[Eugene]

To All:

I totally agree that winning matters. What bothers me is the fact that there are only 8 games that are played. Excluding the variable presented by the alliance partner and assuming that the team scores do not fluctuate much (the data supports that), if the ranking was done by scores than no matter how the teams were paired, the better performing teams would consistently be an top whereas by using the win/lose system the ranks would fluctuate far more. If you look at the table that I compiled, you would notice that the difference in ranks by different methods is huge. One of the statistical methods (which is what we should use in our argument if we were to consider ourselves scientists, as in fir_S_t) that is used to evaluate the validity of the test is the correlation of scores between that test and other tests. In that aspect, the current scoring system would not be considered valid.

Eugene