|
On May 21 2011 09:23 Sentient wrote:Show nested quote +On May 21 2011 05:56 VIB wrote:
No they don't. Did you ever read any book on game design? Balancing is a very simple process. It's a brute force repetition of trial and error. Nothing more than that. It's pure intuition. Pure guess work. There's zero science. There's zero math. Blizzard has no statisticians at all. I agree with your sentiment though. I don't think posts like these are that helpful. Balance is about perception and gut instinct, not about statistics. It's about game design -- losses should feel "fair". From an entertainment perspective, the breakdown of win percentages is less important than whether or not losses feel like the game has conspired against you. As someone else mentioned, mirror matches can feel incredibly unfair at times, which hints at balance problems even though the win percentage itself is exactly balanced. tldr: Balance is a subjective opinion and is reached by consensus, not by statistics. Balance can be objective, it doesn't need to be done with only intuition. I was only saying that how games are made today, it is made with intuition, no one is trying to make a perfectly mathematically balanced game. And SC2 certainly isn't any different. Instead of brute force test-patch iterations blizzard does. We could have built a game designed around a balanced model. But SC2 would have to be redone from scratch for that to happen.
Other than that I do agree with what you're saying. Statistic inference is worthless balancing. And above anything else, before we start any objective discussion about balance, we must first define what "balance" means. There's a lot of disagreement to what a balanced would be. Most people arguing about balance are comparing apples to oranges and thus won't get anywhere.
|
On May 21 2011 04:11 Sleight wrote:Show nested quote +On May 21 2011 03:28 Jombozeus wrote:On May 21 2011 03:17 sylverfyre wrote:On May 21 2011 02:59 infinity2k9 wrote:On May 20 2011 15:00 avilo wrote:
Just because someone made an incredibly long post does not make it mega awesome or even remotely accurate.
The most obvious thing completely wrong with the "post" is to not look at pros for balance. In any RTS game or game you always look at the top level for balance because these are the people playing the game at the highest level and are actively trying to "break the game."
Yeah seriously, not going to quote this all but i don't see why people think this is so particularly great just because it's a shitload of words. By this logic BW isn't balanced well enough cause PvT is easier for P for 99% of people. I can't be bothered to go into this any deeper cause this whole thing is basically useless and something that could have been said in about 100 times less words. But the problem with the pros is that they're inherently outliers. We don't exactly have a magic number attached to every pro showing their skill (or even their skill in each matchup) even when we attempt to model it through Elo ratings and such. Statistically the pros are going to always have very odd looking win ratios, and it's extremely hard to draw conclusions from them. Hence the conclusion drawn should be that statistics is not a good way to measure imbalance, NOT that pros are not a good way to draw conclusions FOR statistics. Since we are discussing imbalance at the maximum potential, the statistical outlier is the prime consideration, not to be ignored. To assume that imbalance is equal at any level is absurd as previously stated, the skill:winrate ratio does not scale linearly. If we cannot use statistics, what can we use as a metric to examine data? There is ONLY statistics. Within the model presented, the OP does a great job of supporting and defining his argument. Everyone saying imbalances vary at skill levels have a VALID point, but that doesn't make it true. If every Bronze Z loses to 3 Rax an imbalance metric of 2, and every Master Z loses to 2 Rax an imbalance metric of 2, then the imbalance value would be the same overall, assuming the MU was otherwise in harmony, for sake of an argument. The methodology presented argues this: If a race is overpowered against another race, it should exist at a similar level regardless of direct causation or mechanism across all levels for purposes of general game balance. (ie different means at different levels but same net result of imba)What it does NOT argue is this: Racial imbalance is uniform in mechanism across the spectrum (ie 2 rax is always the cause of OP). This approach to balance allows for exactly one thing: Identification and stratification of the most gross (meaning large) imbalances in a game for presence alone. Why such an imbalance is present is up to debate. This means that he game CAN BE balanced as the author proposes at the largest scale and that tweaks in units and fine mechanics must be trade offs in overall power to solve issues at the highest level. Well done, OP. Very neat read.
