2016 Usage-adjusted Win Projections, Part 2

Oct 18, 2015; Portland, OR, USA; Portland Trail Blazers guard Damian Lillard (0) celebrates with C.J. McCollum (3) after making a free-throw to cut the Utah Jazz lead to one point during the fourth quarter of the NBA preseason game at Moda Center at the Rose Quarter. The Blazers won 116-111. Mandatory Credit: Godofredo Vasquez-USA TODAY Sports
Oct 18, 2015; Portland, OR, USA; Portland Trail Blazers guard Damian Lillard (0) celebrates with C.J. McCollum (3) after making a free-throw to cut the Utah Jazz lead to one point during the fourth quarter of the NBA preseason game at Moda Center at the Rose Quarter. The Blazers won 116-111. Mandatory Credit: Godofredo Vasquez-USA TODAY Sports /
facebooktwitterreddit
Mandatory Credit: Godofredo Vasquez-USA TODAY Sports
Mandatory Credit: Godofredo Vasquez-USA TODAY Sports /

For methodology, see Part 1.

Here’s the goods everyone — my 2016 Win Projections are below:


And here’s a couple handy charts to help explain why my projections look like they do:

“Top 10”: Blended rating above +5.0 per 100.
“Star”: Blended rating between +2.5 and +5.0
“Above Average”: Between 0.0 and +2.5
“Below Average”: Between -2.5 and 0.0
“Waste of $”: Below -2.5

So yes, there were hours and hours of adjustments and tuning, yet projections still look extremely close to Vegas’ numbers. All my win totals are within five of Pinnacle Sports’ betting lines, except for eight teams:

8. Charlotte Hornets:
Wins vs Vegas: +5.9
Key stat: Cody Zeller’s 2015 RPM was +4.35 per 100.
Key discrepancy: A disproportionate amount of the Hornets’ future success has been associated with Michael Kidd-Gilchrist. While he is arguably their best player (+3.2), my projections actually give that title goes to Cody Zeller (+3.4) by a hair. Also I have Big Al regressing a little bit upwards to the mean, Nic Batum being an absolute stud in Box Plus-Minus, and the Hornets having just three players below replacement level (11th best minutes distribution in the league).

7. Atlanta Hawks:
Wins vs Vegas: -6.1
Key stat: Tim Hardaway Jr and Dennis Schroder’s are fourth-last and fifth-last in my projected Wins Created (-1.9 and -1.8 respectively).
Key discrepancy: The Hawks project to give 24% of their precious minutes to those in the “Waste of $” category (aka “Dumpster Fires” aka “Below -2.5 per 100”). But RPM doesn’t know Mike Budenholzer exists.

6. Los Angeles Lakers:
Wins vs Vegas: -6.2
Key stat: The Lakers are first in expected minutes distributed to those in the “Below Average” category, i.e. between -2.5 per 100 and 0.
Key discrepancy: Bettors tend to overrate the Lakers and Kobe Bryant, so it’s no big surprise that my projections see them more lowly. Their young talent look promising, but as I’ve written before – the vast majority of rookies and sophomores don’t contribute positively. Also, the only Laker I project to add net value against a 0.500 lineup is Roy Hibbert (+0.72 per 100).

5. Milwaukee Bucks:
Wins vs Vegas: -6.4
Key stat: Ex-Bucks Zaza Pachulia and Jared Dudley’s projected Wins created are +5.6 and +4.0 respectively.
Key discrepancy: Many fans think that the Bucks haven’t lost much and are on the upswing. While their young players are exciting and are improving, losing Zaza and Dudley makes RPM very skeptical over 82 games.  And, though I love him, John Henson’s plus-minus stats were miserable in 2014 and 2015. This is a point of contention for box score stats and plus-minus stats. In fact, his RPM projection disagrees with my BPM projection more than any player in the league (except Pekovic):

buckssucks
buckssucks /


4. Brooklyn Nets:
Wins vs Vegas: -7.2
Key stat: Brooklyn projects to give more of their available minutes to below-replacement level than any other team, at a ridiculous 47%.
Key discrepancy: Joe Johnson and Thaddeus Young project to be just good, not great. Brook Lopez has also recently up and down and injury-prone so he’s about average as well. Finally and sadly, Jarrett Jack is not a plus-minus darling. He projects to be one of their truly terrible players, at -3.4 per 100.


Warning: approaching “WHAT THE?!?” territory. Here are the 3 strangest results.

3. Boston Celtics:

Wins vs Vegas: +9.6

Key stats (Boston gets 2):

a) Boston projects to give a league-best 76% of minutes to players rating above zero.

b) By 2015 Real-Plus-Minus, Marcus Smart should have been rookie of the year, at +2.4 per 100.

Key discrepancy:

As I have been

opining

and

whining

about for over a month now,

Boston’s roster is absolutely filled with very good players who are not stars

. Whether or not they all fit together is of course a different story. But as a start, my usage projections actually like their fit, netting them an extra win or two. And as you probably know, Amir Johnson *is* a plus-minus darling. Here’s a brief overview of their roster:



2. Miami Heat:
Wins vs Vegas: -11.9
Key stat: Miami’s roster estimates to use a combined 112% of possessions – this collective 12% the Heat needs to shave off their usage is the most of any team.
Key discrepancy: First off: while Miami has some big names, they have big names of *old* players who haven’t really contributed in the last few years. All of their plus players are past their prime. The best per-100 player they have in my projections is Deng, at +1.44. Secondly, most of their big names were good in the past because they were able use a lot of possessions, which won’t be happening as much this year since they’ll be sharing floor time. Specifically, Wade, Amar’e and Bosh:

heatusagedrop
heatusagedrop /

1. Portland Trailblazers:
Wins vs Vegas: +16.4
Key stats (Portland gets 4):
a) Portland’s combined projected usage is just 93%, the second-lowest in the league.
b) Damian Lillard’s Box-Plus-Minus the last three seasons was: +0.3, +3.5, and +5.2, and he just turned 25. My combined projection for him this season is +4.54, 14th-best in the league.
c) C.J. McCollum and Al-Farouq Aminu project to contribute more than +2 points per 100 this season.
d) Portland projects to give just 5% of their available minutes to below-replacement players (5th best in the league).
Key discrepancy: Ok, please don’t leave the Nylon Calculus website never to return. This one’s the weirdest I know, but stick with me. Yes, Wes Matthews was good. LaMarcus Aldridge was very good – and I’m not sure my Box Plus-Minus projections entirely capture how much he contributed to Lillard’s rating. But one thing is for sure – Lillard is a high-usage point guard who both puts up stats and doesn’t have a bad defensive +/-.

To this end, my projections see him taking a 2015-Russell Westbrook type role this year: doing it all and filling a big void.