PDA

View Full Version : RSCI: OAD v undrafted prospects



niveklaen
04-07-2017, 03:27 PM
There seems to be a lot of inconsistency on what RSCI rankings people think indicate whether it is a failure for a player to not go OAD or to end up undrafted. Draft express has these great pages that show where/when each incoming freshman class member goes in the draft.



http://www.draftexpress.com/RSCI/2016/



I put together a chart to compare players at each rank (1-35, too lazy to go lower) in the OAD era - 10 years starting 06-07 through 15-16 freshman classes. standard small sample size caveats and edge effects for the last 2 years. Caveats aside, with only a few exceptions, OAD odds descended and undrafted odds ascended very tightly based on a player's RSCI. If you swapped 4 & 5, 8 & 10, and 15 & 16 the relationship would be almost absolute, but real life does have a bit of noise.



What I found was that players can effectively be lumped into blocks based on their rank to predict likelihood of being OAD and of going undrafted. I think it is useful to keep this in mind when projecting future OAD likeliness and discussing who was a bust.



1-6 - 60 players. 47 OAD. 5 undrafted. Duke 6-7 OAD, none undrafted. UNC 1-5 OAD, 1 undrafted

7-16 - 100 players. 33 OAD. 37 undrafted. Duke 1-3 OAD, 2 undrafted.* UNC 1-7 OAD, 3 undrafted

17-29 - 130 players. 9 OAD. 73 undrafted.

30-35 - 60 players. 0 OAD. 50 undrafted.



*our two undrafteds here are Jeter and DT. Not sure what the best way to treat them (or any of the transfers) is.

...and of course they are both edge effect players. This year's sophomores, juniors, and seniors each still have a chance to be drafted. This has essentially no impact on top 6 guys as only Cliff Alexander remains undrafted from this group. The 7-16 group has 15 guys who could still be drafted and the 17-29 has 27 so risk of going undrafted at all could be overstated for those two groups. Excluding the edge group, 7-16 has 22 of 70 undrafted and 17-29 has 46 of 91 undrafted.

The whole chart is too big to post, but here is a summary of each ranking's outcomes:


RSCI

1: 9 OAD LOTS, 1 2AD LOT BARNES

2: 8 OAD (6 LOTS (RIVERS),2 1STS), 1 2AD 1ST, 1 UNDRAFTED KANSAS
3: 9 OAD LOTS (INGRAM, OKAFOR, IRVING, B.WRIGHT), 1 2AD 1ST UCLA
4: 6 OAD (5 LOTS, 1 2ND), 1 2AD, 1 3AD (HENSON), 2 UNDRAFTED
5: 9 OAD (6 LOTS(PARKER), 2 1STS, 1 2ND), 1 3AD (LAWSON)
6: 6 OAD (3 LOTS, 2 1STS(TYUS), 1 2ND), 1 2AD LOT, 1 SR (SINGLER), 2 UNDRAFTED (MACADOO & BAD H TWIN)
7: 5 OAD (1 LOT, 1 1STS, 2 2NDS), 1 2AOD 1ST, 1 3AD 2ND, 3 UNDRAFTED
8: 3 OADS (1 LOT, 2 1STS), 1 2AD, 1 3AD(ELLINGTON), 5 UNDRAFTED
9: 2 OADS (1 1ST, 1 2ND), 5 2ADS, 3 UNDRAFTED (JUSTIN JACKSON)
10: 5 OADS, 2 2ADS (ED DAVIS), 3 UNDRAFTED
11: 3 OADS, 2 3ADS(HENDERSON), 1 SR(KELLY), 4 UNDRAFTED (JETER)
12: 4 OADS, 4 2ADS, 2 UNDRAFTED
13: 4 OADS, 2 3ADS, 2 3ADS(HAIRSTON), 3 UNDRAFTED
14: 3 OADS (WINSLOW), 2 2ADS, 1 3AD(BULLOCK), 4 UNDRAFTED
15: 0 OADS, 2 2ADS, 1 3AD, 1 SR, 6 UNDRAFTED (HICKS & PINSON & THORNTON)
16: 4 OADS, 2 3ADS, 4 UNDRAFTED
17: 1 OAD, 3 2ADS(E.WILLIAMS), 1 3AD, 1 SR, 4 UNDRAFTED
18: 1 OAD, 3 2ADS, 2 3AD, 1 SR (MASON), 3 UNDRAFTED
19: 0 OADS, 0 2ADS, 1 3AD, 3 SRS(NOLAN, ZELLER), 6 UNDRAFTED
20: 1 OAD, 2 2ADS, 1 3AD, 1 SR, 5 UNDRAFTED
21: 0 1ADS, 2 2ADS, 0 3ADS, 1 SR, 7 UNDRAFTED (LUKE, LANCE)
22: 1 OAD, 0 2ADS, 1 3AD, 8 UNDRAFTED (STRICKLAND)
23: 1 OAD, 1 2AD, 1 3AD, 7 UNDRAFTED
24: 0 OADS, 02ADS, 2 3AD (HENDERSON), 1 SR, 7 UNDRAFTED (ALLEN)
25: 0 OADS, 1 2AD, 3 3ADS, 1 SR, 5 UNDRAFTED (ZOUBEK, AMILE)
26: 1 OAD, 1 2AD, 3 3ADS (HOOD), 5 UNDRAFTED
27: 1 OAD, 2 2ADS, 1 3AD, 1 SR, 5 UNDRAFTED
28: 1 OAD, 1 2AD (MARSHALL), 1 3AD, 1 SR (GBINIJE), 6 UNDRAFTED (SCHEYER)
29: 1 OAD, 1 2AD, 2 3AD, 1 SR, 5 UNDRAFTED (COOK)
30: 0 OADS, 2 2ADS, 0 3ADS, 1 SR, 7 UNDRAFTED (MATT)
31: 1 SR, 9 UNDRAFTED
32: 1 2AD, 9 UNDRAFTED (SEMI)
33: 10 UNDRAFTED
34: 2 2ADS, 1 SR, 7 UNDRAFTED
35: 1 2AD, 1 3AD, 8 UNDRAFTED