And why exactly would you think the "methodology presented" is in fact correct? I think most people would argue that that is totally false in fact. For an extreme counterexample, imagine automaton 2000 playing TvZ. I don't think Zerg could ever beat marines with flawless micro. Unless you want to imply that a human skill cap actually creates the situation described, but it seems a rather arbitrary perspective.
Balancing SC2 around statistics would be a rather silly approach, especially considering that we want more from SC2 than simply a 50% win rate among all races at a pro level. Coin flipping is a fair game but it's not a very interesting one. Using anecdotal evidence from players while at the same time considering their potential bias is really the most logical way to go about balance in my opinion. SC2 could be a fair game in the sense that every race is competitive at a pro level but there can still exist fundamental problems that would be hard to reveal through pure statistics.
|
any1 else think this guy is secretly a blizzard employee feeding the TL community information?
nobody does this much statistical analysis and writing just to make a 1st post on teamliquid, unless they have a hidden agenda.
conspiracy.
|
I understood pretty much all of it. The only thing I disagree with is that you can reference balance to sub diamond players. In those leagues statistics are irrelevant. You'd have to define a secondary function relating player skill to league, else it doesn't work.
Overall though, tremendous post, its saturated with awesome stuff, I love your applications of math to the game. This is truly a lot to handle, must have taken you so much time to make.
Thanks for posting.
|
Our statistics from GSL would show MKP and MVP winning most of their games against other terrans. Conclusion? It’s not terran that’s OP, it’s the players. That assumes direct correlation between the skill of a player in different matchups, which isn't even the case.
I think for the sake of analysis the game should be looked at as 9 different games, and players to have different ELO for each matchup.
Also, perhaps a revolutionary thought:
What if the endless debate about (im)balance is in the flaw of the whole concept of competition.
Or more precisely, the attempts to quantify the efforts involved in heterogeneous competition (which coincidentally is all human competition, because we are different). It's like measuring a bird flying and an elephant running on some unified scale of effort. It doesn't work. The abilities and the efforts are simply unique to each individual.
I don't know why it has never been taken seriously to have pro-leagues of only random players, or even better - that you have to play a game of each matchup once. In chess, they play turns, one with white, one with black, and so forth. In football they take turns playing home and away. Only in Starcraft you get this weird social experiment where a swimmer and a runner compete, with the pool liquid density, and the runner's shoe weight, being adjusted to find some elusive concept of balance.
Obviously, this process is doomed to be always corrupt and disputed, because it can't be objectively measured. But I guess that's the hook - the hook is this paradox of perpetual injustice. A game which is perfectly just is also usually solved and uninteresting. But hook them up with endless dynamic injustice, and they will play it and play it, till the end of the world. (:<
Side note regarding the argument about gender equality:
+ Show Spoiler +A summarised report on college admission rates show that 40% of college entrants are female and 60% of college entrants are male. Is this “good enough”?
To the untrained eye, the numbers look pretty close to 50%. But in reality this is a massive difference. The true numbers crop up when we compare them with each other. So out of 100 entrants, 40 are women and 60 are men. That means 150% as many men enter college as women. When presented that way, the magnitude of the difference becomes much more evident and it becomes clear there is sexism problem with the college admission system. I'm a very strong believer in equality for both sexes, however this case does not necessarily mean injustice of the admission tests, but, for example, injustice of the whole cultural environment and educational process. It's hard to twist it that tests in mathematics are sex-biased inherently by their design - but rather our culture as a whole does not manage to operate in such way that both sexes develop equal cognitive mathematical ability.
