Blazers 2020-2021 Win-Loss Prediction from BPM

Discussion in 'Portland Trail Blazers' started by bobf, Dec 22, 2020.

  1. bobf

    bobf Well-Known Member

    Joined:
    Oct 19, 2008
    Messages:
    3,959
    Likes Received:
    3,142
    Trophy Points:
    113
    Here is my 2020-2021 Blazer win-loss prediction based purely on the past two years' BPM's and a guesstimate of player minutes. I ran the same prediction method for the past two seasons but using actual minutes played:
    2018-2019 predicted 47-35. Nurk's BPM way outperformed prior years & accounts for almost all of the prediction error.
    2019-2020 predicted 35-39. Bingo.
    2020-2021 predicts 46-26. :)

    FTE - Full Time Equivalency - e.g. 0.68 means a player plays 68% of a full game every game.
    PBPM - Predicted BPM. Weighted average of prior two years BPM with most recent at 2x.
    PWINS - Predicted win contribution. BPM x FTE x 2.5. Historically a full-time BPM correlates with +2.8 wins.

    Players with no data get a PBPM of -3.0. To be conservative, I lowered Hood's PBPM by -1.0 due to his injury.

    Blazers BPM Predictions.PNG
     
    Darkwebs, B-Roy and UKRAINEFAN like this.
  2. B-Roy

    B-Roy If it takes months

    Joined:
    Oct 14, 2008
    Messages:
    31,402
    Likes Received:
    24,364
    Trophy Points:
    113
    Carmelo is bad if you're trying to actually win games.
     
    tester551 and bobf like this.

Share This Page