2017 draft analytics

Home for all your discussion of basketball statistical analysis.
Crow
Posts: 5955
Joined: Thu Apr 14, 2011 11:10 pm

2017 draft analytics

Post by Crow » Thu Mar 23, 2017 10:58 pm

If you have ratings or discussion on the topic and want to get started on displaying it, have at it here (or your own thread if you prefer). I know we've had a little earlier.

Opportunity to provide analysis to DraftExpress again would seem possible or likely (?) if you are inclined.

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Fri Mar 24, 2017 3:04 am

My question is does anyone have a projection system that actually likes Fultz over Ball? I'm yet to see one, and I know Ball is helped by playing on a wildly better team, but I just don't see how Fultz is #1 currently on most boards above Ball. He's obviously very good, he passes the eye test and I think he's the 2nd best prospect in the draft, just not sure why Ball isn't the favorite to go #1. Curious what other people think about this.

Crow
Posts: 5955
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2017 draft analytics

Post by Crow » Fri Mar 24, 2017 3:46 am

I barely know the names but at first glance I assume Fultz's higher usage (and points) gives folks more confidence he can carry a big load of scoring responsibility in NBA. Ball has good maybe even better stats except for usage (and points).

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Fri Mar 24, 2017 4:32 am

That's true. Maybe Fultz is seen as able to make any team good because he's a more traditional scoring PG, while Ball can make a good team great with his passing and thus is more useful in the right situation but perhaps not as useful in general.

To me, Ball seems like a shorter Ben Simmons who can legit shoot it.

Crow
Posts: 5955
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2017 draft analytics

Post by Crow » Fri Mar 24, 2017 4:47 am

I feel Simmons is more likely to disappoint than meet or exceed expectations. Far more likely, at the first few years.

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Fri Mar 24, 2017 5:11 am

Agreed, the amount he's likely to struggle to score is going to really turn casual fans off from him as a prospect if he isn't throwing multiple highlight passes every night.

Crow
Posts: 5955
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2017 draft analytics

Post by Crow » Fri Mar 24, 2017 5:38 am

I think it would be fair to say eyes / subjectively selective stat guys liked Simmons more than the comp systems? Yes or is my memory off?

If someone wanted to do more legwork on comparing the two approaches, they could maybe pit GMs & Draftexpress (by real and mock draft pick #s) against the system guys in DX analytic system projections year 1. Who looks better after almost 2 seasons? There was a second,analytics,article but it did not include any evaluation of outcomes. After one year, ok, probably too soon. After 2 maybe still but why not look? After 3 there really should be a detailed look. If you don't look, you might not learn as much or the right things. There is the issue of picking for need & fit vs. plain talent level but most insiders tend to say they usually try to pick the best available player.

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Fri Mar 24, 2017 7:01 am

I think most projection systems liked Simmons, but yeah scouts certainly liked him more than those systems. If I have time in the coming few days I'll see if I can dive into the scouts vs. system some more. After 2 years, we should at least see any clear misses/hits from one or the other, even if the overall picture is a bit blurry still.

Crow
Posts: 5955
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2017 draft analytics

Post by Crow » Fri Mar 24, 2017 5:31 pm

That would be great. Don't feel rushed.

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Sat Mar 25, 2017 1:31 am

Just getting started, but already seems like the blend of analytics models is doing the best by a hair. The correlation between actual rank, which is an average of Win Shares and VORP rankings, is 0.3802 for the analytic model average. For DraftExpress, it is 0.3768 which is close to how the analytics did, but just a tad bit worse. The GM's were not as close, with a correlation of 0.2626. I'll keep looking into this, not really sure why the GM's did so much worse by this view than the others.

Dr Positivity
Posts: 297
Joined: Thu Sep 20, 2012 6:44 pm

Re: 2017 draft analytics

Post by Dr Positivity » Sat Mar 25, 2017 3:02 am

jgoldstein34 wrote:My question is does anyone have a projection system that actually likes Fultz over Ball? I'm yet to see one, and I know Ball is helped by playing on a wildly better team, but I just don't see how Fultz is #1 currently on most boards above Ball. He's obviously very good, he passes the eye test and I think he's the 2nd best prospect in the draft, just not sure why Ball isn't the favorite to go #1. Curious what other people think about this.
Fultz has far more advanced ability to create his own shot in college. Ball took Morey-gasm levels of no midrange shots this year and averaged 3.1 FTA/40. Fultz averaged 26 pts/40 and 7.6 FTA/40 and seems to have a very isolation friendly game. He is considered to be a surefire 20 point a game scorer by scouts (not that it means he is, just ask Jahlil Okafor) which still holds a lot of power for teams especially one at the bottom of the standings frankly, while still being a good passing prospect. This is before considering Ball's weird jumpshot which could blow up in a team's face I guess.

I have Ball ahead for numbers reasons, but that's the logic behind it as far as I can see. Analytics users know that the ability to create your own offense in college isn't as predictive as other areas of the game, but that hasn't crossed over to the mainstream that much

EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Re: 2017 draft analytics

Post by EvanZ » Sat Mar 25, 2017 4:38 am

I see big red flags in both these guys FT%. I think Ball has serious shooting concerns at the next level because of his release and form, and Fultz 3P shooting in college is scary because his FT% is so low (65%). Very rare to see above-average point guard shooters with such low FT%'s. I think this draft is overrated.

jgoldstein34
Posts: 240
Joined: Fri Sep 30, 2016 6:38 pm

Re: 2017 draft analytics

Post by jgoldstein34 » Sat Mar 25, 2017 1:54 pm

I guess picking Fultz #1 is the safe pick because he is a more traditional prospect, and my system has him #2 so it's not like he's some bad prospect that doesn't project well.

As for the FT issues, I'm way more worried about it for Fultz than Ball. Ball shot 165/205 (80%) in his 2 high school years we have available:
http://www.maxpreps.com/athlete/lonzo-b ... -stats.htm.
Ball will have some trouble figuring out how to score in the NBA, but if you look at his high school 2p% and 3p% they're so similar to his college that I think he'll find a way to keep up his crazy morey shooting style in the NBA and do it well. His shot is gross, but it legit goes in. Between HS, non-FIBA events, and college he's 232/558 (42%) from 3P. It works. It's gross, but it works.

Fultz's high school stats aren't available online, at least not anywhere I could find, and all his FIBA FT shooting adds up to 9/14 (64%) so I'm more worried that he may just strangely be a bad FT shooter relative to his 3P%.

I don't think this draft is overrated, there's just a ton of guys this year who could legit be all-nba level players at some point in their career. That being said, this draft doesn't have much depth. It'll be great for the teams picking top 12, but outside of that it'll be worse than usual. Really top heavy draft, but a great top.

EvanZ
Posts: 912
Joined: Thu Apr 14, 2011 10:41 pm
Location: The City
Contact:

Re: 2017 draft analytics

Post by EvanZ » Sat Mar 25, 2017 4:50 pm

The problem with Ball is that he's so limited with how he can create his own shot from deep. NBA teams are going to scout the hell out of it, and right now, it looks like he's way too easy to defend out there.

JoshEngleman
Posts: 127
Joined: Thu Oct 23, 2014 12:13 pm
Location: Wilmington, NC
Contact:

Re: 2017 draft analytics

Post by JoshEngleman » Sat Mar 25, 2017 6:54 pm

Question for anyone here that might be more familiar. I was trying to get data from National Statistical, but for some reason, their sign-up page is blank. They aren't active on Twitter, but it's clear from the homepage that the data is still being updated daily. Does anyone know what their deal is, or if there's another data source out there in a similar price range?

Post Reply