprised at how difficult it was to get a hold of this data (intended to use Mountain Project API). I intended to build a front end that would allow users to do something like "Show me the very best climbs per grade within a X mile radius, for Bouldering, Sport, Trad, (maybe ICE). Idea started from daughter and I discussing things like, if you could only do one V2 in the United States, what would it be?
Thanks, glad you liked it. Yes, the OpenBeta database is the foundation of everything here, it was built with a custom crawler. There is a lot of info that is not available through the MP API, like individual user ratings as opposed to just the mean, which is all you can get from the MP API. Your idea (more of a route-finder) would be a great contribution since my map is really geared toward area searches, not individual routes/boulders. You should try it with the OpenBeta data! Also, we don't have rating data for boulders yet, but will definitely add it in the future.
You mention "However, there is a problem with RQI: hard routes get fewer ascents and thus fewer votes". That is pretty intuitive. It also seems to me at times, "easier" routes are often rated lower simply because they are easy, and at times, stronger climbers can cause their ratings to be lower than they otherwise would be. Did you see this in your exploratory analysis. I am not sure how it would be corrected honestly, just curious. It would also be interesting to see a chart of the percentage of climbs that qualify as "Classic" by grade using your method.
"It also seems to me at times, "easier" routes are often rated lower simply because they are easy, and at times, stronger climbers can cause their ratings to be lower than they otherwise would be."
This could definitely be true, I did look at how average stars, RQI, and ARQI change with difficulty, and they all tend to increase with difficulty. The increase is least pronounced with RQI. I also quickly plotted the % of classics by grade just now, and it peaks in the 12+/13- area at around 10-12%. You can look at the additional figures (and plotting notebook) here:
https://github.com/rytheranderson/OpenBeta_blogs/tree/main/EstimatingRouteQuality/figures
It is hard to say exactly why this happens. Just brainstorming:
- It is possible that harder routes generally have better quality rock.
- FAs may make more of an effort to make sure the line is good before trying harder routes (since more work will go into the send).
- People may be more psyched to send harder routes and rate them better because of that.
- As you said, people may be more critical of easier routes just because they are easier.
And, yeah, I'm not sure if this can be (or needs to be) corrected.