|
The OP was right when he said in introduction that this kind of threads are destroyed by nitpickers 
(or people that made their idea based on what they want, then argue only on things that favor their point)
|
On May 21 2011 09:17 ploy wrote:Show nested quote +On May 20 2011 15:00 avilo wrote:Just because someone made an incredibly long post does not make it mega awesome or even remotely accurate. The most obvious thing completely wrong with the "post" is to not look at pros for balance. In any RTS game or game you always look at the top level for balance because these are the people playing the game at the highest level and are actively trying to "break the game." Really all the OP is saying (but ironically not doing in many of his examples in the OP) is: "don't be biased with your balance judgements." Nothing new...and there's no need to go into "intricate mathematics" or math at all for any of this...the OP is overcomplicating things, and likely has not enough experience to legitimately comment on balance or imbalance in the first place. The most qualified people to talk about balance are the pro players and people high up on ladder that are playing the game everyday versus other good players. But 99% of these players are trying to practice and improve themselves and not even worry about balance in the first place, though everyone QQ sometimes. imo OP is just trying to re-invent the wheel on balance discussions aka having a discussion about how to discuss things lol...there's about one of these posts per month or so that pop up with some guy that thinks he's mega smart and mystical with "the maths" -_- there's just so many things wrong in the OP and ironically "biased." Do we really need another thread discussing how we should be discussing things and hordes of low post count people going, "wow you're so smart and amazing."  Nice effort sure...but i think a bit misplaced. Also, the entire premise of the thread doesn't work because there is no definitive model for imbalance. The model everyone uses for imbalance is...guess what? Their personal bias and opinion. Notice my use of italics for emphasis. Funny coming from the biggest balance whiner around... only about their own race, of course.
If you have nothing to say then don't talk at all. Don't be one of "those" people that reference random things from months ago that aren't relevant. Did you just come into this thread to do that? Where in this thread did I utter a word about what I (personally) think of the balance of the game?
|
Excellent work but you're completely ignoring the fact that Activision Blizzard is a corporation, which exists to generate profit and to raise the value of their stocks. You seemed to have spent a lot of time on "affinity" section (as well as advice for would-be pros), but I think that's clearly misleading, at least until SC2 sales (including upcoming 2 expansions) plateau.
It's clear that Blizzard sees the future (i.e. money) in e-sports, but that is, well, future. Their revenue comes from the sale of games by and large. So I wasn't surprised at all the introductory race, the Terrans, to have most-finished look and it has been the most dominant race at release. (compare it with zerg's artwork and non-existent sound/voice acting) So the race that new players are most likely to play has been also the strongest race, unlike your "affinity" claim where less popular race would be more likely to be treated favorably by patches. Why? I strongly suspect that Blizzard (or any sane corporation who is obliged to make profit for shareholders) did not want to frustrate the general public market who just finished tasting SC2 by finishing single player campagin then jumped into the sea of Battle.net for some live action.
When each expansions hit, the number one priority for Blizzard is profit. Everything else can wait. Once all the expansions are out and sales become complacent the balance will take the first seat. And even then, you really can't trust a corporation to behave in a completely neutral manner at the cost of making money. For example, Blizzard released the first patch in years for WarCraft 3 (no one thought any more patch is coming for that game), and interestingly the patch buffed "Human" race after all those years. At the same time we learn that WC3 is rocking in China and Human race is by far the most popular race there. Does it get you thinking?
|
On May 21 2011 05:46 Sleight wrote:Show nested quote +On May 21 2011 04:54 Jombozeus wrote:On May 21 2011 04:11 Sleight wrote:On May 21 2011 03:28 Jombozeus wrote:On May 21 2011 03:17 sylverfyre wrote:On May 21 2011 02:59 infinity2k9 wrote:On May 20 2011 15:00 avilo wrote:
Just because someone made an incredibly long post does not make it mega awesome or even remotely accurate.
The most obvious thing completely wrong with the "post" is to not look at pros for balance. In any RTS game or game you always look at the top level for balance because these are the people playing the game at the highest level and are actively trying to "break the game."
