|
I just spent 10 minutes looking at my match history. Here are my last 33 games.
In order from most recent to oldest...
1. T 2. T 3. T 4. T 5. P 6. P 7. T 8. P 9. T 10. P 11. T 12. P 13. T 14. T 15. T 16. P 17. T 18. T 19. T 20. T 21. P 22. P 23. T 24. T 25. T 26. P 27. T 28. P 29. P 30. T 31. T 32. T 33. P 34. Z
Twenty-one T, Twelve P. Don't get me wrong, ZvZ is my least favorite matchup, but I wouldn't mind playing it SOMETIME just to get practice! Just really annoyed lately at the sheer amount of Terrans in a row I'm facing. If you don't believe me and want to look up my statistics I'm FabIntegral.581 on ladder. I started playing T in 2v2 for this reason as they are broken there. =/.
PS. I feel 34 is a sufficiently large sample size. Maybe I'm an outlier, I'm just tired of it!
PSS. Not saying anything about balance, I'm playing fine atm vs P and T.
   
|
I'm considering starting to play Protoss cause i really hate playing TvT and pretty much everyone plays Terran 
Hope Blizzard buffs Zerg so we have less Terrans.
|
Haha, ye I know the feeling.
I can go for complete playing sessions without ever seeing a ZvZ at all.
I guess on average, let's say every 10 games it's: 5T 4P 1Z for me.
And yes ZvZ is also my weakest MU since I almost barely get to play it ^^
|
I play far more P than T at my level T_T
I hate PvP so much
SO FUCKING MUCH
|
I really hate TvP
Can't beat the FE build for the life of me
Single reason I'm considered to go to P
|
EDIT: I get owned (I know nil about statistics), read below.
^_^
|
Wow must be something wrong with your matchmaking system. How the hell do you rarely get TVT?
|
34 is not a sufficiently large sample size for what? he didn't make any conclusions or inferences from the data, just talked about how rarely he has seen zerg.
Also, in many statistical tests, 34 is actually plenty, considering the nature of the test being done. In other words, if your null hypothesis was that there are equal amounts of zerg, terran, and protoss players versus the alternative hypothesis of there being less zerg than t or p (which we already know to be true from large amounts of data) then I am quite confident you would reject the null in favor of the alternative based on the 34 games, although i didn't take the time to go through it myself.
I am just tired of people saying that there "isn't a sufficiently large sample size" and then feeling smart because they remembered one line from their statistics class, even though that has no relevance to the discussion. It really makes you look foolish.
|
Seems like all i got today was zvz and i lost them all because i have no idea how to zvz. Consider yourself lucky.
|
thedeadhaji
39489 Posts
The chances of getting no zergs 33 times in a row, assuming equal probability of facing each race (which is inaccurate, but let's assume this for simplicity's sake) is the probability of facing a T or P 33 times in a row, so that's (2/3)^33 right?
That's a pretty small chance ><
|
/agree with petered.
34 is a huge fucking sample size.
The odds of this happening is extremely low even given the current race distribution.
|
It is obviously a large enough sample.
Let the null hypothesis be that you meet Zerg more than 1/3 of the time Let the alternative be that you meet Zerg less than 1/3 of the time
It's pretty simple to see that:
+ Show Spoiler +> dbinom(1,34,0.33) [1] 2.043908e-05 > dbinom(0,34,0.33) [1] 1.220515e-06 > 2.043908e-05 + 1.220515e-06 [1] 2.165960e-05
the p-value is 2.165960e-05 so we would reject null... under almost any reasonable levels.
We can also do an estimate of Z vs other races ratio and compute the variance of the estimator. It's 3AM right now though so I will cut myself some slack. Maybe tomorrow :3
|
I'm in the same boat. ZvZ is my worst MU but some of that is because, outside of customs, it's easily the rarest matchup. Of course there are times when I get it 3-4 times in a row, but...
