(07-19-2017, 10:43 AM)37thchamber Wrote:Interesting method. I don't think averages is a good way to go though, because it's still skewed. Especially with so few people to compare. That's how you end up with these insane times, I guess.
Still good stuff, though. Different take on things, gave me a few ideas for stats stuff to do in future.
Cheers man! Yeah I felt that going with averages would've been a fair method and as soon as I saw the sub 4.22 times I felt like although that's unrealistic in the NFL, this is the NSFL. I didn't want to change the formula which I felt worked well for the majority of people on all the drills, just because some outliers focused on one attribute. I also feel like when I do this for next years prospects that although these are impressive combine times, if I changed it now then it wouldn't be a fair playing field for every prospect in the combine after S2. What are your thoughts on including 50 TPE players? I feel like if I included everybody (which meant I'd have to start collecting data at the start of each year) it would raise the averages considerably depending on builds, meaning we wouldn't see as quick and often 40 times.
