| Forum Index > BW General |
|
wasted
Germany1789 Posts
edit: for example: when a player, who will reach A rank, bashes through the lower ranks, those games can hardly give any information about the balance on the played maps. imho, only the games in which both opponents are on a somewhat equal skill level produce usable statistical data. | ||
|
e1-Daft
Latvia19 Posts
There would be interesting percentage of PvT wins on the the Gorky Island map ^^. A nightmare for P. | ||
Plexa
Aotearoa39261 Posts
| ||
|
gravity
Australia2198 Posts
On December 25 2005 22:59 mitsy wrote: part 1 PGT - Rush Hour 2.0 [06] 9896 (3.3x more than luna) PvZ 40.8/59.2 PvT 53.5/46.5 TvZ 51.7/48.3 PGT - Luna The Final [06] 2985 (1.5x more than lotem) PvZ vs 302 326 48.1/51.9 PvT vs 553 432 56.1/43.9 TvZ vs 378 344 52.4/47.6 PGT - Lost Temple 2.4 [06] 1994 (2.1x more than r-point) PvZ vs 135 162 45.5/54.5 PvT vs 358 349 50.6/49.4 TvZ vs 304 291 51.1/48.9 PGT - R - Point 1.0 [06] 931 (1.4x more than p2h) PvZ vs 62 51 54.9/45.1 PvT vs 212 182 53.8/46.2 TvZ vs 88 76 53.7/46.3 PGT - Plains to Hill 2.1 [06] 684 (3.4x more than forte2) PvZ vs 90 100 47.4/52.6 PvT vs 108 100 51.9/48.1 TvZ vs 65 68 48.9/51.1 Another way to measure the relative balance of the maps is to take the square-root of the sums of the squares of the imbalance relative to a 50:50 matchup. For example, Lost Temple would get a sqrt(4.5^2+0.6^2+1.1^2) = 4.67 "imbalance rating". I like this method because it is harsher on maps that have a single extremely imbalanced matchup. For these 5 maps, the "imbalance ratings" are as follows (lower is better): Plains to Hill 3.4 Lost Temple 4.67 Luna 6.82 R-Point 7.22 Rush Hour 10.0 These statistics largely argree with misty's, but they put Rush Hour at the bottom. They would also provide a relatively convenient/objective way to compare the relative balance of different maps. | ||
|
gravity
Australia2198 Posts
On December 26 2005 04:21 Plexa wrote: notice that the general trend is that pvz on *most* maps z>p (excluding rpoint) and the reverse with t where p>t =/ For some reason people keep denying there is any PvZ imbalance though. Sure, it's possible to make a map where P>Z, but that doesn't matter if the vast majority of played maps favour Z. The PvT thing I think is more influenced by the lack of skills of the lower level players on PGTour compared to pros, since T is the most micro-intensive race, but there could be an issue there too (it's hard to get pro-level information, but at least at pro level it isn't common to say "oh no, that T player got a lot of P matches, he's screwed" compared to saying the same about P and Z). | ||
Liquid`Jinro
Sweden33719 Posts
On December 26 2005 05:00 gravity wrote: For some reason people keep denying there is any PvZ imbalance though. Sure, it's possible to make a map where P>Z, but that doesn't matter if the vast majority of played maps favour Z. The PvT thing I think is more influenced by the lack of skills of the lower level players on PGTour compared to pros, since T is the most micro-intensive race, but there could be an issue there too (it's hard to get pro-level information, but at least at pro level it isn't common to say "oh no, that T player got a lot of P matches, he's screwed" compared to saying the same about P and Z). No, the problem is that most P players are fucking TERRIBLE at PvZ -_- You might not have noticed, but as of late zergs have been losing tons to Ps (reach is the exception) at the professional level, and zergman even said he thinks the new generation of protoss players are a lot scarier PvZ. PvZ is balanced, the only 'imbalance' I guess is that it takes a lot longer to learn than ZvP does.. Kind of like how you can learn PvT 10 times faster than you can learn TvP =.= | ||
|
gravity
Australia2198 Posts
On December 26 2005 06:04 FrozenArbiter wrote: No, the problem is that most P players are fucking TERRIBLE at PvZ -_- You might not have noticed, but as of late zergs have been losing tons to Ps (reach is the exception) at the professional level, and zergman even said he thinks the new generation of protoss players are a lot scarier PvZ. PvZ is balanced, the only 'imbalance' I guess is that it takes a lot longer to learn than ZvP does.. Kind of like how you can learn PvT 10 times faster than you can learn TvP =.= Just because P have been doing a bit better recently doens't wipe out the historical record and suddenly make PvZ balanced on "normal" maps. Maybe if things stay this way for a long time and there isn't a wave of anti-Protoss Zergs the way July sparked a wave of anti-Terran zergs, or Zergs learn how to beat the newer Protoss style, or whatever. | ||
|
SoMuchBetter
Australia10606 Posts
| ||
Bill307
Canada9103 Posts