Yeah seriously, not going to quote this all but i don't see why people think this is so particularly great just because it's a shitload of words. By this logic BW isn't balanced well enough cause PvT is easier for P for 99% of people. I can't be bothered to go into this any deeper cause this whole thing is basically useless and something that could have been said in about 100 times less words. But the problem with the pros is that they're inherently outliers. We don't exactly have a magic number attached to every pro showing their skill (or even their skill in each matchup) even when we attempt to model it through Elo ratings and such. Statistically the pros are going to always have very odd looking win ratios, and it's extremely hard to draw conclusions from them. Hence the conclusion drawn should be that statistics is not a good way to measure imbalance, NOT that pros are not a good way to draw conclusions FOR statistics. Since we are discussing imbalance at the maximum potential, the statistical outlier is the prime consideration, not to be ignored. To assume that imbalance is equal at any level is absurd as previously stated, the skill:winrate ratio does not scale linearly. If we cannot use statistics, what can we use as a metric to examine data? There is ONLY statistics. Within the model presented, the OP does a great job of supporting and defining his argument. Everyone saying imbalances vary at skill levels have a VALID point, but that doesn't make it true. If every Bronze Z loses to 3 Rax an imbalance metric of 2, and every Master Z loses to 2 Rax an imbalance metric of 2, then the imbalance value would be the same overall, assuming the MU was otherwise in harmony, for sake of an argument. The methodology presented argues this: If a race is overpowered against another race, it should exist at a similar level regardless of direct causation or mechanism across all levels for purposes of general game balance. (ie different means at different levels but same net result of imba) What it does NOT argue is this: Racial imbalance is uniform in mechanism across the spectrum (ie 2 rax is always the cause of OP). This approach to balance allows for exactly one thing: Identification and stratification of the most gross (meaning large) imbalances in a game for presence alone. Why such an imbalance is present is up to debate. This means that he game CAN BE balanced as the author proposes at the largest scale and that tweaks in units and fine mechanics must be trade offs in overall power to solve issues at the highest level. Well done, OP. Very neat read. Contrary to popular belief, anecdotal evidence from pros usually do. Statistics is the only metric to measure data? Since when have we concluded that a metrics is necessary? The assumption you make with the net result is absolutely preposterous. Its grossly abusing inductive reasoning. The stats themselves show that at different levels, the win% of different races is different in each matchup. I don't understand how you can convince yourself that is a valid argument. As there are still those who have not realized, identification of short term "imbalance" is easy with statistics, we say "hey, we see 55% winrate over terran as zerg at X point master level, hence zerg is more imba than terran at X point master level." That does NOT mean: 1. Zerg is imbalanced compared to terran at all levels 2. Zerg will exhibit the same winrate vs. terran tomorrow, next week, or next month due to a new paradigm shift 3. Zerg players and terran players will exhibit the same level of increase in general skill at the same rate 4. We shouldn't listen to IdrA because of his cognitive bias towards the zerg race 1,2,3 are assumptions that lapses in logic, while 4 is a conclusion the OP made with the utmost lack of respect for pro gamers. Would you go up to a scientist and tell him: "Hey, I know you're a scientist, but because you have cognitive bias I don't believe you should be able to make conclusions about science."PS: Short term can be as short as an infinitely small amount of time YES! That's exactly the point. You DO say that in the current scientific community. There is a reason paper publication works like it does. Peer review is established so that one person's findings have to hold up to expert in the same area who have NO PERSONAL GAIN. Your quote is EXACTLY why we don't use a single lab's results or a single paper. No one cares if you feel that way if it doesn't hold up to other un-invested parties. So let's pull IdrA a ZvT expert's statment against more ZvT experts AND TvZ experts, and see if they all hold up. Soon enough we are sampling a monster pool and back at statistical analysis.
You're carrying this analogy out of proportions lol.
Scientists always come up with conclusions. You are required to come up with a conclusion in a scientific paper. Of course if IdrA says something, there will be other pros to argue with or against him, but no one will say his opinion is invalid because of bias.
To all these new posts:
Go is different from Starcraft. Go has an astronomically high amount of possible moves, but there are only two variables. There are thousands of variables in Starcraft with equally astronomically high amount of possible moves.
The variable for go is the black and white piece, while Starcraft has many different pieces, hence the analogy is a bit flimsy. You can theoretically get Go to be played at a 100% level given all the responses possible with a supersupersupersupercomputer that doesn't exist, because it has a finite amount of possible moves.
You can never get a computer to beat a player 100% of the time. If you are to do that, freedom of information must be given to the computer, or in other words, fog of war must be removed. Go gives perfect information.
|
mind blown.
this was an excellent read, totally agreed!
|
I have more analysis on specific mechanics within SC2 but am not sure if anybody would be interested in seeing them.
