NFL Machine Learning - QB Passing Yards Regression Model (Part 1)

Поділитися
Вставка
  • Опубліковано 23 січ 2025

КОМЕНТАРІ • 5

  • @scottparrish7244
    @scottparrish7244 Місяць тому

    I found your reddit post which led me to your video here as I was searching to find others who had undertaken the same project as myself: using machine learning to predict passing yards on a game by game basis. I use R rather than python. I used far less data to train (2021 through 2023). Then tested on the first 13 games of the 2024 season. The MAE using Random Forest was 55.56 and the rmse was 72.78. I ran multiple tests using a myriad of different predictor variables of rolling three-game averages for atts, comps, comp_pct, yds/comp, yds/att as well as rolling 3-game average defensive stats in the same categories.I also had pro football focus grades for a myriad offensive and defensive areas. I ran many, many tests. NONE of the predictor variables that I plugged in (there were about 33 in all) were as predictive as each game's over/under and point spread. -- I'm not a datascientist. I'm self-taught. I used chatGPT's newest model (o1-preview) to walk me through the steps.

  • @leavonfletcher4197
    @leavonfletcher4197 2 роки тому +1

    Jeff Garcia played for Cleveland, Philly and San Francisco. If I remember right, he came out ot San Jose State. Had a pretty good career.

    • @Timbryan22
      @Timbryan22  2 роки тому +1

      Very cool, the play-by-play data gives me a nice trip down memory lane all the time

  • @malone1020
    @malone1020 Рік тому

    Getting this error when attempting to import the passer data. Any ideas?
    AttributeError: module 'pandas.core.strings' has no attribute 'StringMethods'