Hollinger's Draft Rater

John Hollinger does this annually. Blake Griffin is ranked 2nd this year behind Ty Lawson, who scored surprisingly high. Overall, Griffin is the 6th highest rated prospect since 2002, behind Dwayne Wade, Carmelo Anthony, Kevin Durant, Mike Conley, & Lawson. Many of the rankings, both past and present, are suspect.

Hollinger offers a semi mea culpa on Eric Gordon's low rating last year: "Gordon is perhaps easier to understand because he was playing hurt at Indiana and his primary skill (shooting) didn't show through statistically, but that doesn't excuse the others."

Whatever.

7 months ago mikey p 33 comments 0 recs |

Story-email Email Printer Print

Comments

And this is one reason why I believe statistical analysis in basketball is in its infancy (compared to sabermetrics). All of Hollinger’s stats seem to be totally arbitrary and full of crap. PER is probably the most useless statistical rating of all time.

Stuck in limbo.

by PaperClip on Jun 18, 2009 5:07 PM PDT reply actions 0 recs

Hurm...

I think the problem is the nature of the game. Fielding statistics in baseball are at the same stage as statistics in basketball. For example, the spots where plays are made vary to such a degree (and this is one of almost countless variables) that it becomes too difficult to gauge player ability. Baseball (hitting and pitching) are tailor made for deep statistical analysis. Everyone is pretty much in the same spot at regular intervals.

If the runner is stealing second, the second basemen will probably be closer to second base when the bat is swung. You can count on that pretty regularly I’d imagine. When someone shoots a three-pointer, players react is so many different ways. If you’re Dwight Howard and you are playing with guys who don’t/can’t rebound, you will end up having a larger share of easy rebounds compared to someone like Luis Scola or Kendrick Perkins who play with other rebounders. What statistic accounts for that?

I’ve stopped looking at +/- at all. I find it ridiculous that Kevin Durant is near the bottom on the season. He’s one of the league’s best players!

by Jordinho on Jun 18, 2009 10:50 PM PDT up reply actions 0 recs

"He’s one of the league’s best players!"

Says who?

"So what are the odds that the Clippers can compete next season given their limited flexibility? Slim and none."

by John R on Jun 19, 2009 9:03 AM PDT up reply actions 0 recs

Maybe you should just go out and watch the guy play

He’d have a much better WIN score with a more mature team around him. Not that this matters.

Basketball is not rocket science. Just turn on the TV, watch a game where Kevin Durant is playing, and get back to us.

by Jax on Jun 19, 2009 9:59 AM PDT up reply actions 0 recs

I saw plenty of them

And I saw him lose to the 08-09 Clippers 3 times.

So again, what is best about him?

I see he’s a forward but is only shooting 45% from the field, I see he commits more turnovers than he gives up assists, I see that he is 6’9 but doesn’t even record 1 block per game, on and on.

His WIN score, whatever, would go up if he were better, regardless of who was around him.

He’s a nice player, but he’s pretty far from one of the best.

"So what are the odds that the Clippers can compete next season given their limited flexibility? Slim and none."

by John R on Jun 19, 2009 12:22 PM PDT up reply actions 0 recs

I get it - you see his stats - and that's about it

Once again, please go watch the games and enjoy the players. That’s really what basketball is all about. Once you do that, you’ll begin to see why you can’t really analyze basketball players based on statistics.

Having said that, his WIN score, whatever, would go up if the players around him were better. As I said above.

by Jax on Jun 19, 2009 5:08 PM PDT up reply actions 0 recs

Well...

First John R, there is really no reason to be deceptive about statistics on the internet: he’s shot 45 percent over two seasons. But he shot almost 48% last season with a 42% three point shot. That makes his effective FG% 51%. If he shot 51% but was not a three point shooter, you’d have nothing to say. So say nothing.

Second, nobody looks at assist/turnover stats for 6-9 forwards. He’s not a distributor.

Third, his win share total is related to how many games the team wins, so having better players will inevitably get him some win shares he wouldn’t normally get.

