|
On August 20 2015 14:12 unifo wrote: How much of an advantage playing random really is?
Where is the STATISTICAL evidence to support the advantage or disadvantage of being random? I would like to see current stats of random players win rates before really diving deeper into this issue. What kind of statistical evidence would you even want? If they're in ladder, their win rate is probably 50% because if it weren't, they would rise or fall on the ladder until it was. That is no different for random players. You could compare their mmr with random to their mmr with their best race, but that wouldn't reflect random advantage; that would reflect how good their off race matchups are.
Now if you took two players with the same mmr who perform comparably in, say, PvP, then had one queue with their main race while the other queued as random, and then only looked at the PvPs, you could draw conclusions about the strength of the random advantage in PvP. I don't think anybody has done that kind of analysis, nor does anyone have the resources to do so.
|
On August 20 2015 14:27 JimmyJRaynor wrote: i play as Random because its the most fun. to add to my advantage sometimes i'll lie to my opponent and act "manner" by immediately telling him my race. but, i'm lying to my opponent. the additional advantage is higher when my opponent believes i'm being honest. As a proud Random player, I think you should be ashamed of yourself. This kind of behavior gives all Random players a bad name.
|
On August 20 2015 14:27 ChristianS wrote:Show nested quote +On August 20 2015 14:12 unifo wrote: How much of an advantage playing random really is?
Where is the STATISTICAL evidence to support the advantage or disadvantage of being random? I would like to see current stats of random players win rates before really diving deeper into this issue. What kind of statistical evidence would you even want? If they're in ladder, their win rate is probably 50% because if it weren't, they would rise or fall on the ladder until it was. That is no different for random players. You could compare their mmr with random to their mmr with their best race, but that wouldn't reflect random advantage; that would reflect how good their off race matchups are. Now if you took two players with the same mmr who perform comparably in, say, PvP, then had one queue with their main race while the other queued as random, and then only looked at the PvPs, you could draw conclusions about the strength of the random advantage in PvP. I don't think anybody has done that kind of analysis, nor does anyone have the resources to do so.
Yeuch. Random is clearly underrepresented in GM league. Random is extremely underrepresented in tournaments.
The underlying concept of ladder matchmaking would actually lead to random players being overrepresented in higher leagues and underrepresented in lower leagues - IF there was indeed a random advantage. We actually witness the opposite.
If we wanted 'balance', random players should be given extra minerals at game start until their - let's say Code S - representation mirrors the amount of random players in the game overall.
Highly ironic thread looking at your sig.
|
Russian Federation80 Posts
On August 20 2015 12:53 ChristianS wrote: Does anyone know whether tournaments even allow pros to select random? It seems unlikely that NO ONE would have tried it. Plenty of pros have pretty solid off-races and on some maps the advantage is pretty strong. TLO used to be strong with all three races, and in something like a Bo5 it would seem strong to throw your opponent off if nothing else.
There are randoms, btw if you look carefully at liquipedia tourney pages, they have race distribution on the bottom, and yes it has random there. Look http://wiki.teamliquid.net/starcraft2/Random, not so many random players out there to be honest, and there still are some, so it's not like it's forbidden.
|
Russian Federation80 Posts
On August 20 2015 13:30 ChristianS wrote:Show nested quote +On August 20 2015 13:09 BluemoonSC wrote: why even bother with an option to random if it reveals their race at the start? on 4 player maps, yes scouting is difficult. but learning 9 matchups is also difficult.
you just might have to play a safer build until you can scout. honestly, 1 worker at the start is worth the intel on what race they are and what early indicators of their build they have. Alright let's cut the bullshit. I'm not sure anyone is reading this far into the thread but let's try it. Nobody knows what the metagame will look like a year from now, particularly since the design is still up in the air, so imagine a hypothetical which I think isn't too implausible. Imagine that PvP is at a point where everything is adept versus adept, and if you don't go adept you lose (isn't that more or less where we are now?). Because of this the most popular build is to take gas more or less immediately in order to start on your adepts and warpgate as soon as possible. The later you take gas, the further behind you get. Meanwhile imagine one of the other matchups becomes extremely macro focused. The most popular build by far is nexus first, followed by a gasless 1 gate expand. If you take early gas in this matchup you're at least moderately behind. Even if the Protoss player sends a worker right at the beginning to scout, should they take gas immediately? In one of three cases they're dead without it, while another puts them behind with it. It's not even rock paper scissors, because even if they could mind game their opponent perfectly they'd still have no way to predict their race. It's a pure dice roll. Now maybe nothing like this hypothetical metagame will ever happen, but there's no reason it can't. With random advantage left in the game the imbalance can range from miniscule to game-breaking just depending on the whims of the metagame. Even if you think random players should have an in-game advantage for being willing to play all nine matchups, why choose such a weird and unpredictable one like this? If you just made their units cost slightly less, or shortened their build times by 1%, or let them spawn ten seconds earlier, you could achieve the same effect (giving random players an edge to make up for having to learn so many matchups) without potentially game-breaking situations coming up.
