It's interesting because we don't see JSA placing well at many tournaments globally -- is JSA winning more often but more narrowly?
Said it in the other thread so just gonna repeat it. That's relatively low sample size and looks about as accurate as the last batch of numbers we got. A lot of performances look off in one way or another (Shas and Dashat). Winrate is not a useful metric with ties and major/minor wins all making a huge difference in average TP/OP, the metric that really matters.
All of those metrics aren't worth squat. We do not know how many players each faction, how good they were (their own ranking level), how many events, what scenarios, what additional rules (Limited Insertion, etc.). If a faction is deemed "weak" so the only people who take it to a tournament are total pros who find a meta niche for it, they will swing up. If the same faction is deemed "weak" and it's taken mostly by amateurs "for the lulz", it will swing down, hard. If another faction is deemed "OP" but it's taken mostly by amateurs as a crutch, it will swing down anyway, even if some pros use it to its best potential. Et caetera, ad nauseam.
If that were for a single faction, same ITS rules (e.g. 300/6, no specials), same scenario or same set of three scenarios, then sure, why not. This instead is a jumbled up trail mix of games of unknown properties, randomly (but unevenly) reinforcing and undercutting the faction's strong points. In some cases there's just 41 (FORCO) or 61 (QAPU) games played for the whole faction.
Yeah, the individual faction sample size is a bit lacking overall. Plus the whole case of a large amount of possible missions and various tournament modifiers like specials and LI can really swing stuff around the place. A statistic based on non-LI and LI tournaments will have wildly different looking results, especially for factions that are so cheap you can barely even make a LI list ( Insert broad gesturing at Ariadna here. ) or those that are naturally able to capitalize immensely upon the format ( Broad gesturing at IA here. ).
Compared to the last time we got similar numbers it's a fraction of the sample size. And the last time around there was the full data export with even more games covered weighing against it, providing actual performance graphs, TP, OP, etc as well as fitting expectations a lot better.
Well... a sample size of 300 isn't small per se, but mostly I mean that the sample is either borked or hasn't been cleansed of byes and not cleaning out byes which are literal false positives and fairly easy to clean means there's a non-negligible risk of the win percent being affected. Other than that, it's both reinforcing what I was expecting from some factions and it's completely contrary to what I would expect. JSA performing great is completely contrary to what I'd expect considering that it's a difficult faction and it should be under-performing if it's also popular with newbies. That said, I don't think we'll ever see very good stats for this game since we never get match-made against players of similar ELO. It might be possible to make a complicated meta ELO to weigh matches for performance based on games played by individuals in cross-meta games, but I think that's a bit over the top. Mostly I'd look at popular and unpopular units/loudouts if I were CB and try to sus out the what and why.