Random is broken


I got the 2nd time in my life 5 times IN A ROW Viper with ultra 2.
That is a chance of 1/(78^5). It’s unlikely you’ll ever get that in your entire life, I got it TWICE already, and another 4 times streak with another character.

I play random exclusively, and my character statistic shows not a equal usage AT ALL. the random button is everything, but not “random”. wonder how that damn thing works.


the random result is probably related to your past inputs or the system time. anyway the pseudo random sequence doesn’t seem to be uniformly distributed. I also experienced in some fun sessions very unlikely series of character choices.


The wonders of probability… which will never be understood…


Stop trying to make problems out of things that don’t even need to be made problems out of.

Just pick a damn main, voila. :coffee:


If you don’t like the characters that pop out when you pick random then don’t pick random and pick a real character


So you picked random, and then you are angry that the selection doesn’t make sense?



I came into this thread hoping you would be talking about how Cammy, Akuma, Seth etc are broken characters because they are based on luck instead of skill (In a ratio of majority)

Damn ;(


I pick random so I get a random character. seeing any character has a chance of 1/39 every time (if you include ultra, that’s a chance of 1/78). I just don’t want the same character(s) over and over and over again. I swear, I get Oni more then anyone else but I could count on one hand how often I get, Cody for instance. It also seems that I get 9/10 times U2 for Balrog and Gen ._.

Random IS my main. I just play the game for the lolz. I’m actually decent with a good portion of the cast.

where are they based on luck? Akuma is the only special character imo, in that you doesn’t play him anymore, but only “use” him. Setups, Setups, Setups nothing but setups.


I’d like a roulette based random feature instead of an invisible option. it happens to me too, sometimes the same character will appear 1/3 of the time.


Last time I played my cousin in a home game we agreed to pick random, which I almost never do, and the game gave us both our mains with the correct ultras :o




I just played again, first character: oni.

played 15 matches or so, and got him 4 times.

god I hate this game >_>


Are you serious?
You’re looking waaaaaay too hard into something that isn’t there.
What exactly are the benefits of playing Random? :rofl:


fun (for the most part).

I always played a good portion of the characters in longer sets, so ended up playing random naturally. now it’s my main. to no surprise - I even won a small all-random local tournament. lol

I’d just love if it would work like it’s supposed to, I always get my weak characters and never my stronger ones (and this is not perception based). tho, that also led to the fact that former weak characters do just fine now. x)


Dude, just close your eyes, move your arcade stick around and select a character, then hold up or down and randomly press a button to select an ultra. Tada. You’ve just figured out the original way to do random select.


that strips the “I beat you with random”-feeling tho. and I pick the character, so I can’t blame the random button anymore :3


deal with it, IT’S MARVEL BA… oh wait. wrong game: fix that broken shit capcom!!!11
inb4: super street fighter 4 random arcade edition v2013


This is a misconception of probability and is a common mistake by those inexperienced in actual probability. In fact one of the rules of randomness is you actually expect streaks to happen!

You are right the chances of getting Viper with U2 5 times in a row is that number…for a sample size of 5. But this is not happening in a vacuum!

Let’s say for the sake of simplicity there is a 1/10 chance of getting Viper with U2. The chance of getting her 5 times in a row out of 5 attempts is 1/10^5 or 1 out of 100000, but out of 6 attempts the chance of getting 5 in a row changes. Why is this? The number of permutations that give 5 in a row changes

H=Getting Viper
T= Getting someone else


We see there is now three arrangements that gives 5 in a row. The chance of getting 5 out of 6 is 18/10^6 and the chance of getting all 6 is 1/10^6 so the total chance is actually 1.9 out of 100000 or almost double the chance (1 out of 50,000).

The math gets complicated fast, but suffice to say if you do something long enough you are bound to find a streak eventually. In fact did you guys know one of the ways people can tell a set of numbers is not randomly generated is by the ABSENCE OF STREAKS! Because probability dictates that eventually by random chance streaks happen.


Streak happening != Lack of randomness

That is a very common myth. You would have to do a much longer trial to show that the random numbers are not random.


If you don’t think it’s true I question the knowledge of your own character.

“Oh, She jumped over my head, time to block the crossu-”

NOOOPE 260 damage LETS GO


I’m well aware of how you calculate probability, that doesn’t change the fact that the probability for that streak is exactly 1/(78^5), which as I stated, is so low that it equals to win 200 times the lottery or be killed 51556 times by lightning (attention, naive examples). and yeah, I got that exact streak twice already, the odds for that… btw, I got the first one when I started playing and rage quit during the 5th match and stopped for a week. I was THAT mindfcked.

as for “random not being random”, I have well over 20k fights with my random account and there are more then a handful of characters that have almost double the uses of others. that is a decent sample size. Might be still not sufficient to use that as a profound disproof, but it doesn’t reflect the empirical value of a ~1/39~ spread at all. sadly, ultras selected are not saved, would be interesting if that comes close to 50%.