| Forum Index > BW General |
|
Mindcrime
United States6899 Posts
On March 31 2012 09:41 Keone wrote: In fact, it wasn't until recently when everyone started saying "Terran Imba". You know why? Everyone went through Swarm season and Six Dragon season, but one person broke everything. You must be new. | ||
|
TwoToneTerran
United States8841 Posts
On March 31 2012 09:28 Lightwip wrote: It's as if you recognize your own argument's faults and grow defensive in denial. Then you flame and flame, with no arguments but that I simply do not understand! How wonderful. http://www.teamliquid.net/tlpd/korean/maps/84_Bifrost http://www.teamliquid.net/tlpd/korean/maps/42_Nostalgia I can't deny, both of these maps are very terran. But the stats still aren't that significant. One had nearly no games on it(and the future iterations of Bifrost are even worse for your case) and the others deviate only slightly from the aggregates. The fact is, Terran gets maps that are hard for them too. Dreamliner, Battle Royal, HBR. And yet after all that, they still have the largest win %. The fact is, terran has the best chance to win, and the best chance to make bonjwas because of an inherent imbalance in their favor, enhanced by multiple favorable conditions. Let's not pretend a map or two in terran's favor made terran the race of bonjwas, because everyone has ups and downs in maps. But one factor that HAS always remained constant is that terrans have a favorable situation in matchups over long stretches of time. Your argument is basically "I know it, I believe it, that settles it." Your argument is that statistically there is a slight terran bias over the course of Broodwar's history, my argument is that yeah, that's true, and it's because the VAST majority of maps made before 2008 or so were atrociously imbalanced in favor of terran. I speak from experience of having to play on these terrible maps and having watched shit like 1 hatch lurker being standard. You named maps made in the last 3 years. Do you have any experience with older maps, the maps that make up the vast majority of your sample size? Not everyone has ups and downs in maps. Zerg was complete fucking trash for years because of awful maps that completely ignored their strengths(It's why Zerg was the weakest race before Savior and has been by far the strongest since, people recognized Zerg strengths and maps became larger and easier to play with more than 2 bases). Also Boxer made the terran bonjwas, as Mortality already plainly told you. Players who were not Terran started playing Terran at a higher frequency, and that happened to contain the smartest and best players up until Savior. The bonjwas are a minor percentage of the games played (Though by their relevance in leagues they played more than the average player) and should not be purely dictated by the overall large sample of statistics you brought forth. It's like saying the Moon's gravity causes us to sway violently like it does the Ocean. You needlessly attribute vast sample sizes to individuals which is a frequent and terrible statistically fallacy. Addendum, your argument was ALSO an ad hominem, saying my points only existed because I root for a specific player. I just think it's pathetic you attribute the great success of the best players in history to "lol terran imba." It's a fluffed up version of that exact same mentality. Confirmation bias at its best. | ||
|
Lightwip
United States5497 Posts
Dunno why you hate on MSLs. I see about as many flukes in the OSL as in the MSL. July and Effort are two rather recent examples. There's also a fairly recent example of a terran-dominated semi there as well. But Zergs do indeed often dominate semis. That's great, they make wonderful fodder for terrans! Sure, there are quite a few zerg-dominated semifinals. I'd argue that zergs are simply very good at being average. Strong ZvP and numbers go a long way to improving results to a marginally higher level, after all. Terrans tend to be either slightly weaker or quite a bit stronger than SL zergs. Good players can, of course, overcome average statistics and get insanely large win records. Yet they play the same race, and therefore have the same problems and advantages as their peers. This certainly influences their results, even if we'd like to think it doesn't. Yet .5-1.5% doesn't make a bonjwa. However, numbers aren't the only things I was talking about. The thing is, everything about terran lets its best players win much more, and that never hurts their chances. I talked about a few of these, including the mirror. Maybe a few percentage points alone don't make it easy for terran to walk away with the most titles and the most bonjwas. But it's part of it, and considering everything else, it's hard to deny that it's quite a bit easier for terran to rise to the top than Z or