Fourth, he lost to the Clippers three times? So you’re saying his team isn’t very good, right? So how does going 1-3 against another bad team mean that the best player on the team is automatically not one of the best in the league? Your logic is nonexistent there.

This pattern of analysis (started by lazy sports writers who only wanted to cover winning teams) that states that players on bad teams shouldn’t get their due, has to stop. He plays with scrubs. That makes him a scrub? This isn’t forty years ago when there were 14 teams. Some of the best players frequently get to share the starting line-up with some of the worst. If you want to complain about someone, complain about the commissioner and his expansion addiction.

by Jordinho on Jun 19, 2009 5:36 PM PDT up reply actions 1 recs

Great post

Thank you for taking the time to debunk John R’s nonsense. I was just being lazy about it. ast/turnover ratio – LOL

I believe last year was KD’s second season and he shot 48% as an SF.

Those posts above are vintage John R nonsense. They really show how little he knows about the game.

by Jax on Jun 19, 2009 10:51 PM PDT up reply actions 0 recs

We aren't talking Win Shares

That was Jax (intentionally?) confusing the issue.

His Wins per 48 is his own. If there is a weakness in WP48, it is that it has diminishing returns when there are multiple great players on the same team. That is to say, if Durant is truly the only good player on his team, we would expect there to be diminished diminishing returns. His weak team would HELP his WP48, not vice versa.

I’m sorry he tricked you into attacking a strawman with his imprecise (intentionally?) use of language.

Now as for AST/TO for 6-9 SF, ALL of the SF in the Finals would ask that you please look at them. Lamar Odom, Trevor Ariza and Hedo Turkoglu respectfully request that you consider playmaking when you consider their value. Though only 6’8 and not 6’9, I’m sure Lebron does as well.

Though I guess my accusation against those who don’t like advance stats stands. Scoring is all that matters. How dare we consider anything else?

It makes perfect sense to consider team performance when evaluating a player. To do otherwise would be nothing but myopic. Kevin Durant would be no better than, what, the 4th or 5th best player on the Lakers? I just don’t seem him being one of the top 30 or so players in the league today.

"So what are the odds that the Clippers can compete next season given their limited flexibility? Slim and none."

by John R on Jun 22, 2009 10:05 AM PDT up reply actions 0 recs

"Jax intentionally confusing the issue"

Yes, Mr. conspiracy theorist, I’m out to get you.

And I stand by my comments. In addition, John R’s typically patronizing comments leave a lot to be desired.

Jordinho, I don’t think you’re too stupid to see the problems with WIN (lose?).

Finally, I didn’t know PER analyzed scoring only.

Kevin Durant would be the second best player on the Lakers, at worst. And he’s in the top 20 players in the league today and likely will be in the top 5 for years to come.

by Jax on Jun 22, 2009 10:19 AM PDT up reply actions 0 recs

Actually...

I was originally talking about +/- in reference to Kevin Durant. I don’t remember anyone saying anything about wins per 48 minutes.

In the case of win shares however, the better the players around you, the assumption is the more wins you’ll have. And the more wins you have, the more win shares to go around. Basically look at it like this: 12 guys contributing to 20 wins vs. 12 guys contributing to 40 wins. Double the amount of win shares to go around yet still 12 guys. Then you get into whether you are going to give out negative win shares or not but still, I’m guessing that a player can’t acquire more win shares than the team has wins. Look at Pau Gasol: he had his career high in win shares this year, even though his stats are about equal or less than his stats in MEM. However, the Lakers won 65 games this year, which undoubtedly had something to do with his WS total.

