No announcement yet.

2016-17 Raptors Win Projection by the Numbers

  • Filter
  • Time
  • Show
Clear All
new posts

  • 2016-17 Raptors Win Projection by the Numbers

    I usually try to avoid posting my own stuff on here, but it's the long dreaded off-season so I thought some content would be useful. I've put together a projection for next season based on WS and BPM from basketball-reference. It's similar to what I did last year with RPM but I think a little more thorough and using two different win approximations gives a nice range of likely records.

    Obviously appreciate it if you click through (there's a lot of background and a little bit of proofing of the method in there) but I'll post the projections and a decent sample of the text here, enough I think to get what I'm going for here.

    What do the Numbers Say? Projecting the Raptors' Win Total for the 2016-17 Season

    Basketball Reference captures an incredible range of basketball statistics, ranging from box scores to shooting percentages from different distances, to more advanced offerings of all sorts. But they are likely most widely known for their two popular catch-all statistics - win shares, a box score aggregator of sorts, that mostly measures production, and box plus minus (BPM), an attempt to measure a player's impact.

    Win shares are easy to translate into wins - just add up the win shares for each player and the team total should be how many wins the team achieves. BPM can be translated into a wins over replacement player (WORP) value. Add up the WORP for each player on the team and you'll get how many wins better than a replacement team you'd expect. A replacement team, by my calculations (and based on the definition of "replacement player" by basketball-reference), is expected to earn about 15.5 wins.
    2015-16 stats:

    ... taking the sum of the WS generated, we get a total of 54.5 wins, 1.5 wins below the actual record of 56-26 last season. Doing the same for WORP results in 41.0 WORP, or 56.5 wins. In this case WORP measured the season very accurately, within a half win of the actual total, while WS again showed roughly a 1.5 win error. For both seasons, an error of roughly 2 wins either way seems to be about the level of accuracy provided.
    Here is the method I've come up with to remove myself (and my biases and assumptions) from the projections as much as possible.

    1) Compile the current roster of players on the team.

    2) From the previous season, take total minutes played, WS/48 and WORP/48 for each player.

    3) Assume all players will play the same minutes they did the previous season.

    4) Sum the total minutes played for the team. It should (but won't) be 19680 (48 minutes times 5 positions times 82 games).

    5) Scale all players' minutes by the ratio of total team minutes to 19680. For example, if the sum from step 4 is only 17890 minutes played, all players will see their minutes from the prior season increase by 10% (since 17890 x 1.1 = 19680).

    6) The new sum of adjusted player minutes will equal 19680.

    7) Use the new minute assumptions for each player and multiply by their WS/48 and WORP/48 values to get total WS and WORP numbers.

    8) Sum the WS and WORP numbers for the entire team to get the projected win total.

    This process has no assumptions about roles, minutes, usage or fit. Its only assumption is that the best indicator of future role, minutes, etc, is the previous season. There are certainly cases where this is not likely. For example, perhaps Norman Powell or Lucas Nogueira will get a larger role this season than last. But perhaps not. Perhaps Lowry will see his minutes reduced. But perhaps not. In any case, this is the method I've chosen to remove myself from the process as much as possible.

    Once again, like last year's projection, we have a minutes deficit, so once again players get a minutes boost. Right away we can say this is unlikely, so a little further down I'll allow myself to modify some stuff to avoid obvious problems like Lowry playing 3300 minutes (40 MPG).
    Let's make those reasonable changes though. Lowry will almost certainly not play more minutes than he did last season, and nor will DeRozan. Meanwhile, we'd expect many more minutes for Carroll and Valanciunas due to lack of injury.
    All told, the team total slides down to 54.7 wins (by WS) and 60.9 wins (by WORP). While having a minute distribution that is feasible.

    These are still very nice projections. Considering that both WORP and WS have seemed limited to small errors (roughly two wins), both in the after-the-fact measurements and the sample projection we looked at, the simplest way to minimize the error for both measures is to settle nicely in the middle. Which means the final projected record for the team as constructed, with as few assumptions as possible on my part, is 58-24.

