- Joined
- May 13, 2020
- Messages
- 181
I was bored today :] So of course I nerded out on disc golf review statistics. The result was me creating charts to help visualize how closely Diamond TRs' ratings match a normal curve. Attached are the results in 3 files.
Hypothetically, someone who has reviewed a large number of randomly selected courses and given them "accurate" ratings should line up pretty closely with the normal curve. Someone who has not played a random sample of courses (for example, choosing to only play very highly rated courses), or has reviewed them "inaccurately", will not line up nearly as well.
Of course, this is predicated on the assumption that, like many things in life, the quality of disc golf courses as a whole generally falls onto a normal curve. There should be many courses that are generally of Typical quality (2.5), while very few that are Abysmal (0.0) or Best of the Best (5.0).
This is also a very imprecise system that any statistician would censure me for using. (Every statistics student knows that you shouldn't apply normal distributions to categorical data :/). There's lots of margin for my choices to sneak in. Chief of these is that I had to choose one point of the actual review curve to scale the normal review curve to match - I stuck with 2.5 (the middle of the ratings spectrum) as much as possible, but sometimes it was visually more obvious what was going on to scale it with 2.0 or 3.0.
Enjoy! I'd love to discuss or answer questions here or in my PMs.
Hypothetically, someone who has reviewed a large number of randomly selected courses and given them "accurate" ratings should line up pretty closely with the normal curve. Someone who has not played a random sample of courses (for example, choosing to only play very highly rated courses), or has reviewed them "inaccurately", will not line up nearly as well.
Of course, this is predicated on the assumption that, like many things in life, the quality of disc golf courses as a whole generally falls onto a normal curve. There should be many courses that are generally of Typical quality (2.5), while very few that are Abysmal (0.0) or Best of the Best (5.0).
This is also a very imprecise system that any statistician would censure me for using. (Every statistics student knows that you shouldn't apply normal distributions to categorical data :/). There's lots of margin for my choices to sneak in. Chief of these is that I had to choose one point of the actual review curve to scale the normal review curve to match - I stuck with 2.5 (the middle of the ratings spectrum) as much as possible, but sometimes it was visually more obvious what was going on to scale it with 2.0 or 3.0.
Enjoy! I'd love to discuss or answer questions here or in my PMs.