To tie or not to tie, that is the question

[McBruce]

Pairs tournament 4164 today, interesting ending:

 

Name Score (%) Rank

A B C

Prize Points

Lin1951+van0602 60.63

1 1 1

1.28

veecon+Robby8 60.62

1 1

1.28

 

One pair's results show 2/26, 60.62% at the top of the list and 60.63% at the bottom.

The other's results show 1/26, 60.63% at the top of the list and 60.62% at the bottom.

 

It would seem we have a rounding discrepancy in the code.

[pgrice]

Pairs tournament 4164 today, interesting ending:

 

Name Score (%) Rank

A B C

Prize Points

Lin1951+van0602 60.63

1 1 1

1.28

veecon+Robby8 60.62

1 1

1.28

 

One pair's results show 2/26, 60.62% at the top of the list and 60.63% at the bottom.

The other's results show 1/26, 60.63% at the top of the list and 60.62% at the bottom.

 

It would seem we have a rounding discrepancy in the code.

 

There is indeed. I've seen examples where one pairs is ranked above an other but has fewer match points. Very sloppy to do your rounding/truncation in the middle of the calculation rather than at the end.

[McBruce]

There is indeed. I've seen examples where one pairs is ranked above an other but has fewer match points. Very sloppy to do your rounding/truncation in the middle of the calculation rather than at the end.

 

This is usually because there is an averaged board in there somewhere. The BBO system calculates the percentages correctly, but sometimes on ACBL Live for Clubs the matchpoints are wrong, even though the percentages copied from BBO results are correct.