|
I think these stats will be closer on may even without a balance patch, it takes time for zerg to figure out exactly how many drones they can make
|
On May 02 2013 06:21 Sissors wrote:Show nested quote +On May 02 2013 06:09 Eventine wrote:On May 02 2013 06:07 Sissors wrote:@Spiral Actually if this is based on the same spreadsheet that also circulated in other topics that is the case. First of all because we aren't talking here about thousands of games. Second if a single tournament has a very 'unbalanced' win rate that can have a significant influence. Case in point WCS europe qualifiers. Removing that from the list didn't make everything magically at 50%, but it did have a significant influence on the stats. Since there more zergs were invited than terran for example, you can also debate if it should be included. @Bronze, Lovebuzz has it almost correct: Actually, what they literally mean is simple: zergs are the least winning race right now in the games included in these statistics. Period Can you point me to the spreadsheet? Or better yet, is there spreadsheets for each months results by tournament? Thanks. Took some searching (yeah all for you  ), but found it again: https://docs.google.com/spreadsheet/ccc?key=0At0PE4rdhsI9dDE0cEprWkwwMGxQdTczTTlLNW1qX1E#gid=0Main thing to notice imo: If you delete the WCS qualifiers (so EU + NA), then suddenly you get (overall): 49.7% win rate for terran, 48.8% for zerg and 51.5% for toss. So those qualifiers have a huge impact. Not strange since both have TvZ rates at roughly 70%! So yeah I stick to my conclusions that I wouldnt conclude too much from these stats. Edit: It indeeds seems that those stats are from that spreadsheet. So would be nice if the link to the sheet was included in the OP, and in general it is alot more informative than only the pictures
Thanks. Damn, I was hoping for more detail. I looked into the data itself, if you want to do a more rigorous statistical analysis, someone has to individually code in the games. The counts aggregated too much of the information, whereas in a proposed analysis, I would want it to be map by map... I don't have the time to code that in. Oh well.
|
On May 02 2013 03:41 tuho12345 wrote:
Zergs won the first 2 major tourneys, they can't complain much tbh
Zerg also won the first 2 GSL's. Or are you arguing back then it was all fine as well?
Not saying they are UP but that logic is broken.
|
On May 02 2013 06:01 BronzeKnee wrote:Show nested quote +On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda...
I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues.
|
On May 02 2013 06:21 Sissors wrote:Show nested quote +On May 02 2013 06:09 Eventine wrote:On May 02 2013 06:07 Sissors wrote:@Spiral Actually if this is based on the same spreadsheet that also circulated in other topics that is the case. First of all because we aren't talking here about thousands of games. Second if a single tournament has a very 'unbalanced' win rate that can have a significant influence. Case in point WCS europe qualifiers. Removing that from the list didn't make everything magically at 50%, but it did have a significant influence on the stats. Since there more zergs were invited than terran for example, you can also debate if it should be included. @Bronze, Lovebuzz has it almost correct: Actually, what they literally mean is simple: zergs are the least winning race right now in the games included in these statistics. Period Can you point me to the spreadsheet? Or better yet, is there spreadsheets for each months results by tournament? Thanks. Took some searching (yeah all for you ), but found it again: https://docs.google.com/spreadsheet/ccc?key=0At0PE4rdhsI9dDE0cEprWkwwMGxQdTczTTlLNW1qX1E#gid=0Main thing to notice imo: If you delete the WCS qualifiers (so EU + NA), then suddenly you get (overall): 49.7% win rate for terran, 48.8% for zerg and 51.5% for toss. So those qualifiers have a huge impact. Not strange since both have TvZ rates at roughly 70%! So yeah I stick to my conclusions that I wouldnt conclude too much from these stats.
Nice thank you. He never should have made graphs from it, or at least he should have included all the information on the spreadsheet on the graphs.