especially P. On March 31 2012 10:09 TwoToneTerran wrote: You named maps made in the last 3 years. Do you have any experience with older maps, the maps that make up the vast majority of your sample size? Not everyone has ups and downs in maps. Zerg was complete fucking trash for years because of awful maps that completely ignored their strengths(It's why Zerg was the weakest race before Savior and has been by far the strongest since, people recognized Zerg strengths and maps became larger and easier to play with more than 2 bases). I'll address this one point, since everything else has not even an attempt at being civil. The last 4 or so years have about as many games played as everything before that. The volume of games have increased significantly since. This pattern of T>Z>P>T has been like this for years. If anything, it has become more T favored since TvZ used to be about equal in the very early years. | ||
|
sGs.Stregon
United States161 Posts
On March 31 2012 07:04 Lightwip wrote: Because using statistical analysis to try to prove imbalance is whining. Oh, maps matter all right. Yet as an aggregate performance, the effect of map balance is mitigated. Even after accounting for fluctuations, terran is still ever so slightly ahead, and that coupled with a few other factors makes the path to bonjwahood much, much easier. You seem to have read the first part. But yes, maps play a HUGE part in balance in Sc:Bw. You cannot deny it, just because it doesnt work into your statistics, this game is what balance looks like, and the X factor to make one race stronger than another is the map pool. Secondly, Terran may seem a little unbalanced, but like I stated, teams generaly only send Terrans out on maps that are Terran favoured or where they have a slight dissadvantage but have things built into the map they can use to create an advantage((i.e. strategic, hard to reach, cliffs, ledges)), which comes back the map pool.. Sc:Bw is balanced. You can look at any statistic you want and claim any race is inbalanced, but inbalances come from the maps. That is why, generally, on ICCup you see mainly Fighting Spirit only. Because that map is basically 33.3/33.3/33.3% balanced between the races, and player skill determines who wins((or protoss cheese)), as compaired to a map like Outlier, or Neo Medusa | ||
|
Lightwip
United States5497 Posts
Maps change things, but patterns still emerge. | ||
|
shaftofpleasure
Korea (North)1375 Posts
| ||
|
TwoToneTerran
United States8841 Posts
Anyhow, I'd like you to cite exactly how many games have been played since 2008(kind of arbitrary but it's where I'm working from) and how many games have been played before, if you wouldn't mind. I'd like to be sure of this. Also note that literally hundreds of games have not been recorded in TLPD from before 2005 because there were a lot more leagues and showmatchy type deals that didn't stick around, so you are missing some past games. Moving on, T>Z>P>T hasn't been consistent for years. It was P>Z for about a year after Bisu beat Savior, althought it was something to the tune of 51% P winrate as opposed to the typical 53-55% we see in the standard race imbalance. There were also times when Terrans stomped Protoss for several months at a time (Back in Reach's hayday, before Nal_rA really came around, Reach was the only protoss with a modicum of ability to beat anyone). Or directly after Savior where, before Flash's domination, there was like 3 Terran league winners (Flash OSL Mind MSL ForGG MSL) compared to like 4 Protoss and a bazillion zergs. The T>Z>P>T thing is true over a long, large sample size, but not in isolated time periods of relevance (Which is all it takes to win titles). | ||
|
sGs.Stregon
United States161 Posts
On March 31 2012 10:24 Lightwip wrote: Fighting Spirit is a very balanced map, but it's completely standard and an inaccurate representation of the pro scene. Not all games are standard, and some races benefit more than others from certain trends in maps. Maps change things, but patterns still emerge. AND there in lies my ENTIRE point.. Maps determin balance is Sc:Bw anymore.. Terrans dont play on maps that they are statistically weak on - hince Terran's SEEM a little over powered, when in reality, the coach's are just maximizing their line ups, by only playing Terrans on Terran favoured maps, for the most part.. This thread is thinly veiled QQ. So you lose to Terrans, it happens. Watch your replays, and find out where you made your mistakes, and try not making them the next time you play. This is not Sc2, this doesnt need anymore balance((which would probley result in breaking one of the races, to appease those who cant play//insert Sc2 terran comment here\\)) And if you are not sold on maps playing a huge part in balance. There is this thing called a BO win.. If you can accuratly scout what build your opponent is going, you can switch into a build order that beats it, statistically, alot. If you are really going todo a statistics thing, you cant leave out BO wins((which doesnt determine which race is op, but instead which player was smart enough to be able to directly counter what his opponent was doing//insert Flash here\\)) | ||