Tons of P and T (I don't really mind as I find those matchups more interesting and dynamic, however).
|
Haha this is kind of good for a learning player like me, as I can focus on learning only two matchups for now. Of course, it's bad for the long run though.
|
It's pretty clear the matchmaking isn't random, I would like to hear Blizzard's reasoning on why they decided this would be a good idea or whether they think it's working as intended or not. For the purpose of the thread I went ahead and did the same thing as the OP, and got similarly idiotic results.
1. T 2. T 3. Z 4. Z 5. T 6. P 7. P 8. P 9. P 10. P 11. P 12. P 13. P 14. P 15. T 16. T 17. P 18. P 19. T 20. P 21. Z 22. P 23. P 24. P 25. P 26. Z 27. T 28. P 29. P 30. P 31. P 32. P 33. Z 34. T
21 protoss, 8 terrans and 5 zergs. I don't hate PvP as a matchup but holy hell has Bnet2.0 spoiled it for me. Basically any other matchup and I probably wouldn't mind too much, but mirrors get boring much quicker than the other matchups... I'm really turned off the ladder as a whole because I know it'll queue me up yet another mirror.
|
i play about 7/10 Terrans, the rest 3 is sometimes all Z sometimes all P, rarely mixed
|
Yesterday i played like 5 ZvZ's in a row, i was sad =(
|
|
Out of my 200 games I've only played against about 5 random people total. That probably has to do with there being so few random players, especially at the higher ranks of diamond haha . My games that I have recorded were: + Show Spoiler +p z t t t t p p p z p t t t z t p p p z p t p p p p z p z t t t p p t p p p p t z t p t p z t z t t p Thats: 23 Protoss 18 Terran 7 Zerg
Some of those might have been random, I only see the race they end up getting though.
I think there's so many Protoss because I feel like I'm at the point where 4 gate suddenly stops being so successful and the Protoss who are doing it are hitting a peak.
Note: I play random
|
On September 13 2010 16:43 Polar_Nada wrote: lol just play shoey!
im goign to try to play some T offrace. I need a break from Z.
gimme some reps polar lol
I'm going to make it even MORE unlikely that he gets zvz matchups
|
On September 13 2010 16:43 Polar_Nada wrote: lol just play shoey!
We played a custom today.. but I was referencing 1v1 ladder only!
|
I don't know if the MM system is truly random or not but it makes you really question the system and the map pool when you get 4 straight PvP's on Blistering Sands.
|
In my first 23 games, I played around 16 terran, 5 protoss and 1 zerg.
|
Some peoples understanding of random seems to be quite lacking. If there was an even distribution of races (1/3) each then yes, over a large enough sample size(more then once person) it would balance out to getting about 1/3 of each race. The races are not evenly distributed, it is skewed to protoss, terran, then last zerg. Random does not mean you will get an even distribution of the races, it means you will get a random distribution. Flip a fair coin 100 times you will prob get around 50 heads. Get over a million people flipping coins 100 times, its almost certian that some of them will be 75+% heads.
tl;dr People think a random distribution is an even distribution, when in reality it doesn't have to be anything of the kind
|
Some statistics to help with the calculations (from sc2ranks.com):
Global race distribution:
Random: 8.05% Protoss: 36.03% Terran: 36.78% Zerg: 19.14%
Not many Zerg D:
|
Calgary25969 Posts
I play random, and it seems like I only get Z when it's ZvZ. It's so upsetting because I love playing ZvT and ZvP
|
Didnt blizz say less than 20% of total SC2 players play zerg?
seems reasonable to me
|
On September 13 2010 23:30 NukeTheBunnys wrote: Some peoples understanding of random seems to be quite lacking. If there was an even distribution of races (1/3) each then yes, over a large enough sample size(more then once person) it would balance out to getting about 1/3 of each race. The races are not evenly distributed, it is skewed to protoss, terran, then last zerg. Random does not mean you will get an even distribution of the races, it means you will get a random distribution. Flip a fair coin 100 times you will prob get around 50 heads. Get over a million people flipping coins 100 times, its almost certian that some of them will be 75+% heads.