On May 02 2013 06:29 achan1058 wrote:Show nested quote +On May 02 2013 06:01 BronzeKnee wrote:On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda... I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues.
You are completely misusing the law of truly large numbers and taking it out of context. It has nothing to do with a small sample size of games.
The law seeks to debunk one element of supposed supernatural phenomenology. For instance, some theologians will argue that the chances that the Earth was created by chance and not by intelligent design are absurdly small. Thus in there eyes, Earth could have only been created by God.
However, when we consider the size of the universe, and the very large number of planets, you'll realize that the Earth being created by chance wasn't exactly that unlikely. In fact, for it not occur would be unlikely! See the math below.
http://en.wikipedia.org/wiki/Law_of_truly_large_numbers
+ Show Spoiler +For a simplified example of the law, assume that a given event happens with a probability of 0.1% in one trial. Then the probability that this unlikely event does not happen in a single trial is 99.9% = 0.999.
In a sample of 1000 independent trials, the probability that the event does not happen in any of them is , or 36.8%. The probability that the event happens at least once in 1000 trials is then 1 − 0.368 = 0.632 or 63.2%. The probability that it happens at least once in 10,000 trials is .
This means that this "unlikely event" has a probability of 63.2% of happening if 1000 chances are given, or over 99.9% for 10,000 chances. In other words, a highly unlikely event, given enough tries, is even more unlikely to not occur.
[edit]
|
On May 02 2013 05:43 MarlieChurphy wrote:
These stats are only representative to timing imbalances within the units. Each race has really imbalanced shit, Zerg's stuff all comes really later. Protoss and Terran stuff come really earlier.
Widow mine range needs decrease.
Void ray prismatic alignment needs it's buff/debuff timers drastically modified. (possiblly 4-5 seconds buff, 10-30 seconds recharge)
Hellbats could use a nerf or two. HP, Armor, but more importantly the way the damage and unit type modifier is layed out. (they need to do less to all and big boost to small/bio)
Ravens - not sure on these yet, but I rarely lose a game where I get a nice ball of them added up. Especially in TvT
Swarm Hosts with proper dichotomy of AA and creep seem really hard for protoss to deal with. (possible nerf of locust range, damage, or hp with increase in spawn time.)
Perhaps you would also like some nerfs to the marine, the medivac and the siegetank? What are you smoking man? Did you know that TvZ was pretty balanced and FUN to watch in WOL before the queennerf (remember DRG vs MKP @ mlg)? Then blizzard buffed the queen a tiny bit and BOOM, we got a long time of zerg dominance.
TvZ in HOTS is fun to watch. Topzergs have no problem beating terrans. Both races have to put a serious amount of practice into the game, and both sides have to micro & macro as hard. The days where zergs could 1a into broodlord infestor are over, and lazy zergs will indeed lose a lot more now, I agree.
|
On May 02 2013 06:29 achan1058 wrote:Show nested quote +On May 02 2013 06:01 BronzeKnee wrote:On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda... I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues.
edit: ignore, misunderstood post.
|
they dont need to nerf or buff anything yet! just wait another 4-6 damn months or so. too many frequent balance patch ruined WOL.
