2015 draft discussion

[cragganmor]

statistical models are also guesswork until they are proven accurate. would be interesting to see what metrics they use and what weights they assign to each. one method that 538 used a lot during the '12 presidential elections was a poll of polls; nate silver also gave a higher weight to each poll that was shown to be historically more accurate. in the end, he called every state right on the nose.

[n8 the gr8]

statistical models are also guesswork until they are proven accurate. would be interesting to see what metrics they use and what weights they assign to each. one method that 538 used a lot during the '12 presidential elections was a poll of polls; nate silver also gave a higher weight to each poll that was shown to be historically more accurate. in the end, he called every state right on the nose.

That's an interesting example. I'm wondering if they're assigned like a modifier based on the conference according to how competitive they're supposed to be then given a standard deviation based on how many players exceed expectations coming out of the conference historically or something like that. Also curious how they calculated Mudiay in comparison to the guys that went to college in the US.

Either way something like that has to be supplementary to the eye-test, but I'm curious to see how it works.

[spree#8]

Apropos statistical models. Here an article from Kevin Pelton (ESPN) how his WARP (Wins above Replacement) ranks this years prospects and some historical comparison:

http://espn.go.com/nba/story/_/id/13092 ... hony-towns

Towns, Russell and Porzingis have projections above the average number one pick, Mudiay is right on that average and Okafor slightly lower. Another indication that we are speaking about a highly skilled draft class at the top.

About the model with the superstar and bust projections is talked a little bit here (note: that model only got used on College stats, so only on College and not on international players):

http://espn.go.com/blog/nba/post/_/id/4 ... lo-russell

And an in detail look at one of the models mentioned there (Statistical Plus-minus) here: https://bballhistory.wordpress.com/stat ... plusminus/

[Don Che]

So based on this....who's best at 4?

[spree#8]

Depends on who is available. Some of the draft nerds think that Towns, Okafor, Russell, Mudiay and Porzingis are five truly outstanding prospects all worth of the 1st pick and the WARP model gives them a statistical backing. So I would say you can't really go wrong if you take one of them.

Interestingly two guys I would be OK with if we trade down (Winslow and Johnson) also get backed up as pretty safe picks by the other model. Guys I am not that high on like WCS aren't backed at all by stats, so taking a guy like them is all about eye test and seeing things that don't manifest in a boxscore. That's all nice, but I think taking such a guy is far more risky than betting on one of the top prospects reaching his potential.

Their is also a great column every year by Chad Ford that I'm still waiting for: his draft tier piece. For that he conducts a survey amongst NBA personnel where they rank incoming prospects in a tier-system. Always a very good read, I will check whether this year one of those comes out.

[spree#8]

I asked for it and here it is:
http://www.espn.co.uk/basketball/draft2 ... ects-tiers

An excerpt (there are more tiers than the 4 I quote):

Consensus is a word that should be used carefully when you talk about the NBA draft.

Occasionally a prospect emerges who is so clearly better than anyone else in the draft (see Anthony Davis in 2012, John Wall in 2010, Blake Griffin in 2009, LeBron James in 2003, Yao Ming in 2002, Tim Duncan in 1997, Shaquille O'Neal in 1992), but more often, if you put 10 GMs into a room and ask them who the best player in the draft is, you'll get three or four answers.

The debates get louder and more diverse as you move further down the draft.

This year is no exception. With less than a week to go before the draft, the Minnesota Timberwolves are debating whom to take at No. 1 -- Karl-Anthony Towns, Jahlil Okafor or D'Angelo Russell. If the Wolves, who employ dozens of full-time pros who work at this all year, can't figure it out, how can we create a consensus ranking?

We want to believe that there's a Big Board in the sky that knows all. It doesn't exist. Reasonable minds can differ on prospects, and as much as we all love ranking players 1-100, it's not the best or most preferred way to do it.

This year, five players have a case to be picked No. 1 -- Towns, Okafor, Russell, Emmanuel Mudiay and Kristaps Porzingis.

Whom do you choose to draft? And how?

NBA teams watch prospects play thousands of hours of games. They go to practice and camps. Hire guys from MIT to create statistical solutions. Work out players, give them psychological tests, do background checks and conduct personal interviews. And, still, there is very little consensus.

Factor in the debate between taking the best player available and which player a team needs most, and the situation further muddies itself.

To make sense of all this, the past few years I've chronicled a draft ranking system called the tier system, which several teams employ.

By this method, teams group players into tiers based on overall talent, then rank the players in each tier based on team need. A more detailed explanation of how the system works can be found here. http://www.espn.co.uk/basketball/draft2 ... ystem-used

So how do things break down? After I talked to several GMs and scouts whose teams employ this system, here is how the tiers look this year.

Players are listed alphabetically in each tier.

Tier 1

Karl-Anthony Towns, F/C, Kentucky


Ahead of last year's draft, we had three players in this category: Andrew Wiggins, Joel Embiid and Jabari Parker. This category usually is reserved for guys who are surefire All-Stars/franchise players. Just three other players since we started this column in 2009 -- Griffin, Wall and Davis -- have been ranked in this slot. You have to be elite to get here.

This year was one of the toughest yet. Five players were nominated for Tier 1 by various teams. However, only one player was consistently mentioned by all of them: Towns. Although some teams don't see major separation between him and the players in Tier 2, others feel that there's a gap and that he's the only one in the draft who has the potential to be a top-10 player in the league.