 . It would be really neat to compare imbalances at the top with imbalances at the bottom. It would also allow us to measure imbalance at D/C/B/A ranks seperately.I don't think it'd be too much of an additional load on the server, unless the feature becomes really popular. In that case, maybe you could allow people to download a "database" of stats (updated every 24 hours), which would be less than 8 KB, and a simple VB program to search and sort the data to our heart's content . It could also be programmed to perform statistical analyses like the ones in this topic. e.g. you might see something like:PGT - ParanoidAndroid1.0 [06] PvZ - 13:14 - 48.1% (29.6% to 66.7%) PGT - Rush Hour 2.0 [06] PvZ - 1118:1488 - 42.9% (41.0% to 44.8%) Giving you the (95%) confidence interval for the actual win % on the map. | ||
|
gravity
Australia2198 Posts
On December 26 2005 07:04 SoMuchBetter wrote: For some reason people keep denying there is any TvP imbalance though. Sure, it's possible to make a map where T>P, but that doesn't matter if the vast majority of played maps favour P. The ZvP thing I think is more influenced by the lack of skills of the lower level players on PGTour compared to pros, since P is the more micro-intensive race, but there could be an issue there too Hurrr hurrr. The difference is that both the TvP balance is typically smaller at all levels than PvZ, and that the imbalance is more lessened at pro level. PvZ imbalance is a much bigger problem, even if there is still some small TvP imbalance. Again, it's not like people go "omg Oov has to fight 3 P's this OSL, he's fucked", but people (not everyone) *do* say "OMG Reach has to fight 3 Z's, he's fucked". Does anyone have the records of various pro players divided up by matchup, by the way? As far as I can remember from seeing them in the past, no P does as well vs. Z and Oov does vs P, for example. edit: also, just changing the races around doesn't make sense with that post. Since when is P the micro race? T is, and that's part of the reason why they do disproportionately badly at lower levels compared to pro and strong amateur levels. Besides, one imbalance isn't an excuse to ignore another (worse) one. | ||
|
EnDeR_
Spain2931 Posts
On December 26 2005 08:51 gravity wrote: Hurrr hurrr. The difference is that both the TvP balance is typically smaller at all levels than PvZ, and that the imbalance is more lessened at pro level. PvZ imbalance is a much bigger problem, even if there is still some small TvP imbalance. Again, it's not like people go "omg Oov has to fight 3 P's this OSL, he's fucked", but people (not everyone) *do* say "OMG Reach has to fight 3 Z's, he's fucked". Does anyone have the records of various pro players divided up by matchup, by the way? As far as I can remember from seeing them in the past, no P does as well vs. Z and Oov does vs P, for example. edit: also, just changing the races around doesn't make sense with that post. Since when is P the micro race? T is, and that's part of the reason why they do disproportionately badly at lower levels compared to pro and strong amateur levels. Besides, one imbalance isn't an excuse to ignore another (worse) one. if oov had to face 3 p's in an osl, i'd vote for him in liquibet. P not the micro race on pvz? excuse me? are we playing the same game? | ||
|
LastWish
2015 Posts
Rebalancing units may be hard mostly because the game is Professionally played. Some little changes may have some unwanted effects, resulting in one-race domination. Only way to try out new changes is to make a new bw server(with ladder), where these balance issue would be practically tested for some time before taken into professional play. Maybe adding a particular unit for each race could fix some major problems(like some what of anti air unit for zerg on lair level). But why adding new units if certain units/upgrades/abilities are almost never used?(in any matchup) Scouts, Valkyries, Dark archons, Ghosts, Queens could be boosted in some way to regain usability and develop new strategies. Also it somehow irritates me that some units are overused in particular matchups lately - like vessels irradiate and defilers swarm in TvZ. | ||
|
rANDY
United Kingdom748 Posts
| ||
|
gravity
Australia2198 Posts
On December 26 2005 09:02 EnDeR_ wrote: if oov had to face 3 p's in an osl, i'd vote for him in liquibet. That was my point. P not the micro race on pvz? excuse me? are we playing the same game? I was talking about overall, not any particular matchup. | ||
|
gravity
Australia2198 Posts
