Saric rookie comparisons

Home for all your discussion of basketball statistical analysis.
Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 8:08 am

http://bkref.com/tiny/5KDpc

Using a few metrics & stats, Saric as a rookie for season is better than Jeff Green and in between Donyell Marshall and Shareef Abdur-Rahim. Drew Gooden is a neighbor too. Other criteria could find other similars.

Using just second half of season would make Saric look much better. On offensive rating his first half to second half of season improvement is much bigger than the others who also generally improved. His O Rtg in second half is the slight leader in group, but below league average. His first half O Rtg was neck n neck with JR Reid for worst in group. His defensive rating was in better half of group but was still below league average and got slightly worse in second half.

He is best or next best on at least half dozen stats. Surprisingly on the shooting percentages.


Other team stats for Saric:

7th best raw plus minus of main rotation guys. Team wins more minutes with him in SF matchups. Only winning 31% of his estimated time at PF, below overall 38% team win rate.

82games says he takes 5 more shots than his PF matchup opponents to score 4 more points per 48 minutes. Leads on rebounding & assists too. Fwiw though, PER thinks the opponent wins the overall production matchup slightly.

You can be low ranked by metrics as a rookie and get better or even very good. There are positive signs for Saric based on this data. But every player is a unique case and one year or either half doesn't necessarily predict what is coming next.

Team results with Saric and Embid on court together are a terrific plus 7.4, mostly due to Embid but good is good. All his other player pairs are negative, some greatly (partly influenced by not Embid time). They should change out most of the rest of the team.

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 2:14 pm

I've been playing around with sim scores, along with a combo of WS and BPM. Basically everything that is on Basketball Reference. There are a lot of holes here, but for a quick pass, the "comps" for most guys look solid-ish. I'm using z-scores for a bunch of stats. No real rhyme or reason yet. Here's what I have for Saric.

Image

Most of the above should be self explanatory, but I'll touch on the Wins/48 and Talent items. Wins/48 is just for this season. Talent is going to be 3 years generically regressed. Talent+1 and Talent+2 are simply the Talent ratings, but for the next two years. Trying to get a visual on growth.

I assume that everything I am doing here is wrong, because I don't really have much of a stats background. I just like messing around in Excel.

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 4:12 pm

I am very interested in sim systems though I am fairly cautious using the results. Would be interested in hearing more, seeing results for more guys. Are you willing to show or link to the dataset, formulas in more detail? For talent when you say "generically regressed" do you mean no weighting system for the various elements included in talent? If so, that has been my biggest beef with other similarity systems.

BRef could imo use a new sim approach. Or a new public rival (to it and Pelton's Schoene).


Guggs is an interesting "realistic" comp. Pippen even more so. I can see the first far easier. Morris might be the floor.

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 5:41 pm

I'm open for anything. All I did was grab the yearly advanced stats from Basketball Reference for each year since 1985. I use an average of the WS/BPM number to give me a generic wins metric for a season. Talent is a wins/48 minute rating, over the past 3 years. I think I did 5/3/1 for yearly weights, but I'd have to look. Either way, it's somewhere around there. Very arbitrary stuff.

The z-scores are just for that particular season's stats, but I was thinking of applying the same process I did for talent. Sim Score is sqrt of the sum of the z-score errors. I have no idea if that's what I should be doing or not, but this seems to be passing the smell test. Here are a couple other guys. Happy to discuss this further, make tweaks, etc. You all have way more knowledge and experience. I'm just an Excel nerd. :D

I can also take people from the past. So if we wanted say, Derrick Favors after his rookie year, I can do that as well.

Quick note on the numbers above the pivot table. They are just averages of the info below, so average age for the comps, average talent, talent+1, talent+2. I'm only showing top 20 in the table, but technically it compares to every person from 1985-today.

Devin Booker
Image

C.J. McCollum
Image

Gordon Hayward
Image

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 5:56 pm

Booker's top comp list is pretty light on great names but there are a few. I am more skeptical of him than many others but I sometimes lag on giving big praise to big potential guys. I accept possibility for big improvement but I am not quick to EXPECT it and it isn't here til it is.

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 6:01 pm

Crow wrote:Booker's top comp list is pretty light on great names but there are a few. I am more skeptical of him than many others but I sometimes lag on giving big praise to big potential guys. I accept possibility for big improvement but I am not quick to EXPECT it and it isn't here til it is.
I thought it was interesting to see the growth of the avg. talent rates. Dion Waiters minus the crazy is interesting.

