Fix timestamp conversion for match data

[obreit]

Sbopen().match() can crash for certain competitions due to a discrepancy in time formats of the last_updated column. For example, competition = 37, season = 42

Small script that can be used to check how it behaves

from mplsoccer import statsbomb

loader = statsbomb.Sbopen()

competitions = loader.competition()

for comp in competitions[['competition_id', 'season_id']].to_dict(orient='records'):
    print(comp)
    loader.match(competition_id=comp['competition_id'], season_id=comp['season_id'])

[andrewRowlinson]

Thanks for noticing this - I changed the code so it works with pandas versions below 2.