On December 26 2005 11:20 no_re wrote: With all these statistics, you are also making the assumption that the average p, z and t pgt player is of the same skill level. For example, newer players to the game may choose a certain race slightly over another, and therefore, by losing more often, give the impression that such a race wins a lower percentage of games than another race, giving the impression of an imbalance. Therefore, say PvT on map M has 60% win for P, it does not mean that when two equally skilled players, one p and one t play on map M, that P will win 60% of the games. There's no reason to assume that the skills are *not* equal over a large number of players, unless you have any real (ie not wild speculation) evidence that they aren't. Also, if good players *do* gravitate more to one race, it's also probably because that race is the strongest, so that only makes things worse. | ||
|
oneofthem
Cayman Islands24199 Posts
| ||
|
wasted
Germany1789 Posts
On December 26 2005 22:52 gravity wrote: There's no reason to assume that the skills are *not* equal over a large number of players, unless you have any real (ie not wild speculation) evidence that they aren't. Also, if good players *do* gravitate more to one race, it's also probably because that race is the strongest, so that only makes things worse. and what reason is there to assume that the skills ARE equal over a large number of players? there is no evidence for that too, so it's merely a wild speculation. there are several things in the game which could be imbalanced, the most important ones being: skill of the players, strength of the race and the maps. if you want to draw conclusions about the possible imbalance of maps out of pgt statistics, you can't do this without taking the other potential imbalances into account as well. | ||
|
[GiTM]-Ace
United States4935 Posts
 | ||
|
Mortality
United States4790 Posts
On December 26 2005 11:20 no_re wrote: With all these statistics, you are also making the assumption that the average p, z and t pgt player is of the same skill level. For example, newer players to the game may choose a certain race slightly over another, and therefore, by losing more often, give the impression that such a race wins a lower percentage of games than another race, giving the impression of an imbalance. Therefore, say PvT on map M has 60% win for P, it does not mean that when two equally skilled players, one p and one t play on map M, that P will win 60% of the games. The sample being used is the entire population. There is no inherant bias in that except perhaps a bias in the population itself. If you want to know, as a PGT player, what your chances are in a given match-up, then there is no reason not to accept this data. The only thing that concerns me is how "balance" changes with different skill sets. Some maps emphasize micro/macro skills like Luna and Azalea, while others like 815 emphasize strategy. There's no way to quantify that. What I'm talking about DOES NOT effect the inherant "balance" of a map, but rather the "balance" as determined by an INDIVIDUAL. For instance, a progamer like NaDa with superior macro skills might go so far as to say that Luna favors him over Protoss. That doesn't mean this is true for the entire population. Lots of progamers talk about how 815 is a P map, but everytime I play it, it seems that T >> P (and believe me, I've been both the T and the P). In other words, this data is interesting, but I feel it doesn't give the whole story. That said, I'm not sure it's possible to give the whole story. T_T You can't quantify someone's strategic ability. | ||
|
panschk[FP]
Germany148 Posts
 Finally we can do some serious statistics backups for our imbalance discussions;)And great posts Bill, doodoohead101 and mitsy, it is important to analyse those statistics because not everyone is able to really understand them and they might make BS points by reading statistics wrong -_- While stats depending on rank are interesting too, I think you should use the stats from all games for the evidence-giving stats: - On PGT are not only gosus, so we don`t need the maps balanced only for gosus neither. Balance the maps for the level of play of PGT is best for PGT, obviously. - When considering every game, the sample size is bigger, and a bigger sample size is always good to make quality interpretations. - The level of play on PGT is pretty good anyway, it's not public battle.net or something | ||
| ||
Kung Fu Cup
2026 Week 18
IntoTheiNu
 1048WardiTV822
RotterdaM680
TKL
 288IndyStarCraft
 180SteadfastSC121
Ryung
116Rex113
CosmosSc2
 56[ Submit Event ] |
|
StarCraft 2 StarCraft: Brood War Britney Dota 2  43689EffOrt  1326Mini 1251Shuttle 920Jaedong 876Light  375ZerO 359Snow 273Larva 243actioN 226[ Show more ] League of Legends Counter-Strike Other Games |
|
Replay Cast
The PondCast
OSC
INu's Battles
Percival vs Cure
Classic vs Clem
Replay Cast
Escore
IntoTheTV X SOOP
RSL Revival
herO vs SHIN
Clem vs Solar
Serral vs Rogue
RSL Revival
WardiTV Weekly
[ Show More ] |
|
|
EPT