The point of playing against random is actually not being able to play "the most popular builds", but playing the build which may fit against all races. I'm gonna shoot my leg now I guess, but I am very bad as P, so I don't really know its race early game problems all that well, but I still watch quite a few P streams.
So, let's get back to the point itself. I am pretty sure that P can play something like 13 gate 14/15 gas against any race right now in HoTS, followed by cybercore. How much time players needs to find opponents base? I assume it's fairly short time on majority of 2 player maps: smth around 30 seconds (I'm still using HoTS time though due to habit), and it's just over the time pylon takes to build. Perhaps 30s is way too low, let it be 60. It's gateway half-way done, if you scout with a probe building a pylon, and it is absolutely acceptable from my point of view.
Now things get worse on 4 player maps, that is true. You can still veto those maps btw, but ye last position scout is bad luck here. I wouldn't be surprised if map pools had purely 2p maps upon LoTV release. Even here you forget a very reasonable point, that even against specific opponent race scouting last can be devastating. This itself is bad, but my own feeling is that in LoTV this "advantage" of random is even smaller then that in HoTS.
|
On August 20 2015 15:19 Aiobhill wrote:Show nested quote +On August 20 2015 14:27 ChristianS wrote:On August 20 2015 14:12 unifo wrote: How much of an advantage playing random really is?
Where is the STATISTICAL evidence to support the advantage or disadvantage of being random? I would like to see current stats of random players win rates before really diving deeper into this issue. What kind of statistical evidence would you even want? If they're in ladder, their win rate is probably 50% because if it weren't, they would rise or fall on the ladder until it was. That is no different for random players. You could compare their mmr with random to their mmr with their best race, but that wouldn't reflect random advantage; that would reflect how good their off race matchups are. Now if you took two players with the same mmr who perform comparably in, say, PvP, then had one queue with their main race while the other queued as random, and then only looked at the PvPs, you could draw conclusions about the strength of the random advantage in PvP. I don't think anybody has done that kind of analysis, nor does anyone have the resources to do so. Yeuch. Random is clearly underrepresented in GM league. Random is extremely underrepresented in tournaments. The underlying concept of ladder matchmaking would actually lead to random players being overrepresented in higher leagues and underrepresented in lower leagues - IF there was indeed a random advantage. We actually witness the opposite. If we wanted 'balance', random players should be given extra minerals at game start until their - let's say Code S - representation mirrors the amount of random players in the game overall. What exactly are you arguing? That the racial distribution at the top shouldn't be 33/33/33 T/Z/P, it should be 25/25/25/25 T/Z/P/R? You know there's way fewer random players than any of the other races, right?
For fuck's sake, I wasn't arguing that overall random players are imbalanced in a WoL broodlord/infestor kind of way, hence no claims about the top level being dominated by random. I didn't even say that, in general, it's easier to be a random player than to pick a race. If a player chooses random and winds up in a TvT, they are better off than if they chose Terran and wound up in that same TvT. That's the imbalance I'm talking about.
Highly ironic thread looking at your sig. If you're going to imply that people you disagree with are stupid, go to battle.net forums or something. If you're on Team Liquid act like a fucking human being. And if you're gonna try to act superior, at least use the word "ironic" correctly.
|
On August 20 2015 14:46 Brutaxilos wrote:Show nested quote +On August 20 2015 14:27 JimmyJRaynor wrote: i play as Random because its the most fun. to add to my advantage sometimes i'll lie to my opponent and act "manner" by immediately telling him my race. but, i'm lying to my opponent. the additional advantage is higher when my opponent believes i'm being honest. As a proud Random player, I think you should be ashamed of yourself. This kind of behavior gives all Random players a bad name.
if i can get my opponent thinking about the "morality" of my actions... i'm already diverting his mind from the task at hand and winning that aspect of the game.
deception is part of many games. someone once said "all war is deception". am i driving while drunk or making questionable tax deductions or even violating the games terms of service? nah, i'm playing a game for fun.
deception is part of the game.
|
On August 20 2015 05:17 Clonester wrote: The random player has to be able to play all races. The race-picker not. Thus the random player has to learn 9 matchups against 3. It is literaly the reason to pick random in tournaments to have this starting advantage. Most random players want to learn all races best possible and thus announce what race they have, but dedicated random players who will attend in the automated tournaments will not. They not only pick random to play 3 races but also to win. If you want to be shure, use one of the starting workers to scout.
bullshit. The random player "has" to learn 9 different all-ins and execute them. he doesnt need to do jack besides that.
in LotV a workerscout sometimes aint enough.
a Random player will never play the real matchups either since hes forcing his opponent into suboptimal safer builds.