The more the merrier. I love mathcraft, even if its theory mathcraft. Thanks for all the effort you put into it.
|
On May 21 2011 11:06 Jombozeus wrote:Show nested quote +On May 21 2011 05:46 Sleight wrote:On May 21 2011 04:54 Jombozeus wrote:On May 21 2011 04:11 Sleight wrote:On May 21 2011 03:28 Jombozeus wrote:On May 21 2011 03:17 sylverfyre wrote:On May 21 2011 02:59 infinity2k9 wrote:On May 20 2011 15:00 avilo wrote:
Just because someone made an incredibly long post does not make it mega awesome or even remotely accurate.
The most obvious thing completely wrong with the "post" is to not look at pros for balance. In any RTS game or game you always look at the top level for balance because these are the people playing the game at the highest level and are actively trying to "break the game."
Yeah seriously, not going to quote this all but i don't see why people think this is so particularly great just because it's a shitload of words. By this logic BW isn't balanced well enough cause PvT is easier for P for 99% of people. I can't be bothered to go into this any deeper cause this whole thing is basically useless and something that could have been said in about 100 times less words. But the problem with the pros is that they're inherently outliers. We don't exactly have a magic number attached to every pro showing their skill (or even their skill in each matchup) even when we attempt to model it through Elo ratings and such. Statistically the pros are going to always have very odd looking win ratios, and it's extremely hard to draw conclusions from them. Hence the conclusion drawn should be that statistics is not a good way to measure imbalance, NOT that pros are not a good way to draw conclusions FOR statistics. Since we are discussing imbalance at the maximum potential, the statistical outlier is the prime consideration, not to be ignored. To assume that imbalance is equal at any level is absurd as previously stated, the skill:winrate ratio does not scale linearly. If we cannot use statistics, what can we use as a metric to examine data? There is ONLY statistics. Within the model presented, the OP does a great job of supporting and defining his argument. Everyone saying imbalances vary at skill levels have a VALID point, but that doesn't make it true. If every Bronze Z loses to 3 Rax an imbalance metric of 2, and every Master Z loses to 2 Rax an imbalance metric of 2, then the imbalance value would be the same overall, assuming the MU was otherwise in harmony, for sake of an argument. The methodology presented argues this: If a race is overpowered against another race, it should exist at a similar level regardless of direct causation or mechanism across all levels for purposes of general game balance. (ie different means at different levels but same net result of imba) What it does NOT argue is this: Racial imbalance is uniform in mechanism across the spectrum (ie 2 rax is always the cause of OP). This approach to balance allows for exactly one thing: Identification and stratification of the most gross (meaning large) imbalances in a game for presence alone. Why such an imbalance is present is up to debate. This means that he game CAN BE balanced as the author proposes at the largest scale and that tweaks in units and fine mechanics must be trade offs in overall power to solve issues at the highest level. Well done, OP. Very neat read. Contrary to popular belief, anecdotal evidence from pros usually do. Statistics is the only metric to measure data? Since when have we concluded that a metrics is necessary? The assumption you make with the net result is absolutely preposterous. Its grossly abusing inductive reasoning. The stats themselves show that at different levels, the win% of different races is different in each matchup. I don't understand how you can convince yourself that is a valid argument. As there are still those who have not realized, identification of short term "imbalance" is easy with statistics, we say "hey, we see 55% winrate over terran as zerg at X point master level, hence zerg is more imba than terran at X point master level." That does NOT mean: 1. Zerg is imbalanced compared to terran at all levels 2. Zerg will exhibit the same winrate vs. terran tomorrow, next week, or next month due to a new paradigm shift 3. Zerg players and terran players will exhibit the same level of increase in general skill at the same rate 4. We shouldn't listen to IdrA because of his cognitive bias towards the zerg race 1,2,3 are assumptions that lapses in logic, while 4 is a conclusion the OP made with the utmost lack of respect for pro gamers. Would you go up to a scientist and tell him: "Hey, I know you're a scientist, but because you have cognitive bias I don't believe you should be able to