Accuracy of Head-to-Head data is uncertain

[jleonard2099]

After simulating a month's worth of games, a comparison of actual stats to replay stats was done for a handful of teams.

In each case, there were instances revealed where "actual stats" were 0 but the replay recorded values.

At first I questioned the accuracy of my routine for reading in the stats, but the majority of other stat numbers were not 0.

It should be determined where the "actual stats" is populated from, and if this information is accurate, or incomplete.

• If incomplete, then the replay is likely accurate but that begs the question of how to complete the data.
• If inaccurate, then we need to first be sure the values used were populated correctly in the array or determine other sources of the inaccuracy.

What was telling to me is the data looked identical in the original program; for this reason I'm also adding the "legacy-bug" tag (although it may not apply if it's an issue of data missing).

[jleonard2099]

Here is a screenshot example of what I witnessed.

[jleonard2099]

I have since determined that the "actual" stats for teams are saved in the COLBBTMS file, if entered in for the team. As this secondary file was a late addition to the game, it has not been completed for all teams. Replay stats are just that - but actuals are to give you an easy comparison to how the team really did that year. However, such an effort would have until now required manually entering the data for all historical teams.