Aligulac doesn’t samples just the pro scene, but the near top of the ladder, with random nonames included. It’s monthly, and its representation over the years is the largest statistical winrate data source available for SC2.
It does not. His inclusion criteria aren’t seriously defined in his first post, and when asked about the period covered he links sometimes to the latest period available (which is way shorter than monthly average).
Interestingly enough, he also gives links to the whole Aligulac database when asked where to check his data, but only to a single period when asked about his source. And said period isn’t cleary defined, as “latest” constantly changes.
I’m saying his protocol description is lacking (allowing some choice over what players are included. At some moments he gives links only to one single period (which reduces statistical power, and would allow a deliberate choice of the period) at others he suggest using the whole database (including WoL + HotS + LotV) which wouldn’t make any sense to describe current balance (and would explain how he could get protoss having favorable results over zergs). It’s cloudy at best.
Also, upon trying to find back his profile, to take a more detailed look at his thread I found that his RFTW profile sheet now didn’t linked to tEhbAtZ ; but to a barcode. Don’t you find this interesting, barcode ? Considering his silence going for unusually long, the similarities of your views, and your interesting understanding of a thread now long gone from main page. 