make conclusions about science."PS: Short term can be as short as an infinitely small amount of time YES! That's exactly the point. You DO say that in the current scientific community. There is a reason paper publication works like it does. Peer review is established so that one person's findings have to hold up to expert in the same area who have NO PERSONAL GAIN. Your quote is EXACTLY why we don't use a single lab's results or a single paper. No one cares if you feel that way if it doesn't hold up to other un-invested parties. So let's pull IdrA a ZvT expert's statment against more ZvT experts AND TvZ experts, and see if they all hold up. Soon enough we are sampling a monster pool and back at statistical analysis. You're carrying this analogy out of proportions lol. Scientists always come up with conclusions. You are required to come up with a conclusion in a scientific paper. Of course if IdrA says something, there will be other pros to argue with or against him, 1[1]but no one will say his opinion is invalid because of bias.To all these new posts: Go is different from Starcraft. Go has an astronomically high amount of possible moves, but there are only two variables. There are thousands of variables in Starcraft with equally astronomically high amount of possible moves. [2] The variable for go is the black and white piece, while Starcraft has many different pieces, hence the analogy is a bit flimsy. You can theoretically get Go to be played at a 100% level given all the responses possible with a supersupersupersupercomputer that doesn't exist, because it has a finite amount of possible moves. [3]You can never get a computer to beat a player 100% of the time. If you are to do that, freedom of information must be given to the computer, or in other words, fog of war must be removed. Go gives perfect information.
[1] Incorrect. That is EXACTLY what you will say. Listen to the first balance conversation Day[9] had on SOTG after SC2 officially released for the exact methodology of explaining this. Effectively, it comes to this; You, meaning every single person, cannot have an unbiased opinion in any meaningful way. Results are based on data and anyone who believes data does not lie has never worked on major publications in the sciences. The gold standard is a Peer-Review paper because ONLY IF other people find it not just plausible, but recreatable and demonstrable, is it acceptable. If you say ZvT is imbalanced, you have to show the exact situation and go thru and show each moment of contention in context. Data is meaningless unless it is unassailable.
[2] Incorrect. You literally have never played Go to say this. Playing White vs Black is one of the infinite variables. Every move in Go after the opening series has between 50-250 possible followups. Like Chess, Go has unbelievable depth of strategy. There are positional imbalances that change and grow as the game progresses, their are space advantages, their is the advantage of playing first wiht black, etc. Variables mean values subject to change. Every single piece in Go is subject to change within a few short moves.
[3] Incorrect. You could easily build a computer to beat a human 100% of the time in a game of incomplete information given the right definition of "winning." Ever played against a good Poker AI? They obviously have perfect calculation of hand value, don't tilt, and a perfect memory. Furthermore, there is variability inserted into its play that makes it very difficult to read while you yourself are an open book after the first 30 hands. Heads up? Yeah people can sometimes get super lucky and beat the computer before it compiles a useful hand history, but at a table? A computer always wins given enough time.
Starcraft is the same idea. There are strictly superior strategies with perfect execution, we just havent identified them, then you script down an AI to execute the most powerful tactics blended with just the right timing sense and variability, I bet it could beat MC, Nestea, or MVP forever. In a single event/game/hand? Sure a human might win. But play a BoX and as X grows, computers will dominate, as the stack size expands relative to the blinds, poker AI's wipe the floor with humans.
|
Very interesting read to the OP, I'd be interested to see what conclusions you come to
|
1. lol. Don't go all smartypants and explain what a peer review is, everyone knows. You can disprove a scientist with peer review, but you cannot say his opinion is invalid. I don't see the concept that is so hard to grasp. You put IdrA and Cruncher in a room to have a debate, they will give you different answers on balance but their opinions are equally valid. You put a Keynesian economist and a neo-classical economist in a room and the same thing happens.
