lbenz730 / ncaahoopr_data Goto Github PK

In the roster files, the player_id column is misaligned with the player_id column in corresponding boxscores, as well as the player_image column. For example, in [https://github.com/lbenz730/ncaahoopR_data/blob/master/2022-23/rosters/Wyoming_roster.csv](the Wyoming roster), Hunter Maldonado is listed as played_id 5107866. However, his player_image is https://a.espncdn.com/combiner/i?img=/i/headshots/mens-college-basketball/players/full/4280267.png. The image is correct. And then in the boxscores, such as https://github.com/lbenz730/ncaahoopR_data/blob/master/2022-23/box_scores/Wyoming/401494418.csv, Maldonado has player_id 4280267. Scanning a few roster files, this mismatch appears very prevalent, with only a handful of player_id and player_image values aligning.

The master_schedule.csv files in pbp_logs/ are broken for the past two seasons, as they each only include a handful of games.

where to find the best real data in the world to analysis, thanks.

I loaded these data into pandas and when i ran a unique over position i got some positions that were not "F", "C", and "G"

Here's each of the labels in position and the value counts (I'm only pulling in data from the 2013-14 season to present)

Not sure if I'm looking in the wrong place but are there PBP logs for 2020-21 games that occurred before the new year?