As far as 6-9 SFs go, you named Odom, Ariza, Turkoglu & James. Let’s break this down to completely dissolve your argument:

Odom: 08-09 AST/TO ratio = 2.6/1.8
Ariza: 08-09 AST/TO ratio = 1.8/1.1
James: Distributor. If you reread my post I clearly say that Durant is a 6-9 SF and also not a distributor. Therefore his assists, or lack of assists, are irrelevant. He is also the main scorer on the team, and as a result is forced to make plays on his own a lot. This can be a source of excessive turnovers. Charles Barkley has a 3.9/3.1 ast/to ratio for his career. He’s three inches shorter as well, which would make it easier for him to dribble without turning the ball over.
Turkoglu: Not a distributor. His assists come from driving and passing off to reliable shooters: Nelson, Pietrus, Lewis & Lee. However, he was used as a PG in the finals at times and showed how misguided Stan Van Gundy was: 08-09 Finals AST/TO ratio = 3.8/2.6

What I just showed you is a great example of how stats are misused in basketball. The first mistake is saying that non-distributors are distributors when you can easily check what their assist averages are. Second is calling someone a distributor because of how many assists they get. Five assists a game does not give a player distributor status. Paul Pierce has more assists per game than Turk. Is he a distributor too? Third, if you actually watched the players you mentioned you would immediately know which ones are distributors and which ones aren’t, irrespective of their stats. Ariza, Odom & Turk? No. James? Yes. Durant? No.

Finally, despite what you may think, scoring is the most important stat in basketball. The problem with basketball analysis isn’t just that not enough game situations can be objectively measured, but also that there aren’t enough metrics to analyze why a player scores what they score. Do you know why, specifically, Hedo Trukoglu scored 19.5 pts/game last year but only 16.8 pts/game this year? You can guess, maybe, but at the end of the day we don’t know why his shooting percentage went down… unless your name is Shane Battier. He gets special stats that tell him where certain players are less effective, and then defends them toward those spots. I don’t know what those stats are, and neither do you.

At the end of the day, the most important box score totals are PTS, AST, REB, STL & BLK. Only two of those are not directly related to scoring a basket, and even those (REB & STL) are only as important as they relate to scoring (REB = scoring/eliminating second chance points & STL = PTS off turnovers or PTS avoided). If a player makes a shot, the last four stats are irrelevant. Having someone who can reliably bypass the last four stats (not having to be fed the ball by a capable distributor in order to score, not allowing a chance for a defensive rebound, not allowing the ball to be stolen later in the possession, and not being blocked) is a very valuable thing. The Magic had no two-dimensional players at the core of their team, and still made it to the finals. This is because they shot well. All those other stats are crutches to fall back on when shots aren’t falling. If the shots are falling (kind of like eff-51%) then you don’t have to worry too much about a 6-9 scorer’s assist/turnover ratio.

by Jordinho on Jun 22, 2009 9:43 PM PDT up reply actions 0 recs

The fact that you exposed

John R melding his rook stats into the total, ignoring the 48% FG this year is devastating.

by Jax on Jun 23, 2009 12:37 AM PDT up reply actions 0 recs

can you just copy and paste for those who are non insiders. Thanks

by highriser on Jun 18, 2009 5:42 PM PDT reply actions 0 recs

Sorry...

I support excerpting, but no one should copy and paste entire articles, especially from subscription sites. It’s bad form at least, and probably illegal (though I’ve been told different things on that). At any rate, I don’t condone it, and would remove it.

In this world, you must be oh so smart or oh so pleasant. Well, for years I was smart. I recommend pleasant. - Elwood P. Dowd

by Steve Perrin on Jun 18, 2009 6:04 PM PDT up reply actions 0 recs

Welcome back, Steve-O!

Stuck in limbo.

by PaperClip on Jun 18, 2009 6:18 PM PDT up reply actions 0 recs

if we cant read the whole article whats the point of the excerpt ? its like saying the answer is 3!, when we don’t know the question…

by andrewexd on Jun 18, 2009 6:58 PM PDT up reply actions 0 recs

I'm sorry, too

Don’t feel right about that. I did copy the Clipper related points.

