What I have for each player is class year and ratings. I am about halfway done adding in school prestige and coach (human v sim, here i accounted for if the school is just this year a simmy). I can sort the data any way with that. I will try and work on getting some of the other ideas done and posted throughout the week.

Oldave, I almost always start 1 Freshman or have one play very significant minutes at Illinois. I don't know if getting rid of all the Freshmen really helps that much?

This post could not be converted. To view the original post's thread, click here.

Quote: Originally Posted By zhawks on 10/26/2009Oldave, I almost always start 1 Freshman or have one play very significant minutes at Illinois. I don't know if getting rid of all the Freshmen really helps that much

i know, thats true, heck, i am starting one now (but that is odd for me), but im looking at it this way...we know there are data points among these 864 players that we would like to get rid of because they actually represent players who play very little if at all. i know that not all of these players are freshmen (and i concede that many frosh play significant minutes or start).

so, lets start with an assumption. lets say that 3.5 players per team do not play a significant role or are inferior players pressed into duty due to a lack of other options. i might choose to say there are 3 such players per team, or 4, but we'll say 3.5 for now. remember, this includes the slow, unathletic, mediocre ballhandling soph who you plug in at backup SG for 10-12 minutes per game and hope does not kill you. maybe there are not so many of these type on top5 teams, but when we are talking about the bottom half of the big6 conf, i suggest there are more than a few of these.

im going to so that, on average, 2.5 out of those 3.5 players (maybe more?) are freshmen. so, i am saying that only one or so of these fringe players is a nonfreshman.

If you can buy those assumptions (and probly a few of you cannot) then what I am asking is to eliminate 180 of these bad data points (72 teams * 2.5 players per team).

I would be conceding that we are leaving 72 "bad data points" in the study. While i would like to eliminate all the bad data points, this is the best way i can think of to eliminate the vast majority of them.

Why do I think these data points are "bad"? because the question i would like to answer is "what are the ratings averages for all the significant players (starters and key bench players) on all big six teams?" and i feel like eliminating the freshmen gives us a better chance at a meaningful answer.

does that make any sense?

Quote: Originally Posted By zhawks on 10/25/2009I would lump it by upperclassmen, Sophomores and Freshmen since Jrs/Srs are pretty much the same

i agree with this. but when we look at freshmen (which i agree is something i would really like to see) do we know how big of an impact walkons have?

I am thinking that there might be 0.5 walkons per team (if you look at 10 teams, there might be a total of 5 walkons.

If we assume 4 freshmen per team, then i am suggesting that one out of every eight (or 12.5%) of all freshmen are walkons. and I think that number may be a tad low.

it just seems to me that walkons would lead to a deceiving result and lead us to believe that freshmen are weaker than they actually are as these bums would pull down all the averages a bit.



make any sense?

If it isn't too hard to format, it could be interesting to see what the standard dev. is for each of these ratings so one can better judge what's an exceptionally good or bad rating for a player.

oldave, I only pulled scholarship players.

Quote: Originally posted by jcf16 on 10/26/2009If it isn't too hard to format, it could be interesting to see what the standard dev. is for each of these ratings so one can better judge what's an exceptionally good or bad rating for a player.

Once I get done pulling the school data I can do some more stats. Prolly won't have time until this weekend to get some more good work done though.

I hope you didn't pull all this data manually.

. . Love this. It should be very instructional, especially when just starting at D1. .

the DE averages are sick.

Quote: Originally Posted By zhawks on 10/26/2009oldave, I only pulled scholarship players

nice. that makes me feel much better.

nice job Z. pretty cool stuff