On August 20 2015 16:13 JimmyJRaynor wrote:Show nested quote +On August 20 2015 14:46 Brutaxilos wrote:On August 20 2015 14:27 JimmyJRaynor wrote: i play as Random because its the most fun. to add to my advantage sometimes i'll lie to my opponent and act "manner" by immediately telling him my race. but, i'm lying to my opponent. the additional advantage is higher when my opponent believes i'm being honest. As a proud Random player, I think you should be ashamed of yourself. This kind of behavior gives all Random players a bad name. if i can get my opponent thinking about the "morality" of my actions... i'm already diverting his mind from the task at hand and winning that aspect of the game. deception is part of many games. someone once said "all war is deception". am i driving while drunk or making questionable tax deductions or even violating the games terms of service? nah, i'm playing a game for fun. deception is part of the game.
sorry but youre just being a shitlord. Is winning really that important to you? as you said, youre just playing a game for fun.
|
Czech Republic12116 Posts
Just leave a game against Random every time as I do. Last time I was matched 5 times in a row against the same random player, I send some bullshit message like "I don't play random cheesing <insert a random profanity>" and I just left. I don't care that much and for him it had to be fun. He received a boost to his MMR and the next real game had to be really funny.
Who cares, you don't play for money anyway and those who do don't play random that often.
|
Iv'e gone through periods of playing random. At the start of every game I would tell the other player my race. Otherwise your practice won't be accurate because you will force your opponent to do a safe middle of the road build compared to a build optimized vs the race you are playing. So your practice will be shit.
|
You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim.
|
On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim.
yeah totally. Great evidence mate.
|
On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim.
Pretty much this. Playing random isn't easier than playing a race.
On August 20 2015 16:15 NEEDZMOAR wrote:Show nested quote +On August 20 2015 05:17 Clonester wrote: The random player has to be able to play all races. The race-picker not. Thus the random player has to learn 9 matchups against 3. It is literaly the reason to pick random in tournaments to have this starting advantage. Most random players want to learn all races best possible and thus announce what race they have, but dedicated random players who will attend in the automated tournaments will not. They not only pick random to play 3 races but also to win. If you want to be shure, use one of the starting workers to scout. bullshit. The random player "has" to learn 9 different all-ins and execute them. he doesnt need to do jack besides that. in LotV a workerscout sometimes aint enough. a Random player will never play the real matchups either since hes forcing his opponent into suboptimal safer builds.
Just like a random player can learn 9 all-ins, a person who plays a single race can learn 3 all-ins. Additionally everyone will assume the random player is going to all-in. Still harder for the random player.
|
On August 20 2015 15:54 ChristianS wrote:Show nested quote +On August 20 2015 15:19 Aiobhill wrote:On August 20 2015 14:27 ChristianS wrote:On August 20 2015 14:12 unifo wrote: How much of an advantage playing random really is?
Where is the STATISTICAL evidence to support the advantage or disadvantage of being random? I would like to see current stats of random players win rates before really diving deeper into this issue. What kind of statistical evidence would you even want? If they're in ladder, their win rate is probably 50% because if it weren't, they would rise or fall on the ladder until it was. That is no different for random players. You could compare their mmr with random to their mmr with their best race, but that wouldn't reflect random advantage; that would reflect how good their off race matchups are. Now if you took two players with the same mmr who perform comparably in, say, PvP, then had one queue with their main race while the other queued as random, and then only looked at the PvPs, you could draw conclusions about the strength of the random advantage in PvP. I don't think anybody has done that kind of analysis, nor does anyone have the resources to do so. Yeuch. Random is clearly underrepresented in GM league. Random is extremely underrepresented in tournaments. The underlying concept of ladder matchmaking would actually lead to random players being overrepresented in higher leagues and underrepresented in lower leagues - IF there was indeed a random advantage. We actually witness the opposite. If we wanted 'balance', random players should be given extra minerals at game start until their - let's say Code S - representation mirrors the amount of random players in the game overall. What exactly are you arguing? That the racial distribution at the top shouldn't be 33/33/33 T/Z/P, it should be 25/25/25/25 T/Z/P/R? You know there's way fewer random players than any of the other races, right? For fuck's sake, I wasn't arguing that overall random players are imbalanced in a WoL broodlord/infestor kind of way, hence no claims about the top level being dominated by random. I didn't even say that, in general, it's easier to be a random player than to pick a race. If a player chooses random and winds up in a TvT, they are better off than if they chose Terran and wound up in that same TvT. That's the imbalance I'm talking about. If you're going to imply that people you disagree with are stupid, go to battle.net forums or something. If you're on Team Liquid act like a fucking human being. And if you're gonna try to act superior, at least use the word "ironic" correctly.
