With four new projected starters and seven new players overall, this Detroit team has a whole new look. I took a look at our favorite advanced stats, WS, WP, and PER Differential, to see if we can predict a record for this starting group. To make these estimations I had to first standardize the bench. For this I kept the same bench projections as was the production last year. There is a catch here in that Drummond had a serious impact on the bench stats from last year, however since he only played ~20 mpg I feel as though Billups at 25 mpg, Singler going to the bench, and Datome to compensate and even it out or at least make it close. Next we have to project Drummond as a starter. For this I used a combination of Joakim Noah and Tyson Chandler. I think Dre is going to fall somewhere in between there production. Finally we had to project KCP at SG, whom is the starter as of today. At first I was going to use the averages of the last 6 SG taken in the top 10 from their rookie seasons, Mayo, DeRozen, Harden, Turner, Thompson, and Beal, but after looking everything over Beal is almost exactly the average of these players. So for KCP's production at SG we are going to use Beal production from last season for our projection.
I started with Win Shares, but modified it to Win Share %(WS/Team Wins) to compensate for the varying wins each teams have had over the course of the past seasons. I used the average for the last 5 years or if shorter I used their career numbers.
Last Year Win Share %:
Projected Win Share %:
Drummond: 15.1% - 20.0%
The increase from 54.5% to 76/81% is 1.40% to 1.49% improvement good for a projected win range from 41-43 wins for this year from the 29 wins the lineup produced last year.
Next I took a look at Wins Produced.
Last Year Wins Produced:
Projected Wins Produced:
The projections are using Smith and Monroe's career averages, and Noah's WP from last year for Drummond, even though Dre really outproduced him on a per minute basis. If we add in the benches ~11.5 WP's from last year we get a projected win total of 45.3.
Last we are looking at PER Differential.
Last Year PER Differential:
Projected PER Differential:
This was a little tricky, because if I used just the starters I got a win projection of 57.8! Instead I used the team numbers for each position from their former teams, using the Knicks for the C projection. Considering that the average wins for a playoff team was 50.625, and the average PER Differential was 7.025, Detroits 6.0 yields a win projection of 43.2 wins.
So after running the advanced numbers three different ways, I get win predictions of 43, 45, and 43. Eerily similar to each other, a win total in this range has been good for a 5-6 seed in the East the last few years.