Are Highly Rated NCAA Frosh Systematically Overrated?

Discussion in 'Nuttin' but Net' started by InstiGATOR1, Oct 27, 2013.

  1. InstiGATOR1
    Offline

    InstiGATOR1 Well-Known Member

    Joined:
    Jan 9, 2011
    Messages:
    3,100
    Likes Received:
    36
    Trophy Points:
    48
    Ratings Received:
    +75
    Well since I have been skeptical of the highly rated UK frosh all off season and because I know that Value Add Basketball:

    http://valueaddbasketball.com/ballall.html

    has up both projected and actual data for over two thousand NCAA players for 2013, I thought I would try to look at this issue a bit systematically. Note the site calls the 2012-13 season 2013.

    So just to start I thought I would look the 25 frosh valued added predicted to make the largest contribution last year. This is a bit different than say the top 25 RSCI ranked recruits, as it appears to me that value add basketball maybe based projected value add in part on projected playing time. All but one of the 25 players are in the RSCI top 31 so the lists match pretty well. Also I deleted Stokes of UTn as he enrolled early midway through the previous season and was not a true frosh in 2012-13. To avoid an outlier affecting the results, I also deleted one recruit RSCI rank 50ish and who was no where near what value add suggested and did not even show up in their actual results.]

    Here is some of what I have found:

    1. The 25 players were ranked prior to 2013 from 3 to 359 with an average of about 188. The 25 players projected value add was between 10.00 and 3.39 with a mean value of 5.14.

    2. The 25 players were actually ranked in 2013 from 20 to 1820 with an average of about 522. The 25 players actual value add was between 7.29 and 0.37 with a mean value of 3.16

    3. Of these 25 players 19 were ranked worse than projected and 20 were had valued add numbers lower than projected. [The difference is due to Anthony Bennett being very slightly worse than projected in value add, but ended up rank a slot or two higher based on his actual value add.]

    One might expect among the 2,000+ players about have to do better than projected and half to do worse than projected. Among top players with a college basketball track record you might expect some regression to the mean or true value and have more of them do worse then expected.

    With true frosh who have no statistical track record in college basketball, one might not expect regression to the mean. I guess one could say these players had to play well their SR years in AAU/high school to maintain their high rankings and thus we might expect some regression to the mean among top high school players. Still top recruits are rated mainly on their potential and that should not regress to the mean. So to me it seems that the evidence above indicates some systematic overrating of true frosh in the value add data.

    Now maybe the 2012-13 true frosh were just not as good as expected. Value add basketball does not contain preseason projections for any other year. Also other systems may not have the possible bias for true frosh that appear to be in the 2012-13 value add data. Still clearly they were expected to be pretty good before the fact.

    As a side note, Michael Frazier did exceed his projection. Of UF 8 rotation players last season, 5 exceeded their projections, though value added had Ogbueze rated higher than some rotation players and he did not exceed his projection. Additionally every player on UK's 2012-13 roster except Hood failed to meet their value add projection.
  2. REM08
    Offline

    REM08 Well-Known Member

    Joined:
    Jun 1, 2010
    Messages:
    4,857
    Likes Received:
    194
    Trophy Points:
    63
    Location:
    Arkansas
    Ratings Received:
    +702
    I'll be honest - I only halfway understand how that rating system works.

    The reason I haven't put in the effort to all-the-way understand it is because I am able to rely on a system of my own. I haven't named it yet, but I think of it as the is this recruiting class capable of major contributions during their first year system. In 2010, the answer was yes. 2011 and 2012? Yes and Yes. 2013? Nope. 2014? I'd put my money on yes - although I guess anything is possible.

    Its a gut thing but I like it. Haven't trademarked it yet though.

    IMO, there isn't much evidence for UK recruits being overrated. Last year certainly saw some underwhelming performances. But its just one year. Poythress, who's year most considered a disappointment, put up pretty much the same numbers to Jamal Mashburn his freshmen year. Should all freshmen be judged by how they performed playing many more minutes than they should have with not enough help around them? Goodwin was asked to do EVERYTHING last year much of the time. In reality, he should have had a similar role to Fraizier as far as minutes and the offenses reliance on him. Is it relevant to compare their years?
  3. BEH
    Offline

    BEH Active Member

    Joined:
    Mar 2, 2010
    Messages:
    1,665
    Likes Received:
    6
    Trophy Points:
    38
    Ratings Received:
    +41
    Kentucky will not win a damned game this year.

    Kansas does not have a chance to win a game, even the exhibitions.

    Last year showed how stupid the rating services were. How in hell did Indiana miss the final four?