|
On May 02 2013 06:32 BronzeKnee wrote:Show nested quote +On May 02 2013 06:21 Sissors wrote:On May 02 2013 06:09 Eventine wrote:On May 02 2013 06:07 Sissors wrote:@Spiral Actually if this is based on the same spreadsheet that also circulated in other topics that is the case. First of all because we aren't talking here about thousands of games. Second if a single tournament has a very 'unbalanced' win rate that can have a significant influence. Case in point WCS europe qualifiers. Removing that from the list didn't make everything magically at 50%, but it did have a significant influence on the stats. Since there more zergs were invited than terran for example, you can also debate if it should be included. @Bronze, Lovebuzz has it almost correct: Actually, what they literally mean is simple: zergs are the least winning race right now in the games included in these statistics. Period Can you point me to the spreadsheet? Or better yet, is there spreadsheets for each months results by tournament? Thanks. Took some searching (yeah all for you ), but found it again: https://docs.google.com/spreadsheet/ccc?key=0At0PE4rdhsI9dDE0cEprWkwwMGxQdTczTTlLNW1qX1E#gid=0Main thing to notice imo: If you delete the WCS qualifiers (so EU + NA), then suddenly you get (overall): 49.7% win rate for terran, 48.8% for zerg and 51.5% for toss. So those qualifiers have a huge impact. Not strange since both have TvZ rates at roughly 70%! So yeah I stick to my conclusions that I wouldnt conclude too much from these stats. Nice thank you. He never should have made graphs from it, or at least he should have included all the information on the spreadsheet on the graphs. Show nested quote +On May 02 2013 06:29 achan1058 wrote:On May 02 2013 06:01 BronzeKnee wrote:On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda... I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues. You are completely misusing the law of truly large numbers and taking it out of context. It has nothing to do with a small sample size of games. The law seeks to debunk one element of supposed supernatural phenomenology. For instance, some theologians will argue that the chances that the Earth was created by chance and not by intelligent design are absurdly small. Thus in there eyes, Earth could have only been created by God. However, when we consider the size of the universe, and the very large number of planets, you'll realize that the Earth being created by chance wasn't exactly that unlikely. In fact, for it not occur would be unlikely! See the math below. http://en.wikipedia.org/wiki/Law_of_truly_large_numbers+ Show Spoiler +For a simplified example of the law, assume that a given event happens with a probability of 0.1% in one trial. Then the probability that this unlikely event does not happen in a single trial is 99.9% = 0.999.
In a sample of 1000 independent trials, the probability that the event does not happen in any of them is , or 36.8%. The probability that the event happens at least once in 1000 trials is then 1 − 0.368 = 0.632 or 63.2%. The probability that it happens at least once in 10,000 trials is .
This means that this "unlikely event" has a probability of 63.2% of happening if 1000 chances are given, or over 99.9% for 10,000 chances. In other words, a highly unlikely event, given enough tries, is even more unlikely to not occur.
[edit]
I did not say his sample size is small. I say his choice of the selection of tournaments to include/not include can be questioned. Yes, I would admit that I am pushing it, since he didn't really try hard to pick and choose which tournaments to include.
|
On May 02 2013 06:01 BronzeKnee wrote:Show nested quote +On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda...
You missed the entire point of my post. The statistics is not backed up by a method regarding how it should be interpreted as of context and use. I exemplified this by statements which questions it's validity that are easily answered with a proper methodology in place. as it is now we can't tell for certain if my concerns are correct or if your "systemic disagreement" is correct, if they are both wrong or both right, this is also why people tend to speculate and come to various conclusions, as you pointed out, because we don't know. We are just making shit up what it could mean. This is not actually based on anything that can be considered as accurate.