|
Lightwip
United States5497 Posts
On March 31 2012 11:13 sGs.Stregon wrote: AND there in lies my ENTIRE point.. Maps determin balance is Sc:Bw anymore.. Terrans dont play on maps that they are statistically weak on - hince Terran's SEEM a little over powered, when in reality, the coach's are just maximizing their line ups, by only playing Terrans on Terran favoured maps, for the most part.. This thread is thinly veiled QQ. So you lose to Terrans, it happens. Watch your replays, and find out where you made your mistakes, and try not making them the next time you play. This is not Sc2, this doesnt need anymore balance((which would probley result in breaking one of the races, to appease those who cant play//insert Sc2 terran comment here\\)) You may notice that I specifically said that this analysis has no relevance whatsoever except in the pro scene. Read more carefully. | ||
|
huameng
United States1133 Posts
Also, you claim "Flash, Boxer, Oov, and Nada are also all statistically insignificant" but I don't see why you claimed that. Did you run the numbers? It is certainly not obvious. In the aggregate, they clearly have a huge influence on the winrates -- they've played >800 TvZs between them and won over 66%. I have lots of other problems with this post, but I'd like to start with these. | ||
|
jeremysaint
Canada80 Posts
and i am a little embarrassed that i did not think of it myself!!! | ||
|
sGs.Stregon
United States161 Posts
On March 31 2012 11:18 Lightwip wrote: You may notice that I specifically said that this analysis has no relevance whatsoever except in the pro scene. Read more carefully. You seem willfully ignorant((dumb on purpose)).. You do realize that the Proleague is PRO right? You realize that the Proleague is what im talking about, with coach's maximizing line ups by playing Terrans on Terran favoured maps, right? or are you not willfully ignorant, but mentally stunted =/ Look at all my posts together as a whole, and you would easily realize I am talking about PRO games myself, and simply used a map like Fighting Spirit for my argument of maps creating imbalances in one race or another. | ||
|
Lightwip
United States5497 Posts
On March 31 2012 11:21 sGs.Stregon wrote: You seem willfully ignorant((dumb on purpose)).. You do realize that the Proleague is PRO right? You realize that the Proleague is what im talking about, with coach's maximizing line ups by playing Terrans on Terran favoured maps, right? or are you not willfully ignorant, but mentally stunted =/ I'll let you figure out exactly why your post is so ridiculous. Hint: it involves something you mentioned in your post. | ||
|
sGs.Stregon
United States161 Posts
On March 31 2012 11:26 Lightwip wrote: I'll let you figure out exactly why your post is so ridiculous. Hint: it involves something you mentioned in your post. Mentally stunted it is.. Quit your Terran QQ. Sc/Bw is balanced, and your just bad /thread | ||
|
Tomba
Norway106 Posts
(also tvt is damn hard!) Which leaves less room to practice for the other matchups. | ||
|
puppykiller
United States3126 Posts
| ||
|
GhostOwl
766 Posts
By removing these qualities, a map can work to take away these race advantages, but every map is going to be different, and truthfully, Terran race has units / abilities designed that benefits most from little map features. That itself shows how other races like Zerg / Protoss are weaker than Terran. The race that can abuse the map the greatest is the problem, not the maps itself. And we're not even talking about how flexible a race is (And guess which race has the most flexibility and innovating options? ) User was temp banned for this post. | ||
|
TwoToneTerran
United States8841 Posts
People got into starcraft because Boxer did crazy 1 base micro strategies, so they made more maps with lots of 1-2 base viable shit which benefits Terrans the most, protoss a little, and zerg the least (Zerg was the worst for several years). Years where only every 1 in 5 maps were it was reasonable for a zerg to get their third gas before getting defilers to move out, or protoss had to one base no matter what. | ||
|
Bijan
United States286 Posts
On March 29 2012 15:22 Lightwip wrote: I think the parameters themselves are proof of it, really. ~1% isn't large, but I'd say it's pretty statistically significant. In a fairly weighted contest, the hypothetical terran is the best. I'm no expert, but shouldn't a standard deviation be calculated? Is 1% actually statistically significant? | ||
