[Jaybarnes]

We often notice a difference between MPs reported under history and the main game screen. I know the hands are continually being played and the percentages will change, but for our competition purposes which of them better reflects the outcome?

[mycroft]

Almost always, if you care about "the field's" results, comparing against more will be a better comparison to what "the world" would do.

But note, one of the reasons for team games (and one of the reasons the world championships and many of the top-level events are team games) is that "the world" is kind of random (and frankly, especially the BBO non-tournament field, with its high percentages of "play 5 hands over lunch with whoever to kill time" or "having a good time talking with my friends, using bridge as the reason to get together") and comparing against one, known quantity, other table can be a better way of knowing if "I did well" on this hand (or these 10 hands).

If you're the level of person who needs to ask this question though, "the field" is probably a fine comparison for you (and me).

This strongly applies if you're scoring Matchpoints. For one thing, if there's 3 results on a hand, you're either 100%, 50% or 0% (okay, you could have tied with a score, so 75 and 25% are possible). If there are 11 results on a hand, each "I beat that table" is only 10%.

If you're using the default though and scoring cross-IMPs, a warning that statistics says that the numbers get squashed somewhat. If you bid a vul game that "the other table" didn't and make it, that's 10 IMPs. But if there are 14 comparisons, and half found the game and half didn't, you only get 5. In a team game, +2 to -2 should be thought of as "a wash" (yes, you can lose a match by 1 IMP, but you know what I mean). Cross-IMPed, 16 tables, +2 is actually something, +6 is big, and +10 or more is "slam 'nobody' else brought home" level.