Does anyone want to get/share some data on map win percent compared to map play count?

It would be nice to settle the disagreement regarding the importance of map familiarty/BO's that has been discussed in this thead (among other places). In theory, if map familiarity doesn't really matter, then there would be basically equal odds of a player winning the first time they played a map vs the 5th vs the 10th vs the 50th, etc, on average. However, if map familiarity really does matter, then a player would have lower odds of winning on their initial games on the map, and those odds would increase as they play it more, until the player reaches a general plateau in their win percent on that map, on average.

There are, of course, many other factors in reality, but aggregate data should still be able to address this. So, if someone wants to get and share some data, preferably sourced from matchmaker games, on players' win rates on maps vs the number of times they'd played those maps, that would be swell. So, for example, if you looked at the data for 100 players who each played a map 10+ times, and 40 of them won their 1st game on the map, 45 won their 5th game on the map, and 49 won their 10th game on the map, that would support the point of map familiarity being impactful.

Why does this matter?

It seems that a lot of players view asymmetrical map/BO familiarity, knowledge and experience as a real factor in performance and an additional source of ladder stress, while a number of others (including some members of the matchmaker team) seem to basically discount that as an invalid excuse/coping mechanism. So, settling this with data could have some helpful implications.

pfp credit to gieb

I don't understand what knowing this would actually solve. Say for the sake of argument we find out that on the first play the win percentage is 45% and it climbs to 50% over the first 10 games. What do we do with this information?

It would help establish that map familiarity is impactful, which could impact a number of players' perspectives in FAF, and it would theoretically be a supporting argument in favor of adding mapgen-only queue(s) and potentially some other changes, and it would provide validation for the people who have been arguing that map familiarity/BO's matter.

pfp credit to gieb

Do we really need data for this? We already have a prime example.

Setons. Take any Setons player vs any non-Setons player, the Setons player will punch far above their rating when playing on Setons against someone not super familiar with it.

I understand the argument that good players just know how to quickly figure out what they need to do on any map and don't need a BO, I mean that's just part of getting good. It's why I like map-gen, because it's puts everyone on as equal ground as possible in this game. But it's no replacement for having specific BOs to execute specific actions.

Example: When I still played Setons and was a lowly 1200, I learned and IMO got exceptionally good at the double fac opener for mid and was able to consistently take down players 1800+ including Yudi (Which I will never forget because it gave me a massive ego boost at the time :^) )

Intuitiveness, adaptability, and game knowledge are great and necessary to getting good, but they alone aren't always or even most of the time enough for going against good players with good BOs.

There's my 2 cents.

Nobody argues that map familiarity doesn't matter, of course, it does. People argue whether BOs, not map familiarity, have a significant impact on player performance on the ladder given that:
1.) The significant majority of players don't have BOs anyways
2.) Even if they do they are usually only barely better than the most basic generic one

With that in mind, openings come down to who can better improvise, already knows what to do more or less (map familiarity which requires you to play the map 1, maybe 2 times beforehand), or pure lack. With the introduction of map-gen only you're only removing the map familiarity, but only in theory since generated maps are similar to each other. Play enough of them and you will be familiar with most map-gen maps which again gives you an advantage over players that did not.

@tagada said in Does anyone want to get/share some data on map win percent compared to map play count?:

Nobody argues that map familiarity doesn't matter, of course, it does. People argue whether BOs, not map familiarity, have a significant impact on player performance

I think a significant portion of the complaints regarding BO's don't actually mean BO's by your definition, but instead mean the general familiarity and knowledge of how to reasonably open on and play the map that players typically get after playing it a few times. I think this is also generally more impactful at lower levels than higher levels on average, as the impact of individual map familiarity can be reduced by increased generalized skill and map reading ability in FAF.

Also, FTX said yes when I asked, "do you really deny the concept that people play a map better by like the 5th time they play it than the 1st, and play it even better by like the 10th time, etc in general?"

That seems like denying at least a significant aspect of the impact of map familiarity, and that question is clearly and directly related to the OP and the purpose of this thread, regardless of what you want to call it

pfp credit to gieb