|
On May 03 2011 20:59 Ctuchik wrote:Show nested quote +On May 03 2011 19:41 Elean wrote:On May 03 2011 18:27 Primadog wrote:
Any chance of outputting graphs with confidence intervals? I am concerned that some of the "trends" we see in the graph is simply random fluctuations due to some months having smaller sets of data. Assuming all the match are independant event, the standard deviation for a given matchup is: sqrt(p(1-p)/N) where p≈0.5 is the win rate of a race, and N the number of played games. Let's take the PvZ match up as example. There are 2244 games over 6 months, 374 games/months gives a standard deviation of 2.5% Considering an error of twice the standard deviation, your confidence interval is +/- 5%. Conclusion: the fluctuations we observe for the PvZ match up can very well be due to the sample size. For ZvT however the sample size is large enough to say the matchup was unbalanced. I would love to do his, trying to figure out how now. =P
Ideally you can make a graph something like the second graph on a stock screener:
![[image loading]](http://imgur.com/qLi1l.gif)
95% confidence-interval (2 standard deviation) is the standard that most uses. So the range you will use is the mean+/-2*standard deviation (μ +/- 2σ), where σ=(P * (1-P)/n)^.5. Note that n will be the number of data you have per month, not the total datapoints overall.
|
Lol there are so many details left out, this data shows very little except for players of which race are performing well in tournaments at specific points in time. I mean if you wanted this to be meaningful, much needs to be included. You could have a timeline and add lines showing when there were patches that made significant changes to units of different armies. Youd also have to show how many players of how many races played how many games over what period of time for both the international and korea ones. And even then, whilst the data would be a fair bit more interesting, it still would probably only indicate who did well and when  , which again could be due to other things such as player's health or what players were able to attend what events and what they couldnt, for whatever reason or reasons. Dont look for causation here
|
On May 03 2011 21:26 FireSA wrote:Lol there are so many details left out, this data shows very little except for players of which race are performing well in tournaments at specific points in time. I mean if you wanted this to be meaningful, much needs to be included. You could have a timeline and add lines showing when there were patches that made significant changes to units of different armies. Youd also have to show how many players of how many races played how many games over what period of time for both the international and korea ones. And even then, whilst the data would be a fair bit more interesting, it still would probably only indicate who did well and when , which again could be due to other things such as player's health or what players were able to attend what events and what they couldnt, for whatever reason or reasons. Dont look for causation here
Did you even look at the images?
It has everyone of the things you are talking about >_>
The only thing missing are patch dates, but you can easily find them out by yourself.
|
On May 03 2011 21:33 Binabik wrote:Show nested quote +On May 03 2011 21:26 FireSA wrote:Lol there are so many details left out, this data shows very little except for players of which race are performing well in tournaments at specific points in time. I mean if you wanted this to be meaningful, much needs to be included. You could have a timeline and add lines showing when there were patches that made significant changes to units of different armies. Youd also have to show how many players of how many races played how many games over what period of time for both the international and korea ones. And even then, whilst the data would be a fair bit more interesting, it still would probably only indicate who did well and when , which again could be due to other things such as player's health or what players were able to attend what events and what they couldnt, for whatever reason or reasons. Dont look for causation here Did you even look at the images? It has everyone of the things you are talking about >_> The only thing missing are patch dates, but you can easily find them out by yourself.
You forgot to highlight the How many players part. So I think the sample size as presented is misleading in a big way, as some have already suggested in this thread.
|
Very interesting, I wonder what caused the Protoss crash in the last month or two so In Korea.
|
|
|
Great graphs, i would like to see what they would look like if the points were every two weeks instead of every month but that is just me being picky. I hope you continue to update the graphs as more information comes to hand.
|
Damn that Korean ZvP data doesn't seem accurate. If this is accurate then I'm amazed.
The international scene looks good.
|
protoss hasn't lost a single game in the past two nights of the GSL, definitive proof that they are OP imo
|
NASL stats after 3 weeks
Protoss Terran Zerg
Protoss [17-17] | 50%[21-17] | 55%[15-20] | 43%
Terran [17-21] | 45%[18-18] | 50%[25-27] | 48%
Zerg [20-15] | 57%[27-25] | 52%[13-13] | 50%
Not 70% winrate in NASL for ZvP :o
|
|
|
Strange a young game with a hugely shifting meta game and newly discovered timings and counters has a constantly shifting win percentage that has almost nothing to do with actual balance. There could be no changes for 3 months and i guarantee there would be another huge shift in the korean numbers.
|
On May 03 2011 22:10 MrCon wrote:
NASL stats after 3 weeks
Protoss Terran Zerg
Protoss [17-17] | 50%[21-17] | 55%[15-20] | 43%
Terran [17-21] | 45%[18-18] | 50%[25-27] | 48%
Zerg [20-15] | 57%[27-25] | 52%[13-13] | 50%
Not 70% winrate in NASL for ZvP :o
If I'm understanding your post correctly it's 57% ZvP. That may not be 70 % but it's still really high if we assume those numbers would be a fair representation of ZvP generally in high level play (It's not). So I'm not really sure what your point is.
|
On May 03 2011 17:26 WhiteraCares wrote:Show nested quote +On May 03 2011 17:08 Zealot Lord wrote:All these protoss OP complaints really only started when they saw MC dominate the scene ..And when a certain prominent pro player started shitting all over protoss when in reality it was over his lack of tournament results. This, in a way greenlighted a shitstorm to be unleashed by other narcisistic zerg players who just couldn't live up to their parents praises and promises of a bright future.
