phillyinNJ posted:awesome work! and thanks for stating the thread...good stuff. Found the below attachment on the internet the other day that seems somewhat relevant to the topic...
I saw that graph on @TreadAthletics. I never want to discourage people from crunching the numbers, but I think they made some assumptions that hurt the accuracy of the numbers.
First, the title of "Chance of Pitching in College" is misleading. I'm pretty sure they did not confirm that those 7,000 players in the sample have actually pitched in college. Lots of college position players have a FB number on their PG profile. MSU outfielder Jake Mangum (SEC Freshman of the Year) has an 85 FB on his profile.
Second, there are lots of kids pitching in college who don't have a school commitment listed on PG. Anecdotally, it seem like the likelihood goes down from D1 to D3 to D2 to juco to NAIA. So if they want to look at "Chance of Pitching in College", the lower velocities are most likely under-represented in this group.
If the hypothesis is that higher velocities directly relate to chance of pitching in college, very few would disagree. But if you're trying to put real percentages to that relationship, I don't think you can use this data for that purpose. I think 2019Dad's analysis gives a much more accurate assessment of the relationship between velo and pitching in college.