|
On May 02 2013 06:41 achan1058 wrote:Show nested quote +On May 02 2013 06:32 BronzeKnee wrote:On May 02 2013 06:21 Sissors wrote:On May 02 2013 06:09 Eventine wrote:On May 02 2013 06:07 Sissors wrote:@Spiral Actually if this is based on the same spreadsheet that also circulated in other topics that is the case. First of all because we aren't talking here about thousands of games. Second if a single tournament has a very 'unbalanced' win rate that can have a significant influence. Case in point WCS europe qualifiers. Removing that from the list didn't make everything magically at 50%, but it did have a significant influence on the stats. Since there more zergs were invited than terran for example, you can also debate if it should be included. @Bronze, Lovebuzz has it almost correct: Actually, what they literally mean is simple: zergs are the least winning race right now in the games included in these statistics. Period Can you point me to the spreadsheet? Or better yet, is there spreadsheets for each months results by tournament? Thanks. Took some searching (yeah all for you ), but found it again: https://docs.google.com/spreadsheet/ccc?key=0At0PE4rdhsI9dDE0cEprWkwwMGxQdTczTTlLNW1qX1E#gid=0Main thing to notice imo: If you delete the WCS qualifiers (so EU + NA), then suddenly you get (overall): 49.7% win rate for terran, 48.8% for zerg and 51.5% for toss. So those qualifiers have a huge impact. Not strange since both have TvZ rates at roughly 70%! So yeah I stick to my conclusions that I wouldnt conclude too much from these stats. Nice thank you. He never should have made graphs from it, or at least he should have included all the information on the spreadsheet on the graphs. On May 02 2013 06:29 achan1058 wrote:On May 02 2013 06:01 BronzeKnee wrote:On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda... I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues. You are completely misusing the law of truly large numbers and taking it out of context. It has nothing to do with a small sample size of games. The law seeks to debunk one element of supposed supernatural phenomenology. For instance, some theologians will argue that the chances that the Earth was created by chance and not by intelligent design are absurdly small. Thus in there eyes, Earth could have only been created by God. However, when we consider the size of the universe, and the very large number of planets, you'll realize that the Earth being created by chance wasn't exactly that unlikely. In fact, for it not occur would be unlikely! See the math below. http://en.wikipedia.org/wiki/Law_of_truly_large_numbers+ Show Spoiler +For a simplified example of the law, assume that a given event happens with a probability of 0.1% in one trial. Then the probability that this unlikely event does not happen in a single trial is 99.9% = 0.999.
In a sample of 1000 independent trials, the probability that the event does not happen in any of them is , or 36.8%. The probability that the event happens at least once in 1000 trials is then 1 − 0.368 = 0.632 or 63.2%. The probability that it happens at least once in 10,000 trials is .
This means that this "unlikely event" has a probability of 63.2% of happening if 1000 chances are given, or over 99.9% for 10,000 chances. In other words, a highly unlikely event, given enough tries, is even more unlikely to not occur.
[edit] I did not say his sample size is small. I say his choice of the selection of tournaments to include/not include can be questioned.
There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view.
This law does not exist. If you mean the law of truly large numbers (which I assumed), then you are completely misusing it. The data was not derived from a truly large number, so arguing that it is invalid because of that is senseless. That was my point.
Anyway, what is the problem with his selection of tournaments? What do you think he should have added or excluded? He included all major professional tournaments recently from what I can see, but I could very well be wrong.
|
tvz is a lot closer than I expected
|
well it almost looks like the WoL start, so if we don't buff Zerg now the game will be balanced before LotV release. If we buff Zerg now it will be balanced now and Zerg will be OP again in 2 years.
I just hope Tanks get stomped even in TvT at some point, so we get more awesome Tanks !
|
Tbh he is right that even if the selection is perfectly valid, the problem is that there are really many perfectly valid selections. As example excluding the WCS quals would be no problem, you can give a good reason why you do it, and it suddenly gives a totally different result. You can set the bar higher of which games you include, or lower. You can only look at Korean games, and there you also have different levels.
So in the end there are quite some valid tournament selections that can yield different results. Which is why you always should be careful with statistics: statistics are definately useful, but you really have to watch closely what is included and how significant the result is.