It's sad because it's true. Personally, I don't really care if some pro tries to hide his lack of success behind balance whine, the real problem, that bothers me though is, how it influences the general balance sentiments around here.
NEXGenius once started a poll about to which race he should switch, Tester claimed he would start playing terran, MVP stated Terran was the weakest race, hell even Boxer mentioned once that he was considering to switch.....many pro gamers like to QQ when they lose. Since losing sucks, blaming it on something else other than your own play is only natural (up to a certain extent...)
Nevertheless it's unbelievable how this has started to influence the general mindset of people around here.
I like these graphs, not because they show whether or not the game is balanced but how the general assumptions around here (about zerg being terrible) don't even show up in recent tournament results. Hopefully things like these will make people think again before posting some random QQ in a LR-thread.
|
That ZvT graph makes me sad 
Thanks for making the graph however!
|
So how many games were played in each month? This is important information to give context to the ups and downs. I have to agree with the posts that this is not the best type of graph to indicate balance or that the Korean graph should have been posted if you don't have enough data points.
|
On May 03 2011 21:18 Primadog wrote:Show nested quote +On May 03 2011 20:59 Ctuchik wrote:On May 03 2011 19:41 Elean wrote:On May 03 2011 18:27 Primadog wrote:
Any chance of outputting graphs with confidence intervals? I am concerned that some of the "trends" we see in the graph is simply random fluctuations due to some months having smaller sets of data. Assuming all the match are independant event, the standard deviation for a given matchup is: sqrt(p(1-p)/N) where p≈0.5 is the win rate of a race, and N the number of played games. Let's take the PvZ match up as example. There are 2244 games over 6 months, 374 games/months gives a standard deviation of 2.5% Considering an error of twice the standard deviation, your confidence interval is +/- 5%. Conclusion: the fluctuations we observe for the PvZ match up can very well be due to the sample size. For ZvT however the sample size is large enough to say the matchup was unbalanced. I would love to do his, trying to figure out how now. =P Ideally you can make a graph something like the second graph on a stock screener: 95% confidence-interval (2 standard deviation) is the standard that most uses. So the range you will use is the mean+/-2*standard deviation (μ +/- 2σ), where σ=(P * (1-P)/n)^.5. Note that n will be the number of data you have per month, not the total datapoints overall.
Well, reading up a bit on it it doesn't seem like this would be all that useful. Please correct me if I'm wrong.
*The data behind this is not a sample of a bigger set. It's all matches played during this time period, and I'm not measuring it against a larger population.
*The actual data behind it is binary, ie, a match is either a 1 or a 0, win or loss.
*The sample size for each month is over 1000 games. See ![[image loading]](http://imgur.com/k49GV.png) .
Please correct me if I'm wrong here!
|
On May 03 2011 23:46 Spacemanspiff wrote:
So how many games were played in each month? This is important information to give context to the ups and downs. I have to agree with the posts that this is not the best type of graph to indicate balance or that the Korean graph should have been posted if you don't have enough data points.
Just posted this for reference: + Show Spoiler +
Also, note that I am not drawing ANY conclusions from the Korean graph in the OP, I am merely including it since people requested it.
|
On May 03 2011 23:27 nihlon wrote:Show nested quote +On May 03 2011 22:10 MrCon wrote:
NASL stats after 3 weeks
Protoss Terran Zerg
Protoss [17-17] | 50%[21-17] | 55%[15-20] | 43%
Terran [17-21] | 45%[18-18] | 50%[25-27] | 48%
Zerg [20-15] | 57%[27-25] | 52%[13-13] | 50%
Not 70% winrate in NASL for ZvP :o If I'm understanding your post correctly it's 57% ZvP. That may not be 70 % but it's still really high if we assume those numbers would be a fair representation of ZvP generally in high level play (It's not). So I'm not really sure what your point is.
The funny part is that if the stats were swaying so heavily in the Protoss' favour, the Zerg whine would be unbearable. Thank god for this thread.
These graphs certainly arn't a good way to show any kind of imbalance. But they sure do a good job of showing that Protoss is not infact unbeatable.
|
Well, for those raising eyebrows at the Korean PvZ graph, do note that it only includes 269 games. And it gets crazy just at the ending month of April, one of the six months, or roughly 45 games or so. Really small sample size for that one especially.
|
|
|
|
|
|
EPT
