EPTASL vs FlaSh
| Forum Index > BW General |
|
MineraIs
United States846 Posts
| ||
|
Akio
Finland1838 Posts
 | ||
|
XenOsky
Chile2358 Posts
| ||
Sigrun
United States1655 Posts
| ||
|
butchji
Germany1531 Posts
| ||
|
M3t4PhYzX
Poland4289 Posts
On January 18 2023 08:17 Sigrun wrote: Don't map makers always do this when one player/race is dominating? I remember Savior getting shit awful Zerg maps during his OSL run. yes BW's balance is mostly done with maps been like this for years and years | ||
|
Dante08
Singapore4188 Posts
On January 18 2023 08:17 Sigrun wrote: Don't map makers always do this when one player/race is dominating? I remember Savior getting shit awful Zerg maps during his OSL run. Yes but only slightly, ASL did it to the extreme that season with anti-Terran maps. Sparkle, Third World, Transistor, Gladiator were all bad for Terran. | ||
|
Djabanete
United States2786 Posts
| ||
|
RKC
2848 Posts
Also, during match day itself, the extra effort to win on a disadvantageous map will invariably take a toll on a player's performance in later games in the series. Imbalanced maps can affect a player's performance in an entire series and not just on games played on those maps alone. | ||
|
Avi-Love
Denmark423 Posts
![[image loading]](https://i.imgur.com/581u4wi.png) | ||
|
TMNT
3378 Posts
- Sparkle: TvP 60.4% (255 games), TvZ 48% (373 games), Total 53% (628 games) - Third World: TvP 48.9% (237 games), TvZ 68.5% (143 games), Total 60.8% (380 games) - Transistor: TvP 42.9% (352 games), TvZ 50.3% (567 games), Total 47.4% (919 games) - Gladiator: TvP 49% (878 games), TvZ 53.8% (1221games), Total 51.8% (2099 games) Doesn't look too bad. Meanwhile, the win rate of Terran on the de facto standard maps: Polypoid 54.8%, Vermeer 57% (lol), Sylphid 54.9%. Maps which are actually balanced based on win rate: FS 50.6%, Eclipse 51.1%, Butter 50.2%. You can see a pattern here: - maps where Terran dominates (most maps actually) are considered standard, balanced - maps where Terran have a 50/50 chance of winning, are considered slightly disadvantageous - maps where one of the match-ups drops down under 50% for Terran, are considered extremely disadvantageous | ||
|
Djabanete
United States2786 Posts
Re:Sparkle — My take is that the organizers wanted it to be unfavorable for Terran and they just messed up. Balance is hard to predict. Third World had incredible games. | ||
|
TMNT
3378 Posts
On January 21 2023 16:44 Djabanete wrote: TMNT: You may be right, although if Flash’s games account for a high proportion of one’s sample, then I could see the argument that a map with 50-50 results is unfavorable for Terran. (I am not actually trying to contradict you, since I don’t know what proportion of the games you’re referencing were Flash games, but it’s an interesting wrinkle.) I think the best data for evaluating map balance would be games played by players ranked ~#5–10 within their own race, but I don’t know if anyone has ever spent the time to curate that information. Re:Sparkle — My take is that the organizers wanted it to be unfavorable for Terran and they just messed up. Balance is hard to predict. Third World had incredible games. I see this argument a lot (Flash skewing the win rate), but every time I checked, it really wasn't the case. Sponbbang allows you to filter record by date, so if we look at the stats from July 2021 to now when he's been away, the win rates for Terran are: - Polypoid 55.4%, Vermeer 56.3%, Sylphid 53.9% - FS 51.%, Eclipse 51.5%, Butter 50.2% So no difference at all. Also Vermeer and Butter were created after Flash went to the military (there's a slight difference for Vermeer probably due to the sources I took - the previous one is from eloboard). It really is basic statistics. He has to play like, a number of games equal to every other Terrans combined to have any sort of influence on the whole race win rate. | ||
|
TT1
Canada10090 Posts
Sparkle ended up being T favored once T's figured out how to play it optimally (early exp into wraith contain style while mass exping/taking map control, Snow tried to mindgame this b.o with an early DT drop which ended up doing a ton of damage.. but he still lost), it happened towards the end of the season. That said i still don't think the map was explored enough, that's why it's the best map ever made kek. 3rd World was kinda hard cus T was forced to go 2 base 2-1 pretty much every game cus it's a carrier map. That b.o is pretty versatile tho so the stats were about 50/50. It made for super boring games tho, terrible map overall. But ya obv Terran is broken in the hands of a godly optimizer/mechanics player w/ great decision making like Flash. Tours have to figure out how to keep things spicy, which is doable with maps. It def makes life hard for other pro T players because they're nowhere close to his lvl (RoyaL has that potential tho). If you get super bots to play each race to their maximum ability/efficiency Z would be the best, but with human limitations it's T. | ||
|
RKC
2848 Posts
For example, let's say a super small 1v1 map which doesn't allow for any macro play. Flash may still have a higher chance of winning with early game builds against most opponents due to superior mechanics and star sense. But the chance is still lower than a standard map where he can both outplay opponents in early game and long macro game (which is actually where he really shines above all other players). | ||
|
Freakling
Germany1534 Posts
