Worlds post 10/8 - the projections bit

Alright, here's what you've all been waiting for, the first round projections. This is basically a data dump, and I've colour coded a lot of it, so here's the explanations as to what is meant by what.

In headers, date and player should be obvious. Short, medium and long are the size of the data. The first is the last three full months (plus whatever is in this month, which for the entire field will be nothing). The second is the last six full months, the last is the last twelve months (data model automatically filters out all results from more than 365 days ago). The composite is the overall projection. For short/medium/long, green indicates at least than 50 winning legs for both players in that time period. Yellow indicates between 20 and 49 winning legs. Red indicates less. For the purposes of calculating the composite score, green numbers are counted in full, yellow counts half, and red is ignored, basically to limit or exclude data samples that are too small to be completely or even partially reliable. As for the coding of the composite score itself, green indicates three greens or two greens and a yellow, and should be very reliable. Yellow indicates individual scores that count for two or one and a half samples, so something like a green and two yellows, one of each. These numbers should still be OK, but proceed with caution. Red numbers indicate where there is less data again and are only included when there is at least one partially reliable dataset to go with. I would NOT use these for any gambling purposes, they are just there for completion. Where we don't have any data or not enough data even on the full sample size, that is indicated. A N/A in the short/medium/long column indicates that the player on the right does not have any data at all in that period. Finally, all numbers are percentage chances to win FOR THE PLAYER ON THE LEFT.

With all that said, here we go: