Random Commander Real Random?

Hi guys! I choose random commander every time, but I don’t know why it always choose one or two commander every period like a day. If I play 6 to save games, they always choose the same commander for 4 games, that is so weird.

1 Like

Write down what you get and come back when you have a sample size of 100+

8 Likes

It’s not a huge batch, 17 commanders. But if you get one four times out of six, that is indeed weird.

The point is it happens every day!

The point is true random gives exactly this kind of behaviour.

Random minus picked provided what you are asking for, which is a way worse.

And true random is theoretical impossible to achieve code wise, so the fact we have his random work the way it does is already great.

Furthermore, your anecdotal experience would mean nothing to bring change if your sample size is 6 games for a mode with 17 commanders. Go back to your replay list and jot down the last 100.

1 Like

i experienced it too. one day i chose random, i almost always got zagara. that was at a time where i was not very good with zagara as well.

And now you’re a Zagara expert thanks to random commander selection having you train with Zagara many gamess :wink:

Yes but in true random this behavior is rare, rather than the norm.

If you rolled a die 12 times a day every day for a month and got 5s and 6s more than 70% of the time every single day, nobody would believe you if you tried to claim the die wasn’t weighted.

1 Like

Except it is rare… People don’t report when they get what they expect, they only report the wierd situations

1 in a million events happen to 300 people in the US.

4 Likes

Yeah what are you taking about @WireBender lol.

When was the last time you saw tons and tons of people reporting this issue?

Not to mention OP isn’t reporting it, he’s just saying it without any actual data. What 6 games? Lol please man.

Nobody is saying this doesn’t happen, but if he is that small outlier than that is still normal behaviour and not a bug. And even if it is a bug, this isn’t the appropriate forum to post it.

Inside and out there is no reason to make this topic unless he has substantial data to drive a discussion.

From my experiences, it’s “random enough”. After 2 days of playing, I managed to hit every CO

The main anomaly I got was getting Alarak twice in a row! Which was “extra awful”, since he was 1 of 2 COs at the time (the other being Dehaka) I had on low level (lv1 or lv2)

1 Like

i just tried random and i got a different commander each time.

I don’t think we need more than 100+ sample to test a (1/17)^n event, that does not make sense, we play random commander game for 2 or 3 in one day for normal gamer. And it happens more than 10 days during which I promise I played 6 or more games.

@Arthur47, it can be hard to understand if you haven’t learned this before. Try here: https ://www.khanacademy.org/math/statistics-probability/probability-library

i tried random today as well and got a different commander each time(stetmann, stukov and then dehaka). at least today i got a lot of zerg commanders while yesterday i got a lot of protoss commanders. that suggests that it have a higher chance of selecting one race than the 2 others.

I’m still convinced that the randomization is based on an algorithm of least played vs most played commanders, picking the least played more often than the most. Reason being is I constantly get the same 4-5 CO’s every other match, or Alarak 3 times in row whom I refuse to play on my free will. At one point I got Artanis 7 times in a row and I don’t play him much either because he is too easy to use.

1 Like

it is likely based on the selected level as well. i always choose oblivion express. i might try the same on dead of night later.

I’ve read a couple of different articles about “true” randomizers. Simply put, humans don’t like them.

If you flip a coin 100 times, the most common result would be 50 heads and 50 tails, right? Nope, the odds of it coming up 50/50 is actually around 8%!

You want random to feel better? Make it biased against the choices it already made.

2 Likes

This is true. But on the other hand, “true” random doesn’t exist.

All randomisers are actually just a complex function that usually takes some changing variables such as time or location or a numbered event, and plugs them into a function which can then be run on itself any arbitrary number of times to create a series of numbers.

The problem is, there is a pattern, as determined by the mathematical function. It might be a hard to determine number due to the complexity of the function, but it will still give a pattern. And the problem is it’s very hard to not favor a certain result in a pattern.

Quality random generators are actually very expensive to make, because they have to have layer upon layer of logic to prevent them from getting one number significantly more often than another, to keep them as random as possible. But technically, you can go on infinitely, making it ever more random the further you go, but also more expensive to develop and test, and more resource intensive to run. And then of course there is the ultra complex math behind randomizer optimisation that I won’t even pretend to understand.

1 Like

Very true, I maybe should have elaborated but that’s what I meant with the “true”.

At the same time, my example with the coin is also impossible. The odds would change with the coin, the person flipping, the technique and height, wind, air pressure, temperature, even the proximity of the moon would change it to a tiny degree.


Random fact for those who didn’t know, quite often old games (generally 16 bit or older) would use player action as one of the variables for pulling random numbers (along with other things). Each action would add or subtract from a value that would be directly pulled and compared to a table to get the result.

This was done because the old consoles especially were far to weak to have any kind of pseudo-random generation.