NBA Player Projections

Predicting the ’15-16 NBA season, based on WAR projections of all players:

Before start of the season, can be found here. Enjoy.

The following was posted June 25th, 2015:

Two things actually, draft rankings and career WAR projections of all possible draftees (from college), and the other link: 17 seasons of draft retrodictions, in which all college players drafted between 1998 to 2014 career WARs are projected, compared to actual career results, etc.

The following was posted June 26th, 2014:

Here’s the link to my draft day post – and, more importantly, NBA career WAR projections of all available draftees who project better than NBA replacement level player at some point in their career.

Later I will combine my projections with the current NBA players, but after I also finish my D League to NBA conversions.  I want to turn every stone I can outside of international players (with only international league data) – complete historical stats for international leagues are very hard to come by.  For now, you could open the draftee NBA career projection spreadsheet in the link above and the current NBA players career projections below, combine them, and sort to your heart’s desire.

The following was posted on May 30th, 2014:

This page will probably get added to quite a bit in the next number of days – as I’m wanting to add explanations of methodology and upload more files.

But for now – let’s get started by uploading what I have.  I’m hoping uploading an excel file will work, it’ll give you guys the ability to sort and re-sort to your heart’s desire however you’d like.

Using the last 35 years of my NBA regular season player ratings, I’ve created a model that projects future seasons of any NBA player.  Simply, I took all 35 seasons of player data (already adjusted for year, pace, and team quality) broken down into 14 statistical subsets and seen how they change with age.  I applied all I learned to the model.  It’s obviously not nearly as simple as that (especially optimized playing time predictions), but that’s the basics until I figure out how best to explain everything.  So, here’s the future projections – season by season – of every NBA player who has played enough the last 3 seasons to give me better than a 7% “confidence” level (still REAL low, anything below 20% is pretty iffy) in their future projections AND ALSO at some point in the future projects better than replacement player (greater than 80 HN/48).  291 players “qualified” – here you go:

2015&OnNBAonlyPlayerProjections

Obviously 2012 through 2014 are actual results & ratings breakdowns, the rest are projections (2015 & on). As soon as a player drops below “replacement player” (below 80 HN/48), I end their “career” in the projection (since they don’t accumulate any more WAR).

The WAR I show here is a real basic one (compared to my WAR on past – not projected –  seasons), how many wins over replacement for a guy with that per minute production and playing time on an assumed average 41 win team.

Here’s the results accumulated, with all players ranked by future total WAR (Wins Above Replacement):

2015&OnNBAonlyPlayerCareers

PLEASE, hit me up in the comments &/or on twitter if there are ANY questions – I will be adding a quite a bit of explanations and historical test study stuff in the near future.  At this moment I’m just in a time crunch to get the draft model done.  Suffice to say – I will be projecting college to NBA, and including the results in the next NBA player projection update.  Also – few higher recent (notably missing in these projections) draftees in ’13 & ’14 might actually project above replacement player when I add the weighted college to pro projections (Bennett, Porter, Caldwell-Pope, etc.) – plus playing time limiters may smooth some for the young guys (Olynyk, as an example, projects low future playing time to others comparatively because of high projected foul rates).  So, next update, I’ll have Jabari Parker, Embiid, Wiggins, et al (if they project above replacement level player in their career) included – as well as probably better projections (greater “confidence” with increased data) for ’13 & ’14 draftees.

2 thoughts on “NBA Player Projections

    1. You know, I think I’ll try to make sure I have a glossary sheet on every spreadsheet I post.

      %min: the optimal amount of minutes expected based on statistical skillsets and league history. 50% would be half of available minutes, 24 mpg for 82 games (if you ignored OT).
      HN/48: their hoopsnerd rating per minute, 100 is league average. The 12 columns after WAR are a breakdown of that rating into statistical skillsets, league average is at the top for each. If you resort each, I think they’ll make more sense if you’re curious. Sort Reb, top rebounders. Sort Def, best defenders statistically in terms of bks, stl, & pf. I color coded the related columns to try to make it simpler to understand.

      WAR is the projected Wins Above Replacement if the player actually played the optimal minutes at that HN/48 on an average team.

      Conf is my confidence level of the projection – the lower it is the higher the possibility the player’s actual future playing time and rating can veer from the projection.

Leave a Reply to Hoops Nerd Cancel reply

Your email address will not be published. Required fields are marked *