I made it effing exceedingly clear that I do not expect a 25/25/25/25 distribution. I even wrote "...representation mirrors the amount of random players in the game overall." So you can shove your strawman where the sun doesn't shine.
And the TvT example is completely misleading as well. Obviously a random getting Terran in a mirror matchup with equal MMR is "better off" only under the assumption he trains Terran a third of his available time. If he spent all his training with one race, it's safe to assume his skill level should be higher. If he actually spent all his training with his strongest race, it *should* be significantly higher.
|
On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim. I claim that if a player were to queue as random and wind up in a TvT, he would have a better chance of winning than if he had queued as Terran and wound up in the same TvT. The situations are identical except that in the former case, his opponent has less information. That's the random advantage I'm talking about.
Do you honestly believe that advantage doesn't exist?
|
On August 20 2015 16:41 ChristianS wrote:Show nested quote +On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim. I claim that if a player were to queue as random and wind up in a TvT, he would have a better chance of winning than if he had queued as Terran and wound up in the same TvT. The situations are identical except that in the former case, his opponent has less information. That's the random advantage I'm talking about. Do you honestly believe that advantage doesn't exist?
This is definitely an advantage that exists, the question is does it really make up for all of the downsides of playing random? Evidence seems to say not, otherwise people would be playing random at a competitive level.
|
On August 20 2015 16:40 Aiobhill wrote:Show nested quote +On August 20 2015 15:54 ChristianS wrote:On August 20 2015 15:19 Aiobhill wrote:On August 20 2015 14:27 ChristianS wrote:On August 20 2015 14:12 unifo wrote: How much of an advantage playing random really is?
Where is the STATISTICAL evidence to support the advantage or disadvantage of being random? I would like to see current stats of random players win rates before really diving deeper into this issue. What kind of statistical evidence would you even want? If they're in ladder, their win rate is probably 50% because if it weren't, they would rise or fall on the ladder until it was. That is no different for random players. You could compare their mmr with random to their mmr with their best race, but that wouldn't reflect random advantage; that would reflect how good their off race matchups are. Now if you took two players with the same mmr who perform comparably in, say, PvP, then had one queue with their main race while the other queued as random, and then only looked at the PvPs, you could draw conclusions about the strength of the random advantage in PvP. I don't think anybody has done that kind of analysis, nor does anyone have the resources to do so. Yeuch. Random is clearly underrepresented in GM league. Random is extremely underrepresented in tournaments. The underlying concept of ladder matchmaking would actually lead to random players being overrepresented in higher leagues and underrepresented in lower leagues - IF there was indeed a random advantage. We actually witness the opposite. If we wanted 'balance', random players should be given extra minerals at game start until their - let's say Code S - representation mirrors the amount of random players in the game overall. What exactly are you arguing? That the racial distribution at the top shouldn't be 33/33/33 T/Z/P, it should be 25/25/25/25 T/Z/P/R? You know there's way fewer random players than any of the other races, right? For fuck's sake, I wasn't arguing that overall random players are imbalanced in a WoL broodlord/infestor kind of way, hence no claims about the top level being dominated by random. I didn't even say that, in general, it's easier to be a random player than to pick a race. If a player chooses random and winds up in a TvT, they are better off than if they chose Terran and wound up in that same TvT. That's the imbalance I'm talking about. Highly ironic thread looking at your sig. If you're going to imply that people you disagree with are stupid, go to battle.net forums or something. If you're on Team Liquid act like a fucking human being. And if you're gonna try to act superior, at least use the word "ironic" correctly. I made it effing exceedingly clear that I do not expect a 25/25/25/25 distribution. I even wrote "...representation mirrors the amount of random players in the game overall." So you can shove your strawman where the sun doesn't shine. You're misusing "strawman" too. A straw man is when I tell you what your position is, and misrepresent it. Since I was literally asking you what your position is, the term doesn't apply. As for where I can shove the not-strawman, I'll reiterate: be civil or go away. You're not contributing anything to the discussion by talking like that, and you're being a dick about it.