  • #2
    ^^ Dan, you're a beast.

    I dug up your last season's projection (ass that I am). Just wondering: did you change anything from your model/assumptions that you used last year?

    DanH wrote: View Post
    Hi all,

    With most of the big moves of free agency having settled down, and even a lot of role players off the market with a few notable exceptions, I thought I'd take a look at how the season projects out.

    I'm using RPM here, ESPN's player impact statistic, and then using net points as a predictor for winning percentage using Pythagorean wins.

    Quick walk through: first, I took the RPM value of every player in the league from last season. Didn't adjust them in any way, didn't project improvement or decay. Just last year's value.

    Then I placed all the players on their current teams based on the off-season so far, and all the rumours I could track. Any players still unsigned I've left off completely for now. Most of them are either low impact or low minutes players so they shouldn't skew this too much.

    I made no adjustments to minutes, or guesses on how playing time will change. I simply listed the players and their MPG from last season for each team. I tried using total minutes, but it became too difficult to adjust for players who missed large chunks of last season. By using MPG that fixes most of those issues (though not all, I'll touch on this later).

    Then, having compiled all the players, their RPM and their minutes based entirely on last season, I created a MPG-weighted average RPM for every player on every team. This creates an automatic adjustment - I'm not taking an approach like WS and simply summing the WS of new players, as if their minutes need to be adjusted those numbers will obviously change. This approach simply assumes that if a team, based on their total MPG from last season, has too few minutes to go around, every player will have their minutes adjusted downwards proportionally. Probably not true, but there aren't any instances where this effect is extreme, so I think this is a reasonable approach.

    This yields, for each team, an average RPM (I actually did dRPM and oRPM separately just for additional information). Those RPM values are then multiplied by 5 (as at any time, a team will effectively have 5 of their average player on the floor impacting the game). That value is added (or subtracted, for dRPM) from the league average ORTG last year (103) to give a predicted ORTG and DRTG for the team.

    Then using the Pythagorean approach, win percentage is calculated (I used a power of 14.5, though 16 has been used as well). Multiply by 82 and you've got your number of wins.

    Here is what the Raptors looked like, for example.

    Player..................Team | MPG | ORPM | DRPM | O-impact | D-impact
    DeMarre Carroll......TOR.....31.3......0.96 -0.25.........30.05 -7.83
    Bismack Biyombo....TOR.....19.4.....-2.58 1.59........-50.05 30.85
    Luis Scola.............TOR.....20.5.....-0.32 0.87..........-6.56 17.84
    Cory Joseph..........TOR.....18.3......0.04 0.96............0.73 17.57
    Kyle Lowry............TOR.....34.5......2.55 1.27..........87.98 43.82
    James Johnson.......TOR.....19.6......0.33 1.66...........6.47 32.54
    Patrick Patterson...TOR.....26.6......2.16 -1.82.........57.46 -48.41
    DeMar DeRozan......TOR.....35.0......0.21 -0.36..........7.35 -12.60
    Jonas Valanciunas...TOR.....26.2.....-1.72 1.05........-45.06 27.51
    Terrence Ross........TOR.....25.5......1.75 -3.77........44.63 -96.14
    Lucas Nogueira.......TOR.....3.8......-0.92 -0.34........-3.50 -1.29
    Bruno Caboclo........TOR.....2.9......-1.43 -1.63........-4.15 -4.73

    Total............................263.6............ ...............125.335 -0.881
    Per player average RPM................0.475 -0.003
    Net impact per 5 players...............2.377 -0.017
    ORTG and DRTG..........................105.4 103.0
    Win % 58%
    Wins 47.7

    (Sorry about the formatting there just can't get the table embedding to work).

    So, I did that for every team. And ended up with these rankings.

