[Bug]: Data Not Reconciling

[bbq-x]

NBA API Version

1.6.0

Issue

The last game of the finals (game_id: 0042300404), the aggregated data for the teams is not equal between 'playergamelogs' and 'teamdashlineups' endpoints. I have been able to collect data from both endpoints by game id and period for literally every game the previous three seasons except for the game listed above.

Code

This is the code I use to validate the two sources. I use the API to store to a database and then I pull from my database
def get_stats(table,table_var,id):
sql = f"select {test_json} from {table}_{table_var} where game_id = '{id}'"
df = pd.read_sql_query(sql,con=engine)
df.columns = common_cols

return df

sql = f"select game_id from events_{table_var}"
with con.cursor() as cur:
cur.execute(sql)
games = [item[0] for item in cur.fetchall()] # Returns table data in a one-element tuple

for id in games:
comp_list = []
for table in table_names[1:]:
comp_list.append(get_stats(table,table_var,id))
for col in common_cols:
player_val = comp_list[0][col].iloc[0]
lineup_val = comp_list[1][col].iloc[0]
if player_val != lineup_val:
if col != "TOV":
print(id,col,player_val,lineup_val)

[rsforbes]

@bbq-x - The NBA manages all data, we are just passing that data through the library. You could attempt to contact the NBA, but I'm uncertain of that process.

[bbq-x]

The data is correct on the NBA site. When I break it down by lineup and box score, the aggregations match on the site. But when using the API, the final game between Dallas and Boston, those aggregations do not match. But okay. Thanks for replying.

…

On Wed, Oct 23, 2024 at 10:38 AM rsforbes ***@***.***> wrote: @bbq-x <https://github.com/bbq-x> - The NBA manages all data, we are just passing that data through the library. You could attempt to contact the NBA, but I'm uncertain of that process. — Reply to this email directly, view it on GitHub <#470 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A3VMM4M2HT2WLBGUZQORK7DZ46YF7AVCNFSM6AAAAABQICBSE2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMZSGQ2DMNRZHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[bbq-x]

validation.docx

Attached is a document that better explains my issue. I use playergamelogs and iterate by period. I compare this data to the period data obtained using teamdashlineups.

I have acquired data from these two endpoints for the entire three previous seasons. The document lays out the discrepancies. The parsing of the info is as follows:
[game_id] [team_id] [column] [period] [comparison status] [discrepancy]

For turnovers, they're a mess. [total from playergamelogs] : [total from teamdashlineups]