Season 13's problem with (lack of) randomness?

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
Post Reply
User avatar
qyzz
Posts: 3
Joined: 01 December 2020, 22:37

Season 13's problem with (lack of) randomness?

Post by qyzz »

Most of the games i play in this season (like 99% of them?) get set of points that includes two 4s and two 3s. What's up with that?
From time to time i get game where there's only one 3, but still double 4s. Only on rare occasion i get game with a single 4. Either i'm extremely unlucky, or there's some problem with random generator.
They're not on the same color all the time, so that part of randomizing works, but they do seem to show up way too often.
Raymond-L
Posts: 6
Joined: 15 February 2021, 00:50

Re: Season 13's problem with (lack of) randomness?

Post by Raymond-L »

There are a total of 4 values a chakra could take: 1, 2, 3, and 4. Of which a value can at most appear twice, making the distribution of values having only 4 combinations: Only having 1/2/3/4 appearing only once.
It is in fact a certainty that at least one of 3s and 4s must appear twice.
Humans are generally terrible at recognizing frequency data. I'd suggest you to look into confirmation bias, frequency illusion, and clustering illusion.
User avatar
qyzz
Posts: 3
Joined: 01 December 2020, 22:37

Re: Season 13's problem with (lack of) randomness?

Post by qyzz »

Yeah, i understand bias and fallacies of memory. Thing is that for the last 3 or 4 days i played quite a lot, and most of the games had two 4s and two 3s.
I know that it is ok to happen, but if that happens in most of the games (there was a long stream of games where it happened in every one of them, although today i had more games with more varied setups, so maybe i was just very unlucky) then it does looks weird :).
User avatar
Sanssouci
Posts: 1
Joined: 04 September 2019, 23:51

Re: Season 13's problem with (lack of) randomness?

Post by Sanssouci »

Hi,
Instead of tracking if there is this combination of points in the game: 4,4,3,3 (so a pair of both 3s and 4s), which seems you've been doing, try instead tracking over multiple games which exact number is used only once. (This is equivalent to tracking which point value out of the 8 possible ones: 1,1,2,2,3,3,4,4 is missing).

In theory, if the values are completely randomized, you should end up with roughly 1/4 games which miss each 1,2,3,4. Of course, the number won't be precisely 1/4, but should be roughly there.

I'd be interested to see such numbers/graph myself if you end up tracking it :)

I'm not sure why are you focused on the 3s and 4s specifically, but tracking *each* missing value will give you a better picture of the randomness involved. Also (since only one value is always missing), you always will have either two 3s or two 4s. In fact, you *should* have both two 3s and two 4s in roughly half of the games! (In the cases when 1 or 2 are missing).

Last note, don't focus on the order or the 'long chain' of missing a specific number too much, these are normal and expected even! Humans are bad at recognizing randomness; at the extreme, imagine this sequence: 1,2,3,4,1,2,3,4,... - not very random, is it? :)

(Just to add my 1 data point of observation, I haven't noticed any such pattern in my games so far. If anything, I can't recall many games with two 4s in them ;) )
User avatar
qyzz
Posts: 3
Joined: 01 December 2020, 22:37

Re: Season 13's problem with (lack of) randomness?

Post by qyzz »

I was thinking about writing browser extension to log stats, to avoid my faulty memory misleading me :).

I also did not have many double-4s and double-3s games last season, that's why it caught my attention.
Post Reply

Return to “Chakra”