    Place |East |ORTG |DRTG |Wins
    1 ATL 106.4 98.9 60.8
    2 CLE 107.7 101.6 57.5
    3 CHI 104.8 101.5 50.3
    4 TOR 105.4 103.0 47.7
    5 WAS 101.1 99.6 45.4
    6 BOS 104.1 102.7 45.2
    7 IND 103.2 104.0 38.7
    8 MIA 103.6 105.3 36.2
    9 DET 103.7 106.5 33.4
    10 CHA 100.0 102.6 33.4
    11 MIL 97.7 102.2 28.0
    12 ORL 100.5 107.0 23.6
    13 BKN 99.3 107.2 20.2
    14 NY 100.2 109.9 16.9
    15 PHI 94.6 106.6 12.3

    Pos West ORTG DRTG Wins
    1 GS 108.9 97.4 68.5
    2 SA 108.1 97.8 66.5
    3 DAL 109.4 103.4 56.8
    4 MEM 105.1 101.0 52.4
    5 OKC 106.9 102.8 52.2
    6 LAC 107.7 103.6 52.2
    7 NO 106.5 103.6 49.2
    8 HOU 104.0 101.2 48.9
    9 SAC 103.8 102.0 46.4
    10 PHX 98.6 98.1 42.5
    11 UTA 101.7 102.4 39.0
    12 POR 101.1 101.9 38.6
    13 DEN 101.6 108.0 24.1
    14 LAL 100.9 110.4 17.5
    15 MIN 98.6 108.2 16.9

    Those results seem reasonable to me.

    Some corrections I thought to make (they are not implemented above, that's the raw data there). Paul George returning to Indy using his RPM and MPG from 2013-14 would mean a jump of 7 wins, up to 5th right behind TOR. Chris Bosh returning to his 2013-14 RPM and MPG would mean a 3 win jump for MIA, which doesn't move them up the standings unless you don't do the Paul George adjustment. Kevin Durant doing the same would add 4 wins to OKC's total, moving them up one spot ahead of MEM. I didn't adjust for Melo, as he had a reasonable number of games and minutes and had his usual impact last season.

    So, a couple promising things for Raptors fans: first, the Raptors project to have a top 10 offence again. And they project to improve to a league average defence. Hopefully they can make even larger strides based on improved systems, but it is heartening that purely on the strength of the personnel they are looking to make a 6-7 spot jump defensively. And it appears that the prediction of a struggling Raptors offence might be a little premature.

    Another very promising consideration: the Raptors own the less valuable of the Knicks and Nuggets first round picks this coming draft. And this system places those two teams 3rd and 7th in the lottery. A 7th overall pick would be a pretty nice return for Andrea Bargnani.

    I hope you liked this little early projection game. I tried to keep my own assumptions and biases completely out of it. Any questions are very welcome.


    • #3
      Yeah, completely different approach this year, probably simpler (using a direct translation from BPM to wins rather than point differentials) and cleaner, and using two different stats instead of one (and using BPM instead of RPM, which seems to have less variance year to year) gives a little more confidence.

      Last year the free agency market was pretty muted, but so much has changed this year that doing a full league's worth of teams might kill me, so I'm sticking with the Raps this time around.


      • #4
        Also, if you look at the article in full, you'll see I looked back at last year's projection using this new model to see how it would have worked in retrospect (again, just for the Raps). Never bothered doing that last year. For the record this model would have predicted 54 wins last year, not 48, so I think I've gone in the right direction here.


        • #5
          This makes so much sense I'm now expecting 58 wins LOL.

          Great work Dan!
          The name's Bond, James Bond.


          • #6
            We lucked out in free agency when RR landed you Dan.

            Do you happen to have the WS and Worp numbers for Carroll and JV from their last healthy seasons? By healthy I mean when they didn't miss a large significant chunk of games.


            • #7
              LJ2 wrote: View Post
              We lucked out in free agency when RR landed you Dan.

              Do you happen to have the WS and Worp numbers for Carroll and JV from their last healthy seasons? By healthy I mean when they didn't miss a large significant chunk of games.
              This approach takes a per minute WS and WORP number, then applies them to expected minutes played. So, for example, JV's WS and WORP numbers were higher this past year than the year prior, in terms of per minute production, even though his total WS was lower (WORP he improved so much in that even in fewer minutes he set a career high in total WORP last year). Carroll's were better the year prior (about twice as good).