On January 21 2023 20:54 RKC wrote: Map imbalance is not just about raw numbers of win probability and actual win rates. But also in a qualitative sense - a map that forces a race to play with a limited build and style (while the other race has a wider repertoire of options or limited but more easily executable builds). For example, let's say a super small 1v1 map which doesn't allow for any macro play. Flash may still have a higher chance of winning with early game builds against most opponents due to superior mechanics and star sense. But the chance is still lower than a standard map where he can both outplay opponents in early game and long macro game (which is actually where he really shines above all other players). Probably not a good example. It's actually pretty hard to make a "super small map" (something like PA) that is not heavily Terran favoured, because it's either gonna have very few expansions, which will favour Terran due to lack of P/Z's ability to out-expand them, leading to relative fast and easy even map split situations in many games, which ultimately are heavily in Terran's favour, or it will have a high density of expansions with small distances in between, which again favours the less mobile, more area-defence reliant race. | ||
|
TMNT
3378 Posts
On January 21 2023 20:54 RKC wrote: Map imbalance is not just about raw numbers of win probability and actual win rates. But also in a qualitative sense - a map that forces a race to play with a limited build and style (while the other race has a wider repertoire of options or limited but more easily executable builds). For example, let's say a super small 1v1 map which doesn't allow for any macro play. Flash may still have a higher chance of winning with early game builds against most opponents due to superior mechanics and star sense. But the chance is still lower than a standard map where he can both outplay opponents in early game and long macro game (which is actually where he really shines above all other players). I agree that map balance also has to be judged from a qualitative sense. Like, you can look at the "natural expansion" on Transistor and see it's horrible for TvP or TvZ (I'm actually surprised that TvZ is 50% for this map given the areas Muta can abuse - maybe someone can explain this?) But I don't think limiting a race's build or style on a given map can be considered disadvantageous to them, as long as they're still winning it. I mean, the gameplay itself already forces you to do so. Terran has to go mech in TvP, Protoss has to build Corsair in PvZ, Zerg has to build Muta in ZvT,... On the other hand, wider repertoire for the other race doesn't mean they have a better chance of winning the matchup. Take PvT for example. "Protoss can do anything", a great man once said - while Terran can't do anything until they take a 3rd. Now if we create a map where Protoss can only go Carrier but consistently has 55% win rate over Terran, while Terran has the option to go bio, would you say Protoss is at a disadvantage? | ||
|
Eywa-
Canada4876 Posts
On January 21 2023 20:54 RKC wrote: Map imbalance is not just about raw numbers of win probability and actual win rates. But also in a qualitative sense - a map that forces a race to play with a limited build and style (while the other race has a wider repertoire of options or limited but more easily executable builds). For example, let's say a super small 1v1 map which doesn't allow for any macro play. Flash may still have a higher chance of winning with early game builds against most opponents due to superior mechanics and star sense. But the chance is still lower than a standard map where he can both outplay opponents in early game and long macro game (which is actually where he really shines above all other players). Map balance is purely about statistics, any argument that the overall win rate doesn't accurately reflect map balance should be backed with statistics. A map that has a 50/50 win rate, but one race can only do 1 build is equally balanced as a map where any build is viable, but the win rate is 50/50. If Light wins 50% on a map that he can only do 1 rush build, and 50% on a map where he can do any build, which map is he most likely to win a game on? The assumption that more choice means a player has better control over their win rate is a common gambling fallacy. Of course, maps will have stylistic tendencies that will affect the win rates of individual players... This doesn't matter for overall map balance. I think understanding how the win rate is achieved (i.e. only 1 build is viable) is important analysis for the quality of the map and enjoyment of the map, but it's irrelevant for any balance discussion. There could be statistical differences between ASL results and sponsored match results because ASL has more prep, but the sample size makes it pretty difficult to draw any definitive conclusion and arguably this is irrelevant unless you're top 8 which means it only really affects 1-3 players. It's worth noting though when a map is introduced into a tournament setting, often the race with the most intuitive good strategies will be favored, and we might later discover that the map is actually balanced / imbalanced. This novelty effect could skew perception / ASL stats because maps are often first introduced in ASL. | ||
Liquid`Drone
Norway28900 Posts
So basically Sparkle was overall worse for TvZ no matter what terran opted for (terran basically had some various openings that could give them big advantages or outright win - but if Z hadn't been hurt the first 10 minutes, they'd win almost every time), while for TvP, it became a fantastic map when terran played it correctly and executed flawlessly, but it wasn't a particularly strong TvP map for most people. | ||
|
TMNT
3378 Posts
On January 23 2023 05:48 Liquid`Drone wrote: for TvP, it became a fantastic map when terran played it correctly and executed flawlessly, but it wasn't a particularly strong TvP map for most people. I mean, for "most people", TvP isn't a strong matchup anyway  | ||
| ||