|
On May 02 2013 06:44 BronzeKnee wrote:Show nested quote +On May 02 2013 06:41 achan1058 wrote:On May 02 2013 06:32 BronzeKnee wrote:On May 02 2013 06:21 Sissors wrote:On May 02 2013 06:09 Eventine wrote:On May 02 2013 06:07 Sissors wrote:@Spiral Actually if this is based on the same spreadsheet that also circulated in other topics that is the case. First of all because we aren't talking here about thousands of games. Second if a single tournament has a very 'unbalanced' win rate that can have a significant influence. Case in point WCS europe qualifiers. Removing that from the list didn't make everything magically at 50%, but it did have a significant influence on the stats. Since there more zergs were invited than terran for example, you can also debate if it should be included. @Bronze, Lovebuzz has it almost correct: Actually, what they literally mean is simple: zergs are the least winning race right now in the games included in these statistics. Period Can you point me to the spreadsheet? Or better yet, is there spreadsheets for each months results by tournament? Thanks. Took some searching (yeah all for you ), but found it again: https://docs.google.com/spreadsheet/ccc?key=0At0PE4rdhsI9dDE0cEprWkwwMGxQdTczTTlLNW1qX1E#gid=0Main thing to notice imo: If you delete the WCS qualifiers (so EU + NA), then suddenly you get (overall): 49.7% win rate for terran, 48.8% for zerg and 51.5% for toss. So those qualifiers have a huge impact. Not strange since both have TvZ rates at roughly 70%! So yeah I stick to my conclusions that I wouldnt conclude too much from these stats. Nice thank you. He never should have made graphs from it, or at least he should have included all the information on the spreadsheet on the graphs. On May 02 2013 06:29 achan1058 wrote:On May 02 2013 06:01 BronzeKnee wrote:On May 02 2013 03:27 Integra wrote:On May 02 2013 03:22 Alryk wrote:On May 02 2013 03:05 Integra wrote:On May 02 2013 02:52 Alryk wrote:On May 02 2013 02:45 synd wrote:
I don't trust this statistics at all Any reasoning for that lol? 52/48 is almost definitely within margin of error, although we don't know exactly what the margin of error is. I don't know if including qualifiers was a good idea because of the potential for amateur vs pro matches, but there should be so few of them included that they get drowned out in the sample anyways. Kind of interesting statistics, and I'm not too too surprised by what I see. Even 55/45 is potentially close to margin of error, although in this case it's unlikely. I hope we see buffs to zerg and not nerfs to whatever is necessary - it's usually a much more entertaining way of doing the game. Because it's bullshit, nor sample size, how relevant the sample size or that the numbers actually are showing "balance" are determined or properly explained.How do we know that the sample size is represental for the game a a whole, how accurate does it reflect the MU's on other leagues such as gold, plat, diamond etc, How does it compare to different regions, like EU, NA and KR etc. How do we know that these numbers only relate to race balance issues, it could just as well showing bad designed maps, people with actual lower skill meeting other players with better skill or just the fact that certain players are better at playing on high latency and that just happens to be Terran players, or maybe Terran players ARE easier to play during latency compared to the other races. No methodlogy for data collecting has been determined and no attempt has been made to actually isolate the important variables for proving if one race is better than another, we don't even know what those variables are People will only start to speculate and make their own shit up as of why this has happened, and depeding of their personal bias they will either confirm or reject the statistics. That sc2 statistics guy did the exact same thing for about 2 years and nobody complained. And obviously we aren't looking at gold etc. statistics because it takes too much work. The majority of people however are interested in balancing the game at the pro level, not for bronze players. And while you don't want it to be impossible for a bronze zerg to win, it doesn't make sense to make changes FOR them. Badly designed maps are a fact of life - there's no way around them. The statistics aren't trying to show that. They're simply showing game balance as is. If you notice, the OP didn't actually bring his opinions into the actual statistics. For what they're intended to represent, they work well. Are they perfect? no not really. But pointing out the things you did makes no sense; the survey doesn't aim to address map balance. Also, the low vs high latency games are so few and far between that they're drowned out by normal games. Things like that. I don't really feel like explaining all of statistics though lol. You point out bias, but it isn't relevant bias. It's not something they can prevent. How would you gather statistics that determined how good a map was? That's almost entirely subjective. So by other words, it could mean ANYTHING, and thus is just useless statitsics with no real intent or clear focus or aim. @BronzeKnee; thank you for proving my point regarding speculation and making their own shit up regarding what it it could mean. Yes by systemically disagreeing with you and showing why you were wrong I proved my point... Every statistical discussion on this website turns out like this. People rant and rave and don't think anyone can be objective, that everyone has some hidden agenda... I do think that, actually. There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. The only way you can make a reasonably objective statistic is to choose your sample first, then find out about the result. ie. We discard this month's data, and look at next month using the same selection criteria, and see whether the trend continues. You are completely misusing the law of truly large numbers and taking it out of context. It has nothing to do with a small sample size of games. The law seeks to debunk one element of supposed supernatural phenomenology. For instance, some theologians will argue that the chances that the Earth was created by chance and not by intelligent design are absurdly small. Thus in there eyes, Earth could have only been created by God. However, when we consider the size of the universe, and the very large number of planets, you'll realize that the Earth being created by chance wasn't exactly that unlikely. In fact, for it not occur would be unlikely! See the math below. http://en.wikipedia.org/wiki/Law_of_truly_large_numbers+ Show Spoiler +For a simplified example of the law, assume that a given event happens with a probability of 0.1% in one trial. Then the probability that this unlikely event does not happen in a single trial is 99.9% = 0.999.