              • #8
                The east was pretty crazy last year so you were quite off on the top 8 lol, but you were fairly accurate on the west. Nailed Houston finishing 8th when everyone thought they were a contender still!
                It's Klaw Season. Time to hunt.


                • #9
                  OK, quick and dirty with possibly a few errors, here is an unadjusted win prediction for every team in the league (ie, some minute assumptions will be very wrong, and I probably missed a player transaction or two, but hopefully no big ones). Slightly different approach here, had to use WORP differently (the estimated conversion tends to extremize any outliers, so top notch teams will predict more than 82 wins and bad teams will predict win totals in the teens instead of in the twenties, for example), used a pythagorean win estimate based on BPM rather than the direct BPM -> VORP -> WORP conversions.

                  Wins are rounded, but the standings are in order of exact prediction numbers.

                  1 CLE 64
                  2 TOR 58
                  3 BOS 53
                  4 CHA 50
                  5 ATL 46
                  6 DET 42
                  7 IND 40
                  8 WAS 39
                  9 CHI 39
                  10 ORL 38
                  11 MIA 38
                  12 MIL 37
                  13 NYK 32
                  14 BKN 31
                  15 PHI 19


                  1 GSW 78 (thanks, Kevin)
                  2 SAS 71
                  3 OKC 60
                  4 LAC 51
                  5 HOU 48
                  6 UTA 46
                  7 POR 46
                  8 DAL 45
                  9 MEM 43
                  10 MIN 39
                  11 DEN 35
                  12 NOP 32
                  13 SAC 32
                  14 PHX 27
                  15 LAL 26

                  I was surprised by SAS, but Gasol was really good by both metrics last year and they've retained their superstar (Kawhi), and project to come up short in minutes, so all their guys (including Kawhi) see a minutes jump. Same for OKC - they added some solid players in return for Durant, but more importantly the lack of minutes pushed Westbrook up to an insane number of minutes played (42 MPG) - another one where I'd adjust it but don't want to touch anything, just like the Raps (the 58 wins shown is the result of using the different BPM method and letting Lowry's minutes run rampant).


                  • #10
                    lol at the Knicks only winning 1 more game than Brooklyn.
                    Two beer away from being two beers away.


                    • #11
                      Mess wrote: View Post
                      lol at the Knicks only winning 1 more game than Brooklyn.
                      Using last year's stats means Derrick Rose is viewed as a terrible player. And that Noah is expected to play very little.

                      Not that those are inaccurate assumptions, but they are what are driving the terrible numbers for them. That and the Nets are a little inflated by some solidly average players they acquired getting a big minutes bump.


                      • #12
                        Those win totals do come in a little high, by the way, an average of 43.5 wins per team (should be 41). So something has to give there, but whether that happens unilaterally, or one team is 75 wins worse (GSW 3 win season here we come), or anywhere in between, who can say. I expect a lot of the top teams will see their records damp down a bit, it's a long season... So that might include the Raps, if you wanted to put the brakes on the 58 win prediction. Good news is even knocking those 2.5 wins straight off for every team, the Raps still project at 55-56 wins and 2nd in the East.


                        • #13
                          Okc at 60 seems very high..they missed the playoffs in 2015 when durant missed extended time
                          It's Klaw Season. Time to hunt.


                          • #14
                            KeonClark wrote: View Post
                            Okc at 60 seems very high..they missed the playoffs in 2015 when durant missed extended time
                            I agree. They do have more depth now than they did then, but 60 seems way too high.


                            • #15
                              DanH wrote: View Post
                              I agree. They do have more depth now than they did then, but 60 seems way too high.
                              They are def better than 2015 less durant but yeah if they get 60 give Russ the mvp forsure. I expect them in a lower seed around Houston, behind obv Gs and san antonio, but also clippers Utah Portland and Memphis barring major injuries
                              It's Klaw Season. Time to hunt.