Matthew 6:27

by mikey p on Jun 19, 2009 9:22 AM PDT up reply actions 0 recs

710 ESPN

Just wondering if anyone else was listening to the Mason and Ireland radio talk show this afternoon. They had Jalen Rose on as a guest,and they were talking basketball,mainly what the lakers were going to do in the off season,and who they could afford to bring back or let go.At the end of that conversation,it turned to the Clippers,and not being surprised, the two host, Mason and Ireland began bad mouthing the Clippers,and how toxic the players are and how this will affect Blake Griffins growth in the NBA, Then they asked Jalen Rose what he thought about the Clippers and when he thought the organization would eventually turn things around. Jalen said “let me be the first to say this, I believe the Clippers will make the playoffs this up coming season” which brought sounds of shock to Mason and Ireland, who commented are you serious?, Jalen continued by saying, "The Clippers have one of the most talented rosters in the League, and if they stay healthy the entire season,they will be in the playoffs. He continued by saying that he has been up to UCLA, a couple of days a week,and Baron Davis has been working out up there playing basketball and he is in great shape, and is working extremely hard. Jalen ended his conversation, by saying, “remember i said it, the Clippers will make the playoffs”. I was just a little surprised to hear Jalen Rose say what i’ve been thinking all along.

by heights on Jun 18, 2009 8:41 PM PDT reply actions 0 recs

Jalen's right!!!!

Clipper nation will finally play LOUD!!! 1..2..3..“RING”…stupid KOBE.

by ddub177 on Jun 18, 2009 9:08 PM PDT up reply actions 0 recs

Weak.

I listened to the whole podcast(which cuts off Jalen’s interview) online, they must’ve talked about the clippers after that.

by STUCK IN LA on Jun 18, 2009 9:32 PM PDT up reply actions 0 recs

The thing about Baron

makes me the most hopeful. But wow clippers getting some respect that is very surprising.

In Gordon we trust

by bestclipfan on Jun 18, 2009 9:32 PM PDT up reply actions 0 recs

can he take over as our coach..... :)

by dulciusEXasperis on Jun 19, 2009 8:16 AM PDT up reply actions 0 recs

Jalen..

Thought we would make the playoffs last season too.

by Newtybar on Jun 19, 2009 8:59 AM PDT up reply actions 0 recs

At least one guy is supporting us

Hell, Hollinger had us in the playoffs before the draft order.

And isn’t Mason a so called "clipper " fan? I never bought he was a clipper fan.

by Qlippers on Jun 19, 2009 10:28 AM PDT up reply actions 0 recs

My Open Letter to Hollinger

Posted it on my blog

Dear John,

We are now in year three of your Draft Rater system. I believe that the Draft Rater could be a useful reference point but the way that you overhype it does the formula (and your own reputation) a disservice. It’s especially surprising that after the system’s 2008 performance, this year’s articles come off as more self-congratulatory, misleading, and disingenuous than ever.

The first issue is one of language. When you write about past drafts, you describe the Draft Rater’s performance in the past tense, as if the current system was actually in place at those times. For instance, you wrote “Additionally, the Draft Rater has come up with some big draft steals” or “2002 was a strong season for the Draft Rater, as it correctly recommended avoiding four players (Wagner, Ely, Haislip and Jones) selected among the top 10 collegians and made only one bad recommendation (Logan) in their stead.”

Obviously, these types of comments are misleading because the Draft Rater wasn’t introduced until 2007. The system didn’t actually “come up” with anything nor had it “correctly recommended” anyone. In fact, the complete opposite is true and that’s what makes your article so disappointing. Were these steals and busts discovered by your Draft Rater or were they actually the very information that you used to formulate your system? You claim that The Draft Rater predicted Carlos Boozer but it would seem that the truth is that Carlos Boozer helped create the Draft Rater.

This is the key problem that you fail to address in any of your articles. While the Draft Rater is great when it has the benefit of hindsight, it doesn’t fare so well in the actual moment.

If someone in 2007 had consulted the draft rater, they would have gotten a top 10 (in order) of: Kevin Durant, Greg Oden, Mike Conley Jr., Thaddeus Young, Brandan Wright, Al Horford, Nick Fazekas, Josh McRoberts, Rodney Stuckey, and Jared Dudley. While I’m sure someone might have thanked you for Stuckey over Law and Young over Brewer, you’d have some serious explaining to do for Fazekas and McRoberts. The Rater successfully pegged Big Baby Davis as a steal but, then again, it had him ahead of Jeff Green, who was about even with Kyle Visser and Herbert Hill.

