You know my methods. Apply them.
-Sherlock Holmes “The Sign of the Four“
Hey all. You may have noticed that I’ve stopped posting as often.
Hey now, Relax.
What you may or may not have noticed is that I’ve taken over as Editor over at the Wages of Wins.What this means is that I will be splitting my posting between the two sites.
In fact, this, will be a simulpost.
So if you were missing your daily fix just go over there and you will be sure to find the typical output from me.
Better now?
Let’s rate this Draft!
Projecting the 2011 NBA Draft
I built some models to project rookie performance based on the available stats for the prospective NBA players.
I built two models specifically (go here for the model build parts 1 & part 2 ) and I called them : Yogi and Booboo.
This is a totally legit picture
To give you an idea of the value of the models I decided to look at:
- The probability of landing a better than average player (>.090 WP48) for his first four seasons
- The probability of landing a good player (>.150 WP48) for his first four seasons
I also decided to show this for:
- Any qualifying pick (>400 MP in his rookie Year)
- Any Top 5 pick
- Any Top 10 pick
- Any 1st Round Pick
- And Both models.
And this was done for 1995 to 2009. The table is here:
If I rank the factors in order of predictive value for player success for this data from best to worst I get:
- Both models calling for you to draft the player
- Yogi picks
- Boo Boo picks
- A player being a Top five picks.
Yogi is more picky, Boo Boo casts a broader net and is more accurate.The takeaway though is that the models work.
Last year I did it for the 2010 Rookies and early returns were very good.
See here:
And in summary:
| Greater than .080 WP48? | |||
| Picked by | Outside Top 10 | Top 10 Pick | All Players |
| All Players | 24% | 55% | 31% |
| No Models | 17% | 43% | 22% |
| One Model | 50% | 50% | 50% |
| Both Models | 100% | 100% | 100% |
If we compare the top 10 picks versus model selection and set our benchmark for success at .080 WP48 for rookies, one model picks performed identically to top 10 picks and two model picks were can’t miss prospects in this draft. Only 30% of all picks were successful contributing rookies and the models clearly outperformed that mark.
The GM’s and scouting again get fairly trounced by the math.
The takeaway? Math always wins and it’s worth doing this again.
So here goes.
2011 NBA Draft Rankings
There are 6 good prospects in this draft according to the models and one slightly less attractive one (but he can be had cheap). The best prospect is easily Kenneth Faried, who will in all probability wind up on a pretty darn good team. Kyrie Irving and Derrick Williams the prospective 1 and 2 picks look to be good picks and not busts.
The other three good prospects are Kawhi Leonard, Alec Burks and Markieff Morris (all to be had outside the top 5).
The real steal may end up being Travis Leslie in the second round.
Enjoy the draft!
Tom
06/23/2011
Can you do a post on reviewing your regular season win total predictions (along with the other people you were competing against)?
arturogalletti
06/23/2011
Tom,
Of course. After the draft.
marparker
06/23/2011
FWIW,
Here are my preditions
25= .3+ player
20= .2+player
15==.1+
<15 = bust
all numbers expected after year 3
Highest ever graded out player- Kevin Love , Evan Turner, Demarcus Cousins, Chris Paul,Blake Griffin(all over 28)
I told you so that I'm gloating over- Landry Fields 26, Noah 23, and of course Love
Player I gave a really high ranking who hasnt made it to the nba yet- Mike Green pg Butler, Calathes g Florida
whoops- gave Mike Conley a 22
coulda been- Oden 22
Biggest argument- kevin durant 18 which I still contend he was at 20 but he just kept getting better
I have yet to call someone below average who has turned out to be better than slightly above average and then we're just talking about Redick
all documented over at dave's blog
I have had problems grading point guards who have high turnovers
So I dock any of this issues 10 points to be safe
This years crop
25+
20+Tristan Thompson,Kawhi Leonard,Alec Burks
15+ Kyrie Irving,derrick williams,Klay Thompson,Marcus MOrris,Faried,Tobias Harris,Shumpert,Travis Leslie,Vucevic
bust-Brandon Knight(docked), Jimmer,Singleton,Kemba,Jordan Hamilton,Marshon Brooks,Markieff Morris,Singler,Honeycutt,Parsons, Justin Harper
Some of these busts could catch on in a marginal role. But thats what my 2nd round pick is for.
Rex
06/23/2011
AG, do you adjust for strength of league, or do you just assume all college stats are created equal?
Rex
06/23/2011
The next frontier: international prospects.
arturogalletti
06/24/2011
I’m getting to that.
ClipFan
06/23/2011
ClipsNation thanks you for giving us much needed hope after failing to acquire Iguodala.
ges05a
06/24/2011
I’m sure Arturo has something better than this, but my rough-and ready method for factoring in competition level and “potential” is to:
figure PAWS 40 for each player
add 10% of the player’s PAWS 40 if they play in a major conference
do nothing if they play in a weaker major conference or a stronger mid major (C-USA, Mountain West, A-10, maybe WCC)
subtract 10% of their PAWS40 if they play in a minor conference
then, factor in the likelihood that the player will improve:
if the player is a senior (or the age of a senior), I do nothing
if a junior, I add 5% of their PAWS40
if a sophomore, add 10% of PAWS40
if a freshman, add 15% of PAWS40
Anyway, although this method is unscientific, it gives a better idea than do the raw stats, although it still shows that teams should not pass on somebody like Paul Millsap simply because they played in a minor conference in college. I think doing the adjustments as a function of the player’s PAWS40 prevents the sort of over-rationalization of poor performance by young players that seems to pervade draft night in the NBA (e.g., “Josh Selby was poor last year, but he was just a freshman; he is a first-round talent”). Just being younger doesn’t mean you get credit for playing when you can’t.
J. Glanton
06/24/2011
Arture, why no model on Trey Thompkins? There should be 3 years of data on him, yet he is in the “Don’t Ask” section. I’m asking.
J. Glanton
06/24/2011
Sorry, ArturO
RandomStrring
06/25/2011
Also along the same lines, why is Bismack Biyombo the only international not in the “Don’t Ask” section?
Michael
06/25/2011
Sorry to be critical but I hate what has happened to the proff’s site. It looks like a poorman’s version of this one. I would have made a comment over there but it is now a pain to post comments over there as well. I’ve been reading Berri’s blog for years now and I have to tell you it wasnt broke and did not need ‘fixing’.