|
revy
United States1524 Posts
1) You claim that long term averaging negates the need to do a more thorough job on the time-dependence of matchups. Yes, the power of averaging is great but using a filter window size of "The history of Broodwar" doesn't give a whole lot of chance to see how things may have changed over the years. There really does need to be a much higher bandwidth (shorter time) filter applied to the statistics and have that scrolled over the history of broodwar. Part of what I do for a living is design control systems, I can do all of this except for the data mining portion. It's really something for you to attribute Flash's bonjwa status to the fact that he's terran when we don't even know what the time dependence of the matchups look like. The fact that Oov pwned zergs so bad back in the day has very little relevance to the TvZ statistics right now except for the contributions he made to the base understanding of the matchup. The players now are playing a different game than Oov played. 2) The proleague problem. I have a feeling proleague statistics will exacerbate the imbalance in XvX statistics due to the fact that Coach's will try to put good players out on maps that are good for them (presumably good for their race). The result is that you'll see a lot of Stork on P favored maps (insert any of TBLS, just picked one at random) rolling over whatever fodder the other team sends out. This will show up as racial imbalance when it's truly skill imbalance which is causing the divergence of matchup statistics. There probably should be a separate study carried out splitting out team league statistics from individual statistics. I believe individual statistics should be more accurate as players typically have to play on a more balanced mapset. 3) The effect absolute player skill has on imbalance. This one is unfortunately probably impossible to figure out. The argument goes like this, just because ZvP is favored at 54-46 for Z for the set of all pro gamers does not mean that this is true at all levels of player skill. If you go down to D level, what does the racial imbalance look like down there, I'm willing to bet it's different than for an average pro level. Who cares what the balance looks like at the D level, who cares what it looks like at the average pro level, what is most important is what does it look like at the very top. As you try to focus in on only the top players you run into a problem of not having enough data to make any certain predictions. Beyond that you face the task of quantifying a player's absolute skill. In short the effort is somewhat futile, all we really care about is what does racial balance look like at the uppermost prolevel and how it changes over time. Throw in the fact that we don't know a player's absolute skill level and it really gets messy. Add for good measure the feedback loops that good players have on the pro-scene (I don't believe it's coincidence that we have things like swarm season after JD dominance, the era of the dragons, or that one MSL where terrans got 3/4 semi's slots after Flash's 2010). There's also the unknown feedback effect that Boxer had on the whole scene. The first Bonjwa terran emerges, many people look up to him and select T as their race and then go on to do good things themselves. Were most of the bonjwa terran due to an imbalance, or because the first bonjwa was terran and they all copied (this ignores the fact that Boxer actually counts as 2 bonjwas because of his protoge' Oov). On March 31 2012 10:16 Lightwip wrote: A few fair observations, along with a few not-so-fair ones and a whole lot of rage. Dunno why you hate on MSLs. I see about as many flukes in the OSL as in the MSL. July and Effort are two rather recent examples. There's also a fairly recent example of a terran-dominated semi there as well. But Zergs do indeed often dominate semis. That's great, they make wonderful fodder for terrans! For the hell of it I ran the MSL numbers for the last 11 seasons: So yes, include the MSL and you don't have to go back as far to find more than 1 T in a semis. But that doesn't change the fact that there still have been more toss in semi-finals than terran in the past 11 seasons. | ||
| ||
StarCraft 2 StarCraft: Brood War Stormgate Dota 2 League of Legends Counter-Strike Super Smash Bros Heroes of the Storm Other Games summit1g4812 ScreaM3378 FrodaN1829 Dendi1468 Beastyqt1136 B2W.Neo603 elazer383 JuggernautJason363 Fuzer  328shahzam306 Livibee226 Trikslyr83 QueenE44 |
|
Replay Cast
SOOP
SKillous vs Spirit
Tenacious Turtle Tussle
PiG Sty Festival
Serral vs TriGGeR
Cure vs SHIN
The PondCast
Replay Cast
PiG Sty Festival
Clem vs Bunny
Solar vs Zoun
Replay Cast
Korean StarCraft League
PiG Sty Festival
herO vs Rogue
ByuN vs SKillous
[ Show More ] |
|
|
EPT