If there's anyone specific you want me to post, let me know.

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 6:10 pm

Porzingis
Image

Paul Millsap
Image

Jahlil Okafor
Image

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 6:16 pm

Zingis has a pretty weak imo comp list. I haven't seen enough from him on average to justify the current hype or mega expectations. He might get there but the evidence for it so far isn't that strong.
Being used heavily does not necessarily mean you are great (yet).

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 6:20 pm

Crow wrote:Zingis has a pretty weak imo comp list. I haven't seen enough from him on average to justify the current hype or mega expectations. He might get there but the evidence for it so far isn't that strong.
I see it a bit differently. Tobias, KVH, Gay, Deng. All solid guys. KP has comparable stat lines to them, but he happens to be 7 feet tall as well. I have no idea what that means, but I think it's an interesting caveat.

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 6:24 pm

I'd be interested I seeing some others. Oladipo, S. Adams, John Wall, Lowry, Yogi Ferrell, Seth Curry, Wiggins...

I'll stop there for now.

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 6:29 pm

The four Zingis comps are "solid".. for #2,3 or 4 guys on a good team but the Zingis hype is unicorn level. Yes size is an important variable. Some ignore it to look at positional-less role or talent but it matters in role & position and these things are not now completely unimportant. Maybe not sufficient, maybe constrain understanding but not dead considerations to me.

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 6:34 pm

Oladipo
Image

Steven Adams
Image

John Wall
Image

Kyle Lowry
Image

Yogi Ferrell
Image

Seth Curry
Image

Andrew Wiggins
Image

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 6:52 pm

Thank you.

Steven Adams with high comps to Tarik Black and C Fellacio? Could be signs of a massive overpay.

Ferrell has Billups, Conley, Jet Smith... and a lot of meh backups.

Seth Curry with high comp to Parsons? More damning of Parsons but not a name I wanted to see.

Most of Wiggins' top comps are forwards. B-Ref suggests Thibs has moved him to SF pretty much exclusively. Probably the easiest role for him, given his talents and non-talents and using conventional thinking of positions / roles. May not be that good though. Team wins 46% of his time at SF according to 82games. Barely wins matchup according to PER. Last season dominated matchup at SG and team won 50% compared to 35% overall (and much less without him there).

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

Re: Saric rookie comparisons

Post by JoshEngleman » Tue Mar 21, 2017 7:05 pm

Crow wrote:Thank you.

Steven Adams with high comps to Tarik Black and C Fellacio? Could be signs of a massive overpay.

Ferrell has Billups, Conley, Jet Smith... and a lot of meh backups.

Seth Curry with high comp to Parsons? More damning of Parsons but not a name I wanted to see.
The Adams contract seems like it will end up being a problem. Not that this is any sort of accurate process, but I did a little back of the envelope $ calc. Using $4m per win, he would be more like a $15m a year guy next year instead of the $22m he'll make. You don't really want to start the contract off with negative value, so, eek.

The Ferrell comps are interesting. I'm sure he wouldn't mind having Augustin's career.

Is Parsons really a bad comp? Seth Curry being Parsons is amazing, so long as it doesn't include the injuries.

I like the combo of Larry Hughes/DeRozan for Wiggins, but I don't know if that is a good thing or a bad thing.

Are there any ways to hide these graphics in spoiler tags or something, so I don't just take up this whole thread? Should I switch them out for text tables? Any thoughts for changes to this process? Different stats? Weights? I'm open to any feedback.

Crow
Posts: 6246
Joined: Thu Apr 14, 2011 11:10 pm

Re: Saric rookie comparisons

Post by Crow » Tue Mar 21, 2017 7:14 pm

Using space isn't an issue to me. Immediately fully visible is easy.

Curry having Parsons as a comp isn't "bad" but I consider Parsons as something of an empty calorie scorer. Not really empty, just less diverse a contributor than you might expect / want.

Cuban might have gotten too excited in print for Ferrell. At that salary it is fine but still more likely a minor addition than anything that special.

On weights, I'd weight shooting / scoring efficiency higher than other stuff (probably 3x the standard weight?) especially for perimeters.
Assists matter more with perimeters and rebounds for bigs... in general. Could weight differently (twice standard depending on role?) or just let position-less and weightless(?) approach rip. I'd add RPM or especially DRPM to try to capture the unseen in box score stuff including shot defense. Or add fg% allowed overall or at rim.

Post Reply