And the TvT example is completely misleading as well. Obviously a random getting Terran in a mirror matchup with equal MMR is "better off" only under the assumption he trains Terran a third of his available time. If he spent all his training with one race, it's safe to assume his skill level should be higher. If he actually spent all his training with his strongest race, it *should* be significantly higher. The whole point of this thread is to discuss whether or not the random player should get an in-game advantage in a TvT just because he's skilled at other matchups. Do you see why you're not contributing anything, then, when you say "Well yeah he's given an in-game advantage, but that's because he's skilled at other matchups too"?
|
On August 20 2015 16:45 Pursuit_ wrote:Show nested quote +On August 20 2015 16:41 ChristianS wrote:On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim. I claim that if a player were to queue as random and wind up in a TvT, he would have a better chance of winning than if he had queued as Terran and wound up in the same TvT. The situations are identical except that in the former case, his opponent has less information. That's the random advantage I'm talking about. Do you honestly believe that advantage doesn't exist? This is definitely an advantage that exists, the question is does it really make up for all of the downsides of playing random? Evidence seems to say not, otherwise people would be playing random at a competitive level. Okay well once it exists, there's a few questions. The first is whether it's strong enough to be worth the cost in time spent learning all the races. The answer seems to be no, particularly on the professional level where learning to play even one race sufficiently well takes almost more time than there is in the day. Maybe that will change when macro mechanics are removed, since that's a pretty large part of the investment in learning a new race, but for now the answer seems to be no.
The question this thread is about, however, is whether that advantage should exist anyway. Even if it's not small enough to make it a smart strategy for most players, does it create unfair situations? Does it create fun gameplay? Is it in keeping with the principles that competitive Starcraft is built on? I argue that it is at its heart an unfair advantage, since one player starts the game with more information than the other. The gameplay it encourages is typically coin-flippy for the non-random player and sloppy play for the random player. And it seems like an important principle of Starcraft that both players should start on equal footing, which this system violates.
|
On August 20 2015 17:00 ChristianS wrote:Show nested quote +On August 20 2015 16:45 Pursuit_ wrote:On August 20 2015 16:41 ChristianS wrote:On August 20 2015 16:34 MoreFaSho wrote: You claim that a player playing random has some kind of advantage yet 0 professional players who spend countless hours grinding out every improvement they can have decided to play random even though quite a few are more than proficient with the other races. I think this is relatively strong evidence against your claim. I claim that if a player were to queue as random and wind up in a TvT, he would have a better chance of winning than if he had queued as Terran and wound up in the same TvT. The situations are identical except that in the former case, his opponent has less information. That's the random advantage I'm talking about. Do you honestly believe that advantage doesn't exist? This is definitely an advantage that exists, the question is does it really make up for all of the downsides of playing random? Evidence seems to say not, otherwise people would be playing random at a competitive level. Okay well once it exists, there's a few questions. The first is whether it's strong enough to be worth the cost in time spent learning all the races. The answer seems to be no, particularly on the professional level where learning to play even one race sufficiently well takes almost more time than there is in the day. Maybe that will change when macro mechanics are removed, since that's a pretty large part of the investment in learning a new race, but for now the answer seems to be no. The question this thread is about, however, is whether that advantage should exist anyway. Even if it's not small enough to make it a smart strategy for most players, does it create unfair situations? Does it create fun gameplay? Is it in keeping with the principles that competitive Starcraft is built on? I argue that it is at its heart an unfair advantage, since one player starts the game with more information than the other. The gameplay it encourages is typically coin-flippy for the non-random player and sloppy play for the random player. And it seems like an important principle of Starcraft that both players should start on equal footing, which this system violates.
Unfair advantage? No. Unfun? I think it's safe to say yes given how many people complain about it.
There's no such thing as starting on equal footing. Maps will favor one race or position over another, players' experience and personal skill will be different, races will have different styles and strrengths based on the meta, some people play unorthodox strategies ect. Having all these different inequalities while still having a relatively balanced game it what makes it so fun. Giving random players a small advantage in information given the setbacks of playing random seems like an example of asymmetrical balance to me.
Personally I think playing vs random is usually fun, because it forces me to play a more organic game rather than following my usual flowchart, and random players often do wonky / quirky things that force me to think and adapt on the fly.
|
I think we all agree on the "balance" part of playing random:
- A terran player that picks random and gets terran is in a better position than picking terran from the start. This is what OP refers to as "the random advantage".
- Playing random isn't a good choice if you want to get high on the ladder or win tournaments. While there is the advantage above, the disadvantage of 9 MUs outweighs that advantage.
I think the question then is only if it is fun to play vs random. Some seem to find it extremely frustrating, others (like me) don't seem to mind as much. But that's opinion, which it isn't much use to discuss I feel.
|
|
|
|