In a sample of 1000 independent trials, the probability that the event does not happen in any of them is , or 36.8%. The probability that the event happens at least once in 1000 trials is then 1 − 0.368 = 0.632 or 63.2%. The probability that it happens at least once in 10,000 trials is .
This means that this "unlikely event" has a probability of 63.2% of happening if 1000 chances are given, or over 99.9% for 10,000 chances. In other words, a highly unlikely event, given enough tries, is even more unlikely to not occur.
[edit] I did not say his sample size is small. I say his choice of the selection of tournaments to include/not include can be questioned. Show nested quote +There's this thing called the law of very large numbers. That is to say, if you make enough samples, at least one will skew towards your view. This law does not exist. If you mean the law of truly large numbers (which I assumed), then you are completely misusing it. The data was not derived from a truly large number, so arguing that it is invalid because of that is senseless. That was my point. Anyway, what is the problem with his selection of tournaments? What do you think he should have added or excluded? He included all major professional tournaments recently from what I can see, but I could very well be wrong.
As I understand it, it does not describe the data itself. Rather, it says that given enough samples or criteria selections, there is very likely one that appears abnormal. Assuming that the provider of the data is malicious, he could pick to see which tournaments/cut off to skew towards a certain win rate and include those. Yes, I will admit I am pushing it a bit since he didn't really pick at which kind of tournaments to include/exclude, and the number of natural cut-offs for tournament tiers is somewhat limited.
|
On May 02 2013 03:25 Danners933 wrote:
Honestly Sc2 feels the most balanced as it has even been in my opinion. The only thing I really think needs a looking at is Hellbat drops. As for the percentages it really does look about right. Zerg feels like they need help but they rarely use any of the new tools given to them. Yet they are still winning on the pro level. So give them time to play, I think within a few months Zerg Win Rate should jump.
Yeah I agree. The game is mostly balanced besides hellbats and widow mines drops (which are really silly IMO).
Widow Mines should have cap on the amount of units it can hit (you can set it in the "Search" field, and set "Maximum Count" to any #).
The game where 1 widow mine took out 14 probes (sOs) and he lost just because of 1 WM was crazy.
Widow Mine blast should cap at like maybe 10 units or so.
Even as a mostly Terran player (playing random the other time), I don't like Widow Mines or Hellbats that much.
In BW, nothing was as random or could change the entire course of the game as a Widow Mine (for example).
Reaver scarab blasts were predictable and slow (both opponents could micro against it easily). The scarabs would dud and do no damage sometimes but that was only if the scarab traveled some crazy pathing (like trying to travel around a structure or something, as scarabs had ground pathing and could be blocked by ground units).
Plus, it's rare that 1 or 2 scarabs changed the outcome of the entire game (mainly because units didn't really clump as much) like Widow Mines do.