Peer review is used to prove (to a certain extent) and disprove a theory raised by a scientist, to see if the scientist is in essence right or wrong. The scientist can be wrong, but hes entitled to make his own conclusion. No one will say his opinion is invalid because of bias. I stand by that. You are taking this analogy and twisting in your own words. His opinion may be wrong because he has made a mistake or was outright lying about the data, but no one will say "I reject your hypothesis because you have bias." Not at least until there is sufficient proof that it is the case. Simple: A man is innocent until proven guilty.
2/3. I used to play Go at a semi-competitive level when I was young, don't make wild assumptions. White vs. Black are two variables. You can put in a white or a black piece, and the variable you can plug in is the place where you place the piece. You can map out an entire game of Go with two variables just like you can do with chess.
You cannot do this with Starcraft, and hence the impossibility of bruteforcing a build. The amount of variables, including the lack of information, will always result in some kind of coin flip. For example, after a prepatch 4gate encounter, the computer techs to colossus, while you tech to stargate. He does not know there is a stargate because he is denied this scouting information by vigilant stalker/detection micro. A perfectly microed colossus will never beat a phoenix.
Starcraft is similar to poker, but poker also has a finite number of moves. I say 100% and you say a poker program can beat the player 99% of the time. 99% is so much different from 100%. The whole point is the approach towards infinite, not how large that number is.
|
roooofl, don't want to derail or anything, but there is no good AI poker bot except for perhaps heads up games. Other than that... not even close.
Also, you are oversimplifying chess/go. If you could map out an entire game of go/chess so easily, then the games would have already been "solved" by computers.... which computers are not close to doing at all as of right now.
Edit: On second thought, there are definitely no good AI poker bots. The only bots that can play heads up well are based purely on game theory, which means whatever the other player does means nothing to the bot. No AI involved, just game theory (simplified forms of heads up poker have been solved by game theory).
|
On May 21 2011 14:15 ploy wrote:
roooofl, don't want to derail or anything, but there is no good AI poker bot except for perhaps heads up games. Other than that... not even close.
Also, you are oversimplifying chess/go. If you could map out an entire game of go/chess so easily, then the games would have already been "solved" by computers.... which computers are not close to doing at all as of right now.
Edit: On second thought, there are definitely no good AI poker bots. The only bots that can play heads up well are based purely on game theory, which means whatever the other player does means nothing to the bot. No AI involved, just game theory (simplified forms of heads up poker have been solved by game theory).
The amount of processing power in our supercomputers is a fraction of what is needed to solve every move of go.
|
On May 21 2011 14:15 ploy wrote:
roooofl, don't want to derail or anything, but there is no good AI poker bot except for perhaps heads up games. Other than that... not even close.
Also, you are oversimplifying chess/go. If you could map out an entire game of go/chess so easily, then the games would have already been "solved" by computers.... which computers are not close to doing at all as of right now.
Edit: On second thought, there are definitely no good AI poker bots. The only bots that can play heads up well are based purely on game theory, which means whatever the other player does means nothing to the bot. No AI involved, just game theory (simplified forms of heads up poker have been solved by game theory).
I think computers in chess are pretty damn good, they can beat or do draws against the top players, But in Go it's impossible for a computer to match a top player. Go is a lot more complicated than chess.
|
Wonderful post from the OP. I would like to add some more food for thought and would love it if other users could respond to my comments -
Have you considered the GSTL race matching preferences? Is there a trend in what race certain teams prefer to matchup against other races/teams? Is this pattern entirely player dependent or race dependent? even on a sub-conscience level is there a pattern of a certain race being more favored against other races?
Even if the OP has not considered GSTL I do think, racial imbalances aside, other very interesting patterns regarding player styles, reputations etc can result in very interesting patterns of what a team plays against specific players/races.
|
I totally agree, I think balance should not be left to the community because we have to admit we are all biased towards a specific race. If you ask any pro player they will usually feel like their race is the weakest and no one is going to say their race is the strongest except maybe Genius because he is cocky.
Blizzard needs to stop patching and listening to pro players/community qq and look at the results/games themselves devote some sort of branch in their company that handles balance for sc2 and have them study/analyze games themselves. People who work for their company in order to better develop a game are generally not biased and should be credited.
|
wow what a smart and amazing first post
|
|
|
|
|
|
EPT