bob blue devil
04-07-2017, 05:07 PM
...

Thanks for compiling and posting this. I've been considering the same, particularly in light of people's reaction to Marques's freshman year - not sure why people act like it was a big surprise he didn't go OAD as the #11 rsci guy.

One note, I'm not particularly comfortable with the cutoff points - feels like data mining to choose 6, 16, 29, and 35 as cutoffs. maybe something like 5, 15, 30 and 50 would be more defensible.

Kedsy
04-07-2017, 09:41 PM
11: 3 OADS, 2 3ADS(HENDERSON), 1 SR(KELLY), 4 UNDRAFTED (JETER)


I like the idea of this, but where are you getting your RSCI data? According to this RSCI site (https://sites.google.com/site/rscihoops/home), which I thought was the "official" RSCI, for example, Chase Jeter was #14, Gerald Henderson was #10, and Ryan Kelly was #14, but unless I'm reading your chart incorrectly, you list them all as #11. I didn't check beyond that, except for the class of 2014, where you appear to be mis-ranking Jahlil Okafor (#1, not #3), Tyus Jones (#7, not #6), and Justise Winslow (#13, not #14). I'm not sure if these inconsistencies do much to alter your overall summary numbers, but it makes me wonder if there are multiple RSCI sites with different rankings.

Nugget
04-07-2017, 10:03 PM
I like the idea of this, but where are you getting your RSCI data? According to this RSCI site (https://sites.google.com/site/rscihoops/home), which I thought was the "official" RSCI, for example, Chase Jeter was #14, Gerald Henderson was #10, and Ryan Kelly was #14, but unless I'm reading your chart incorrectly, you list them all as #11. I didn't check beyond that, except for the class of 2014, where you appear to be mis-ranking Jahlil Okafor (#1, not #3), Tyus Jones (#7, not #6), and Justise Winslow (#13, not #14). I'm not sure if these inconsistencies do much to alter your overall summary numbers, but it makes me wonder if there are multiple RSCI sites with different rankings.

The OP is using what Draft Express calls its "RSCI," which they seem to be representing as:

(1) taking the "official" RSCI website data and

(2) tweaking it slightly to account for RSCI's (admitted) flaw in "penalizing" 5th year players whom some of the scouting services included in the RSCI averages didn't rank -- e.g., Thon Maker who was #37 according to RSCI because he was only ranked in 4 of its 6 inputs (he was in the Top 10 in 3 and around #20 in the 4th), whereas Draft Express's version has him at #9 b/c it only averages in those rankings that considered Maker.

But, you are right that even on an input-by-input basis, the inputs Draft Express says it's using don't seem to quite match up to the RSCI site.

Maybe, b/c Draft Express isn't pulling from the "final" RSCI versions? Hard to tell.

niveklaen
04-08-2017, 02:49 AM
I just pulled from the draft express page for my rsci data.