Warrior General Games

how does this work?

suppose two outputs are set 50/50...is it like a coin toss each time? I just had a game where a given situation occurred 4 times--and all four times it went with Output "A" rather than "B"

if that's how it works, then "in theory" you could have an output at 10% and have it show up all 4 times too, right? it would be unlikely but possible, eh?

is this working as intended?

Yea you coul have 2 outputs, one set at 10% and other at 90% and the lower one still could fire more.
Yes it's unlikely but it could still happen. 10% chance every time.

And yea working exactly as intended... how else could it work?

Originally posted by Bane


And yea working exactly as intended... how else could it work?

hard to tell if that's serious or sarcasm.

Remember each event is an independent occurrence. So your 50/50 question isn't really all that unlikely. Your theory question. Of course it is possible, but it is statistically less likely to occur than it does with your 50/50 question. Working as intended. You would really need to spend much more than a few games to really determine if the RNG is working as intended. Sorry, as the question is phrased the question is not really answerable.

It's just probability. Here's a thought example that probably won't help but interesting to think about.

Lets say you have a room with 100 people in it, each with a fair coin (50/50 shot). Since each event is independent each person in the room has a 50/50 shot to flip to a head (your play). So you have everyone in the room flip their coin, and anyone who flips a tails sits down. You should end up with about 50 people still standing after the first one. Flip coins again, 25 still standing. Then 12 (just to make the division easier since it is as likely to be 13 still standing) then 6, then 3, then 1. Thats ~6 flips in a row to get down to 1 person. This also means that one person in the room flipped their coin 6 times in a row to heads.

It seems unlikely but it's not impossible. As toobad said, each time it's a 50/50 shot, there is no historical memory on the flips. Picking one play doesn't make it less likely to be called the next time. So every time it's an even shot for either.

So ending up picking the same play 4 times in a row is statistically unlikely (since in my example at flip 4 there were only 6 of the initial 100 still standing) but its quite possible.

Alindyl is correct.

Sometimes, when you flip a coin 4 times in a row, it comes up heads 4 times in a row. Really nothing more complicated than that.

Originally posted by fogie55
how does this work?

suppose two outputs are set 50/50...is it like a coin toss each time? I just had a game where a given situation occurred 4 times--and all four times it went with Output "A" rather than "B"

if that's how it works, then "in theory" you could have an output at 10% and have it show up all 4 times too, right? it would be unlikely but possible, eh?

is this working as intended?

If you want to make it more likely or less likely you can set at something other than 50% OR you could include auto-adjust.

I don't know what you are getting at. Getting something called four times at 50/50 isn't a stretch. Come back when something was called 20 times and I'll think about it.

Originally posted by Time Trial
If you want to make it more likely or less likely you can set at something other than 50% OR you could include auto-adjust.

I don't know what you are getting at. Getting something called four times at 50/50 isn't a stretch. Come back when something was called 20 times and I'll think about it.

Agreed. I play a lot of poker and it is fascinating how "streaks" happen and how they last... then, without warning, die. I highly suspect a random generator with "some" constrictive input (our beloved AI) will produce the same situation... strings of "same plays" and then you won't see that play again for a while... then another "string" of it happening.

What you want to see is, at the END of the game, did your two outputs end up around 50/50 (taking into account how scores and other stuff effect play calling). As long as the end product is close to what you designed, then it's working correctly. Now... if you end up with 100 plays called on that particular output and end up with 20 of one and 80 of the other... then you got a major issue/beef with the game.