Tier 2

Emmanuel Mudiay, PG, Congo

Jahlil Okafor, C, Duke

Kristaps Porzingis, F/C, Latvia

D'Angelo Russell, G, Ohio State

This is a very strong Tier 2. Mudiay, Okafor, Porzingis and Russell all have a claim to be the No. 1 pick. In fact, Okafor was the favorite to be the No. 1 pick since July, and several teams put Russell on par with Towns. Both received votes for Tier 1 but not enough to move them up.

Tier 2 is reserved for players with All-Star potential. However, each player on the list has a weakness that some teams feel will keep them from being a superstar. For Mudiay, it's shooting. For Okafor and Russell, it's athleticism and defense. For Porzingis, it's a thin body and questionable position. Nevertheless, each of the four has a strength that should propel him to an All-Star Game in his career.

One fun note: Last year we wrote the column before Porzingis, then regarded as a Tier 4 prospect, had withdrawn from the draft. So he's jumped up two tiers.

Tier 3

Willie Cauley-Stein, F/C, Kentucky

Mario Hezonja, G/F, Croatia

Justise Winslow, G/F, Duke

This is a very strong Tier 3, as well. Hezonja and Winslow received Tier 2 votes from teams. However, both fell just short of the cut.

This tier usually is reserved for players who are projected as NBA starters in their careers. All three players should go in the top 10. Hezonja, especially, has some star appeal. Having eight players in the first three tiers makes for a very strong draft, but not quite as strong as the 2014 draft, which had 12 players in the first three tiers.

Tier 4

Devin Booker, SG, Kentucky

Sam Dekker, F, Wisconsin

Jerian Grant, G, Notre Dame

Stanley Johnson, G/F, Arizona

Frank Kaminsky, F/C, Wisconsin

Kevon Looney, F, UCLA

Trey Lyles, PF, Kentucky

Kelly Oubre, G/F, Kansas

Cameron Payne, PG, Murray State

Bobby Portis, PF, Arkanas

Myles Turner, F/C, Texas

Tier 4 typically is late-lottery to mid-first-round selections in a normal draft -- selections 10 through 20. This year, that tier is roughly 9-19 on our Big Board. Lots of talented players in this tier project to be starters or high-level rotation players.

Here, even the consensus in tiers starts to break down.

Guys really are all over the place. Booker, Kaminsky, Lyles and Turner each scored a few votes for Tier 3. Payne also picked up a Tier 3 vote. Still, the majority of teams had them ranked in Tier 4. Oubre, Grant and Looney had some Tier 5 votes, but most of their votes were in Tier 4.

[n8 the gr8]

Thanks for posting that and the other links spree, the bit on statistical models in your last post was a good read.

The gap perceived between Towns and the tier-2 guys is interesting. But at the end of the day the only team that matters is Minnesota and they could be one of the teams that perceives one of the other guys to be at that tier.

The part that you bolded: Hezonja and Winslow received votes to be 2nd tier leads me to believe that WCS did not and isn't going to be a top 5 pick (so the Knicks "interest" was a smoke screen). Thank god.

[spree#8]

Thanks for posting that and the other links spree, the bit on statistical models in your last post was a good read.

The gap perceived between Towns and the tier-2 guys is interesting. But at the end of the day the only team that matters is Minnesota and they could be one of the teams that perceives one of the other guys to be at that tier.

The part that you bolded: Hezonja and Winslow received votes to be 2nd tier leads me to believe that WCS did not and isn't going to be a top 5 pick (so the Knicks "interest" was a smoke screen). Thank god.

Let's hope that the Knicks use a tier-system.

BTW: here is the whole article about the college players. It's from a pretty cool site, if you are into stats analysis, I recommend you browse around there a little...

http://fivethirtyeight.com/features/pro ... aft-class/

[Don Che]

knicks management dont strike me as a group that uses analytics to heavily.

[spree#8]

Jonathan Givony @DraftExpress
Huge game for Mario Hezonja in the ACB Finals against Real Madrid: 18 points in 25 minutes, 5/6 for 3, 1/4 for 2, 5 rebounds. Real won.

Jonathan Givony @DraftExpress
Stanley Johnson reportedly refusing to work out for Charlotte, who are drafting 9th. Hoping to get picked by Detroit (#8) or Miami (#10).

[Don Che]

Can't wait till Stanley is taking names in the league

[DolanCurse2001]

Can't wait till Stanley is taking names in the league

I would not be surprised at all if he's an All-Star. He's my 6th bpa.

[BXGuy]

Johnson has bust written all over him imo. I need more convincing.

[dcapodic]

I don't know about choking, ND is just a good team. They have been all year. I am not sure why guys like Jerian Grant get so little follow when all he does is make his team win. They are a different squad when he has been on the floor with them through his years there. He could be one of those guys that comes out after being there 4 years and hits the NBA running and people say, "Wow, where did this guy come from"....

I REALLY did not think there was any way the knicks would wind up with Grant. I hope my original feelings about who drafts him are correct

[ISIAH_THOMAS]

so glad the heat passed on booker. Dude is going to be very good

[StevoStarks]

so glad the heat passed on booker. Dude is going to be very good

Lethal shooter.

[shakespeare]

KAT and Devin Booker were on the same roster in college and they didn't win it all?

Wow.