    Regression to the mean will be the norm. All the dumb coaches who go and watch these kid are stupid and Calipari is the most stupid. A kid who plays the best in high school will fall on his face in college. It has been this way for the last 50 years.

    Apparently the mean established in stupid studies does not get the input of dumb butt human being who rely on evaluation of real talent.
  4. mdfgator
    Offline

    mdfgator Well-Known Member

    Joined:
    Apr 3, 2007
    Messages:
    4,939
    Likes Received:
    303
    Trophy Points:
    83
    Ratings Received:
    +1,684
    You guys will win some games, keep your head up.
  5. InstiGATOR1
    Offline

    InstiGATOR1 Well-Known Member

    Joined:
    Jan 9, 2011
    Messages:
    3,100
    Likes Received:
    36
    Trophy Points:
    48
    Ratings Received:
    +75
    It is typical to see the UK fans over react to a post about over rating true frosh and of course as could only happen here follow up with post like they are on a UK board. Well no matter how much some here including some moderator suck up to the, I think this is still a UF fan board.

    At least I intend to treat it as such. My interest in this is about Wilbekin and Hill and what to expect of Hill and C.Walker, with a tiny side note about anyone else. As I suggest above, it was not the UK true frosh who were worse than the projections at value add basketball, rather it was about 80% of the highly rated true frosh.

    I do think there might be some evidence of a basketball school only bias as one of the true frosh who did achieve more than projected was Marcus Smart. Still I have not looked at that issue systematically yet.
  6. REM08
    Offline

    REM08 Well-Known Member

    Joined:
    Jun 1, 2010
    Messages:
    4,857
    Likes Received:
    194
    Trophy Points:
    63
    Location:
    Arkansas
    Ratings Received:
    +702
    I don't see how your freshmen premise is able to differentiate between freshmen being "overrated" or this systems inability to accurately project the impact of freshmen. I like the guy's system - for quantifying a players impact after or during a season (when you've got college stats to analyze). I just don't see much reason to believe that freshmen projections are an accurate measuring stick considering a lack of college stats going into the season and a number of other variables in play.
  7. HallGator
    Offline

    HallGator Administrator VIP Member

    Joined:
    Apr 3, 2007
    Messages:
    44,867
    Likes Received:
    1,178
    Trophy Points:
    113
    Location:
    Outer Limits
    Ratings Received:
    +4,540
    Insti, do you have a Cliff Notes version?
  8. InstiGATOR1
    Offline

    InstiGATOR1 Well-Known Member

    Joined:
    Jan 9, 2011
    Messages:
    3,100
    Likes Received:
    36
    Trophy Points:
    48
    Ratings Received:
    +75
    That is a Cliff Notes version above. I have not really written a long version.

    To summarize quickly.

    A. I looked at the top 25 frosh from value add basketball.

    B. I compared their projected value add last year to what they actually did.

    C. About 80% of them were actually worse, some much much worse than projected.
  9. InstiGATOR1
    Offline

    InstiGATOR1 Well-Known Member

    Joined:
    Jan 9, 2011
    Messages:
    3,100
    Likes Received:
    36
    Trophy Points:
    48
    Ratings Received:
    +75
    I just found an April essay in which Ken Pomeroy weighed in on the whole issue of accountability of team rankings in the preseason. I found this quite interesting:

    http://kenpom.com/blog/index.php/weblog/entry/a_look_back_on_preseason_ratings
  10. BEH
    Offline

    BEH Active Member

    Joined:
    Mar 2, 2010
    Messages:
    1,665
    Likes Received:
    6
    Trophy Points:
    38
    Ratings Received:
    +41
    I was trying to be funny and failed.

    I think the reason freshmen have more impact now versus 15 years ago is because there are fewer quality upper classmen on all teams. Up until the one year requirement, many better players were staying two or three years and the quality upper classmen would outshine the quality freshmen. Now we have few quality upper classmen because the pressure to take the NBA money is high.
  11. HallGator
    Offline

    HallGator Administrator VIP Member

    Joined:
    Apr 3, 2007
    Messages:
    44,867
    Likes Received:
    1,178
    Trophy Points:
    113
    Location:
    Outer Limits
    Ratings Received:
    +4,540
    Thanks, that will work.
  12. InstiGATOR1
    Offline

    InstiGATOR1 Well-Known Member

    Joined:
    Jan 9, 2011
    Messages:
    3,100
    Likes Received:
    36
    Trophy Points:
    48
    Ratings Received:
    +75
    I did "cheat" and drop out one IU frosh in the value add top 25 who was something like 50 in the RSCI recruit ratings. He did so little he dropped out off the charts of the actual value add for last season. Had he done better than projected, I certainly would have kept him in the sample.

Share This Page