I say a cap of like 10 units max (again, it can be set in the search field of the data editor under "Max Count" or "Max Target").
Anything that makes the game end (or decides the entire game) due to 1 or 2 things (especially if they are random as widow mines) should be nerfed somewhat.
Randomness isn't a problem itself. I mean WarCraft 3 and Dota have randomness, but nothing in WC3 or Dota have randomness enough that you lose the game outright due to 1-2 bad rolls (there have been plenty of times in Dota where the entire team gets taken out due to something "unlucky", like being caught out of position but they still were able to make a comeback... in SC2 this rarely happens as much due to how 1 or 2 unlucky things, like a Widow Mine taking out 14 or your workers early game, can put you far behind that you probably won't make a comeback).
(Also it's not just Widow Mines either. I think Protoss in general has that problem where the whole investment is on one big Protoss army, and if they lose or win 1 big battle with that Protoss army, the game is decided there. That's why I dislike watching anything involving Protoss. TvZ is more back and forth and more gradual.)
While the game with sOs where 1 Widow Mine destroyed 14 probes is rare, something like that shouldn't be in the game.
tl;rd - Widow Mines should have a "Maximum Count" (editable in the "Search" effect) of 10 or less (rather than unlimited like it is now).
|
Hmmm, the odd part seems to be the EU winrates in TvZ specifically. I wonder if it's some sort of rubberbanding effect from EU Zerg domination in WoL, kind of like a bubble bursting.
|
On May 02 2013 03:22 Eventine wrote:
always nice to see people with demands and complaints on the data and a lack of commitment to actually provide "better" data or analysis.
Agreed. These data sets are interesting as always, and something can be garnered from them regardless of a better set of data existing.
Mentioning GSL tournament results is a nice contrast because that is another very specific set of data.
|
I think the game is quite balanced other than ZvT for now.
The biggest problem is that there is almost only one style from the terran, bio mine whole game long.
Even if the game is completely balanced, I would love bio mine style to get a small nerf so mech would get used and viper/swarm hosts can make some appearance
|
On May 02 2013 06:33 Snowbear wrote:Show nested quote +On May 02 2013 05:43 MarlieChurphy wrote:
These stats are only representative to timing imbalances within the units. Each race has really imbalanced shit, Zerg's stuff all comes really later. Protoss and Terran stuff come really earlier.
Widow mine range needs decrease.
Void ray prismatic alignment needs it's buff/debuff timers drastically modified. (possiblly 4-5 seconds buff, 10-30 seconds recharge)
Hellbats could use a nerf or two. HP, Armor, but more importantly the way the damage and unit type modifier is layed out. (they need to do less to all and big boost to small/bio)
Ravens - not sure on these yet, but I rarely lose a game where I get a nice ball of them added up. Especially in TvT
Swarm Hosts with proper dichotomy of AA and creep seem really hard for protoss to deal with. (possible nerf of locust range, damage, or hp with increase in spawn time.)
Perhaps you would also like some nerfs to the marine, the medivac and the siegetank? What are you smoking man? Did you know that TvZ was pretty balanced and FUN to watch in WOL before the queennerf (remember DRG vs MKP @ mlg)? Then blizzard buffed the queen a tiny bit and BOOM, we got a long time of zerg dominance. TvZ in HOTS is fun to watch. Topzergs have no problem beating terrans. Both races have to put a serious amount of practice into the game, and both sides have to micro & macro as hard. The days where zergs could 1a into broodlord infestor are over, and lazy zergs will indeed lose a lot more now, I agree.
Actually TvZ is shit to watch at the moment, every game is the same. Zerg defends until Muta. Terran build cheap mineral only units and pushes up the map complete with a further 10 min death animation .Getting bored of it already tbh and i suspect in the next 3 months many others will too.
PvT is the only matchup worth watching these days, at least the first 10 mins of the game has some variety with different openings and units.
C
|
|
|
|
|
|
EPT
