"Not played"
Sorry to revisit an old topic. I hope we all know that the "not played" option on the bridgemate is not to be used as a substitute for an artificial adjusted score when for some reason (e.g. insufficient time, receipt of unauthorized information) a schedule board cannot be played. It can be used to bypass the score input for boards that are not scheduled to be played. For instance, when running a two-session pairs event with a Mitchell in one session and a twinned Howell in the other, I was unable to edit the movement to miss the first round of the Mitchell, so I just had the players enter "no play" for those boards. I would also use it if a pair had to withdraw before playing half their boards, so their scores would be deleted.
A correspondent asks me what they should do if they start a pairs movement but are unable to play the last round because they have to finish the game by a certain deadline. Is it better to enter averages for the last round boards, or "not played"? That round was scheduled to be played, but you could argue that you have changed the movement to play only n-1 rounds, so they are no longer scheduled. Does it matter?
Comments
If you're using a barometer movement with no sit-outs, it doesn't matter – you're giving everyone an equal number of matchpoints, so the relative order of the contestants will be the same either way. (It matters even less if you're playing IMPs rather than matchpoints – the scores will be identical.)
If you're using a non-barometer movement, or if there's a pair sitting out, then the choice does make at least a little difference in theory, depending on a) how you score normally reached results on boards where there were artificial results at other tables, and b) how you adjust the results for pairs who sat out more often than other pairs. However, a movement where you don't play all the rounds is unbalanced anyway, so it's hard to know what would be the fairest target to aim for.
I most commonly see directors treat this (very common) situation as a change to the movement, and skip the rounds that they couldn't play. I'm far from certain that that's correct, though; in theory, our rules for handling artificial results are designed to be as fair as possible to the other contestants, so I'd intuitively expect awarding Av/Av on all the unplayed boards to give a fairer result (or to put it another way, I'd expect a balanced movement that was modified by replacing some rounds with artificial Av/Av results to end up more balanced than a balanced movement that was modified by removing some rounds). I am far from certain that I'm correct on this, though!
Of course, the ideal is to identify a balanced movement at the start, and play the whole thing, but this often ends up impossible due to the wrong number of players turning up.
It matters for the percentage (and hence for NGS) whether or not it is barometer. If you curtail a movement for everyone you are changing the schedule and therefore this is one of the times when it is legitimate to use Not Played.
Now that's got me thinking: at matchpoints, how does the NGS impact of an n-board movement compare to that of an (n+1)-board movement where all the scores are moved towards 50% via addition of an Av/Av result to every pair? I'd expect them to be approximately the same (because the impact of the wilder percentages should be approximately cancelled out by the greater number of boards, and they effectively contain the same amount of information about the players' skill levels), although it wouldn't surprise me if they were slightly different.
If it is not barometer and if there is a half table, it could very occasionally affect the ranking list as the pair that would have sat out the last round, if having a good score, is advantaged.
For NGS, if you assign Av/Av scores for the last round, it benefits the weaker players and damages the stronger players, all unfairly.
The number of boards that we are scheduled to play in an event does matter for NGS. Let's say that a 27 (9x3) board event with no sit-outs doesn't then play the last round and we assign Av/Av. For NGS it is equivalent to taking a 24 board event with the real scores (which is what should happen) and then adding on a 3 board event where everyone scores exactly 50% (which is wrong). This becomes obvious if the event is barometer. (It sounded like the OP event was F2F.)
Barrie Partridge - CTD for Bridge Club Live
Thanks for your replies. The question was about an ordinary club pairs, which would not be a barometer tournament. I'll pass your comments on.