The only other example we have of the Draft Rater in action was last year and even you had to acknowledge what a mess that was. However, blaming the one-and-dones doesn’t really help since it’s not like the system did a bang up job with the other collegians. Darrell Arthur and Mareese Speights over Brook Lopez. Darnell Jackson and Richard Hendrix over Jason Thompson. Ryan Anderson was considered almost undraftworthy. And that’s just the big men. Jamont Gordon was your sleeper (better than Westbrook said the numbers) and Courtney Lee was a “fringe second rounder”. Probably the biggest sleeper of the draft, the undrafted Anthony Morrow, didn’t get any help from the Draft Rater either; he wasn’t mentioned at all in your article.

This isn’t the only sin of omission that your article commits though. You constantly highlight the benefits and ignore the mistakes of the system. You commend your current system for passing over Patrick O’Bryant, Randy Foye, and JJ Redick but they were replaced with Shawne Williams, Hilton Armstrong, and Ronnie Brewer. Even if you cop a plea for Williams’ attitude, that’s still not impressive, especially since your current system is working with three years of hindsight and actual knowledge of how these guys performed.

Also, you commend your system for avoiding the likes of Patrick O’Bryant but it took you two years to get the system to be able to do that. He was #6 in 2007 and last year, the Draft Rater still had O’Bryant as having the third most potential of any big man in the draft (#2 was Shelden Williams and #4 was LaMarcus Aldridge).

The bottom line is that the system is just as flawed as the current results. For every Rajon Rondo, there’s a Marcus Williams. You might spare a team from drafting Marcus Haislip but they’ll end up with Casey Jacobsen instead.

I’m not sure if ESPN pushes you to make such bold pronouncements or if you really just have a blind faith in your system that causes you to miss all the problems but the fact of the matter is that the Draft Rater is an interesting reference but it’s hardly as helpful as tool as you like to paint it. You’re already copping pleas for Ty Lawson (citing work ethic and injury issues) but how could anyone really doubt a player with the fifth highest projection since 2002? The only guys with higher marks were: Durant, Wade, ’Melo… and Mike Conley Jr.

So I guess you can take it for what it’s worth and what it’s worth isn’t as much as you seem to think

My Open Letter to John Hollinger @ www.soulhonky.com

by SoulHonky on Jun 18, 2009 10:52 PM PDT reply actions 3 recs

Very eloquently put.

Stuck in limbo.

by PaperClip on Jun 18, 2009 10:57 PM PDT up reply actions 0 recs

And another reason why

I don’t listen to John Hollinger. Seriously lose the stats and start watching some actual games for once. The way to see a players potential is through watching them play and looking at strengths and weaknesses not looking at the box score, crunching some numbers and then making proclamations.

In Gordon we trust

by bestclipfan on Jun 18, 2009 11:22 PM PDT up reply actions 0 recs

He's committed a big statistical sin by succumbing to

Look ahead bias. The system wasn’t in place and events had not taken place back in 2002, the year he applied this system to. You’re correct in that it was probably developed based on those players stats and that’s why they look so good in Draft Rater.

What is Hollinger’s background? Did he even get any sort of degree in stats? It doesn’t seem apparent and if not, why is he so about the numbers?

by Newtybar on Jun 19, 2009 9:06 AM PDT up reply actions 0 recs

I'm with you

the thing tends to be all over the place, and he always has an excuse for the Michael Sweetney’s of the system.

Matthew 6:27

by mikey p on Jun 19, 2009 9:24 AM PDT up reply actions 0 recs

Great post on Hollinger

I was suspicious of his “system” when he he introduced it in 2007. One of my main points was, “Hey, let’s see how it does AFTER you’ve introduced it.” Answer: not so great. You’re post makes an excellent follow up to my 2007 post.

In this world, you must be oh so smart or oh so pleasant. Well, for years I was smart. I recommend pleasant. - Elwood P. Dowd

by Steve Perrin on Jun 19, 2009 10:13 AM PDT up reply actions 0 recs