fbuijsen, on 2014-August-18, 01:52, said:
As I read it, lanford's argument is specifically for knockout matches: it is simply impossible to really give unbalanced assigned scores. Let's say you give an unbalanced assigned score of +3 to both sides as above, and the score ends up as 33-31 for team A and 34-30 for team B (the score being 31-30 without the one board). Clearly, the team B ends up winning the match.
So the so-called unbalanced score assignment is effectively the exact same as as giving an assigned score of 0 IMPs to each side.
True in a head-to-head match, but not necessarily in a three-way (either one or two survivors). This situation actually arose several years ago in a Gatlinburg Regional. I had a brain misfire and gave a ruling from which we deemed that there was no recovery (indeed, it was barring the wrong partner after an opening bid out of turn). We awarded +3 IMPs to each side; the net effect of this could have been that the third team in the KO, who were completely uninvolved in the ruling might have been eliminated (in fact, happily, the margins were such that the error had no effect on the outcome).
For example, in a 3-way, 2-survivor round: A beats B by 2, C beats A by 1, and after an adjustment for director's error B is +4 and C is +2 in the B-C match. So A has 1 win, B has 1 win, and C has 2 wins. The CoC have a team with 2 wins going forward, so C progresses. When more than one team has a win and a loss, it is resolved by net IMPs; A is +1, B is +2. B goes forward, Ignoring the tainted board, each team would have one win, A (net 1 IMP) and C (net 0) would go forward with B (net -1) eliminated.
I am not saying that an unbalanced assigned score of +3 to each party to a director's error in a KO match is wrong; I am simply observing that in a very rare case it might actually affect the outcome rather than just canceling out.