EPTRandom Mirror Match Up Percentages Off - Page 6
| Forum Index > SC2 General |
|
Alejandrisha
United States6565 Posts
| ||
|
awu25
United States2003 Posts
| ||
|
Marradron
Netherlands1586 Posts
On April 11 2011 04:29 Hopeless1der wrote: Thats why the OP said to use SC2Gears to verify that your last 10 or 20 or 50 games are in fact mirrors. He's trying to get statistical data to back up his claims and a lot of people are just shitting all over him for trying to get some help. Yeah thats gonna be biased anyway. Anyone who doesnt feel the problem isn't gonna upload replays. I sure won't. Just look at the poll. It's completely biased. | ||
|
shinarit
Hungary900 Posts
On April 11 2011 02:14 Gnax wrote: That doesn't answer the question one bit. I didn't ask what the chances of playing against protoss is. Try again. To answer your question, yes, you still get a theoretical 1/3 chance, simple math: add up the chances, assuming you roll each race for yourself with 1/3 chance and the player base is distributed on your MMR level as you stated above: If you roll T/Z, your enemy will be T/Z with 4%, with P its 92% (you must add the random's 1/3*playerbase chance too), you add up, divide by 3, and you got your 1/3 chance  On April 11 2011 04:40 awu25 wrote: You can only assume that you should get a 33% chance of mirror if the race distribution is even, which it definitely is not. Same for you. The thread is about random players, they ALWAYS have the 1/3 chance (in theory ofc). | ||
|
Gnax
Sweden490 Posts
On April 11 2011 04:44 shinarit wrote: To answer your question, yes, you still get a theoretical 1/3 chance, simple math: add up the chances, assuming you roll each race for yourself with 1/3 chance and the player base is distributed on your MMR level as you stated above: If you roll T/Z, your enemy will be T/Z with 4%, with P its 92% (you must add the random's 1/3*playerbase chance too), you add up, divide by 3, and you got your 1/3 chance Same for you. The thread is about random players, they ALWAYS have the 1/3 chance (in theory ofc). Thank you. So that means that people can stop using arguments like race distribution to prove that the OPs stats are not valid. Even tough it's a small sample size, with more people adding their own stats it will indeed be a better sample size at some point, asuming people are telling the truth ofc. | ||
|
AmericanUmlaut
Germany2577 Posts
On April 11 2011 04:30 shinarit wrote: Yeah, pseudo-random is much more evenly distributed than real random... So it would be surprising to find anything wrong with it... though we dont know the exact algorithm (or at least i dont, point me somewhere if i can find it there, straight from Blizz possibly). This whole thread is incredible, but as a programmer, this post is my favorite. d Pseudo-random, by definition, is not random but statistically cannot be distinguished from random. A pseudo-random number generator that created numbers more evenly distributed than real random numbers would by that very fact reveal itself as non-random and could then be predicted to some extent without knowledge of the algorithm used ot generate the numbers. Thus, the generator would not be pseudo-random. Thus, you are full of crap. QED. | ||
|
ThaZenith
Canada3116 Posts
TvP TvZ PvZ - 3 Non-Mirrors Shouldn't you get 50% mirrors? Where are you guys getting the 33% from? | ||
|
Gnax
Sweden490 Posts
On April 11 2011 05:02 ThaZenith wrote: TvT PvP ZvZ - 3 Mirrors TvP TvZ PvZ - 3 Non-Mirrors Shouldn't you get 50% mirrors? Where are you guys getting the 33% from? More like TvT PvP ZvZ - 3 Mirrors TvP PvT TvZ ZvT PvZ ZvP - 6 Non-Mirrors | ||
|
Barca
United States418 Posts
Even after that i'd say 60% or more are mirror matchups PvP sucks | ||
|
JDub
United States976 Posts
On April 11 2011 05:02 ThaZenith wrote: If you play Random, you have a 33% chance of rolling the same race as your opponent. For one player, TvP is different than PvT (in one you are the T, in one you are the P), so your count should look like:TvT PvP ZvZ - 3 Mirrors TvP TvZ PvZ - 3 Non-Mirrors Shouldn't you get 50% mirrors? Where are you guys getting the 33% from? TvT PvP ZvZ - 3 mirrors TvP PvT TvZ ZvT PvZ ZvP - 6 non-mirrors. 33%. That said, there is nothing wrong with Blizzard's random number generator -- this thread is actually ridiculous. Small sample sizes and massive confirmation bias and the psychological effect of people disliking mirrors and over-estimating how many times they play them are all combining to have a ridiculous result. If you really want statistics on it, why don't you just do a custom game, play as random, and see how many times you get each race. Do it 1000 times, and I bet you will be right at ~33%, ~33%, ~33%. | ||
|
Humoresque
United States5 Posts
However the sample size of my ladder games is pretty small so it doesn't say much. | ||
|
Yoshi Kirishima
United States10345 Posts
If I were you I'd just trust that Random is indeed Random and the matchmaker does not match players based on race. Yeah thats gonna be biased anyway. Anyone who doesnt feel the problem isn't gonna upload replays. I sure won't. Just look at the poll. It's completely biased. Yup, this is a big problem regarding the "accuracy" of studies/surveys etc. | ||
|
nkr
Sweden5451 Posts
| ||
|
FinnGamer
Germany2426 Posts
| ||
|
Arkless
Canada1547 Posts
Over the past two days I have played 48 games, 6 of which were tvt, and one of those tvt's was vs random. Also, looking at my ladder I just noticed I am the only terran in the top 8, and 1 of 3 terrans in the top 25. | ||
|
JDub
United States976 Posts
On April 11 2011 05:18 Arkless wrote: Please, read the OP. He is talking about Random players only. However absurd the premise of this thread is, it still has nothing to do with race distribution. I don't understand why people keep doing this.I think this is very race/league dependant. And we need some more info. Master's are riddled with P and Z, and I almost never get to mirror anymore (i play terran) Over the past two days I have played 48 games, 6 of which were tvt, and one of those tvt's was vs random. Also, looking at my ladder I just noticed I am the only terran in the top 8, and 1 of 3 terrans in the top 25. | ||
|
dmillz
Canada270 Posts
On April 11 2011 05:17 nkr wrote: 44% in 80 games is actually an indication that things are working as intended. If you have played 8000 matches and you still have 44% mirrors, then we might start speculating. Exactly this. Nobody has played enough games to say whether it is "random" or not, not even close. | ||
|
DerNebel
Denmark648 Posts
| ||
|
CakeOrI)eath
United States327 Posts
+ Show Spoiler + W L P 2 3 T 2 2 Z 5 4 T vs + Show Spoiler + W L P 2 1 T 4 4 Z 6 1 Z vs + Show Spoiler + W L P 0 3 T 2 1 Z 3 6 I've stared logging all my games since reset in a spreadsheet. When I first read your post I thought you were crazy, but my (admittedly small) dataset agrees with yours. 22/51 = 43% I'm so rusty on my statistics I can't give you a precise confidence interval, but I think its a safe guess that 43% is well within it. This means I still think you're crazy but I'll be interested in tracking this over the whole season. | ||
|
Azide
Canada566 Posts
| ||
| ||