tl;dr People think a random distribution is an even distribution, when in reality it doesn't have to be anything of the kind
I don't think anyone has the misconceptions that you have. I already stated I would qualify as an outlier. I am also aware Zerg is seen the least. However, I'm blogging about my most recent SC2 experience and the utter lack of Zergs I've seen to encounter as I've progressed up to the ladder. If you look in almost any division, there are less and less Zerg the higher up you go. I was just talking about how I'm frustrated I personally haven't encountered any Zerg and that it's very statistically unlikely something like this is going to happen.
Yeah, it's going to happen to some people, but that's hardly an excuse for everything. People reject null hypotheses with a margin of error of ~5% depending on what they're studying. My p value would be quite below .05.
|
Match-making isn't random. The system always pits me against three Diamond players when I'm doing well. I will always win against the Z and P, but the third Diamond match is always TvT, and I always fail, and stay stuck in Platinum It knows your game better than you know it.
|
Lol i changed to t from p to avoid all the pvp's =P matchmaking system is good though
|
On September 14 2010 00:58 Chill wrote:I play random, and it seems like I only get Z when it's ZvZ. It's so upsetting because I love playing ZvT and ZvP 
Don't get me wrong, I definitely prefer ZvT and ZvP both over ZvZ. It's just my ZvZ is so terrible atm, I feel like I personally am an outlier (although I think Z is def more rare overall) and whenever I do get matched up against a Z I'm gonna get rolled.
Btw that 34th match vs a Z was a loss =[
|
I have the same thing happen to me. Log on, play a ton of Terrans in a row, log off upset LOL.
|
well i play on eu, but i rarely get a ZvT.
My most played and my worst matchup is ZvZ.. followed by RvMirror and ZvP
|
On September 14 2010 00:58 Chill wrote:I play random, and it seems like I only get Z when it's ZvZ. It's so upsetting because I love playing ZvT and ZvP  Better than getting P in PvP, T in TvT and Z in ZvZ. and the occasional PvZ. I DIDNT PLAY THE OTHER MUS AS RANDOM YET!!!!!!!!!!!! damn my random luck -_-
|
Agh I understand the feeling here in gold xD For me it's usually PPPPPPPPPPPPPPPPPPP all of them 4 gates get really tiring really fast.
I am learning to defend against it better though. Spine crawlers are great :D
|
Meh, just got 10 zvz in the last 15 games ANd like 6 in a row. I wanted to kill myself.
|
you won't believe, but mostly of my matches are against zergs. I even made a post in the bug report forums because I was only getting zergs. I got 10 zvz in a row once, and every then and now I get like 5 or 6 zvz in a row. that is retarded.
|
Odd how that shit happens for some ppl and not for others.
|
On September 14 2010 00:58 Chill wrote:I play random, and it seems like I only get Z when it's ZvZ. It's so upsetting because I love playing ZvT and ZvP 
Me too! I played against 6 zergs in my last 10 games, and guess what? 5 of them were ZvZs! God I hate that matchup so much. I'm considering just 6 pooling every time I play it now
|
Doesn't it also have to do with what maps you have veto'd
|
On September 15 2010 03:13 JudoChopper wrote: Doesn't it also have to do with what maps you have veto'd
Well you'd think that you'd get more mirrors based on certain races will veto certain maps.
I have vetoed...
Steppes of War (more Zergs veto I believe) Scrap Station (ok, ya got me, this is a reason I'm not getting SOME ZvZs! Hate the map!) Kulas Ravine (havne't played since the veto feature came about, as mine ALWAYS worked 100% of hte time and never had issues like others, and am never going back. Most zergs have this vetoed).
PS. I've gotten two ZvZ's lately :o. Although they were both vs the same person, I only played a few games, and I only played 4 games... 50% Z wtf?! Ha jk :p
|
|
|
|