Quote from LEH »Quote from ktkenshinx »Re: matchups
One definite issue we have is a lack of good data about real matchup percentages. Claims of 50/50 vs. 70/30 matchups are hard to test when our data is so throttled. It basically forces us to rely on anecdotal information instead of real matchup data. That said, there is enough consistent performance at top Modern tables and big events that it seems very unlikely there are as many 70/30 matchup-lottery games as people claim. I suspect the reality is exactly what I computed in an earlier analysis; matchups in Modern are about 3%-5% more variable than Legacy. This means that 1 in 20 to 30 games at a GP could be really affected by variance.
While it's always going to be a guesstimation on exact MU percentages, I dislike the view that "anecdotal" information is seen as next to useless on this forum particularly when it stems from multiple individuals whom all have 100s if not 1000s of games under their belt and all saying the same thing. At what point does anecdotal information become "Qualitative data"?
Data like that is not anecdotal. Let's get more of that please! I'm talking about users who complain about matchups without posting any numbers to back up their claims. Some people track matchups in a spreadsheet. We want more of that in this thread. Others just recall negative or positive experiences in recent games and extrapolate it to the entire format. That's the anecdotal information I'm fighting against.