You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It's not clear to me what the behaviour should be, just something I noticed when writing unit tests.
If there is a debate with only unconfirmed ballots, then this debate doesn't count towards the second-order metrics used in calculating draw strength, but the fact that it is drawn does count for the purposes of determining who a team's opponents have been when calculating draw strength. This is arguably a strange inconsistency—we are counting their opponents' record in a team's draw strength, even when it isn't fully accounted for that they faced the opponent.
For now, I've marked these as expected failures in 78ca17f.
The text was updated successfully, but these errors were encountered:
- Moved elimination round tests to separate tests #1073
- Added checks for unconfirmed ballots
- Suppress routine logs to prevent clutter when run from command line
- Two expected failures in draw strength, see #1466
It's not clear to me what the behaviour should be, just something I noticed when writing unit tests.
If there is a debate with only unconfirmed ballots, then this debate doesn't count towards the second-order metrics used in calculating draw strength, but the fact that it is drawn does count for the purposes of determining who a team's opponents have been when calculating draw strength. This is arguably a strange inconsistency—we are counting their opponents' record in a team's draw strength, even when it isn't fully accounted for that they faced the opponent.
For now, I've marked these as expected failures in 78ca17f.
The text was updated successfully, but these errors were encountered: