Balatro

Balatro

View Stats:
Wheel of fortune is rigged
Having studied it properly, decided what hypothesis I'm testing in advance and collected data it's pretty obvious that wheel of fortune is rigged. Deciding to record its uses in advance and recording every failure and run over several events makes it possible to determine to a reasonable (0.05) level of statistical certainty that the developers have rigged this card. Unlike intuitive approaches or working out the probability of an event after it is happened this is a much more reliable approach used in a variety of sciences and academic disciplines. In some fields these results would be good enough to publish (on a statistical level rather than a novelty of results level) I am highly confident that the chances of this card working are not a matter of skill but have in fact been fixed by the game developer so that it will fail slightly less than 75% of the time.
< >
Showing 1-15 of 21 comments
This post breaks down what happens when you use a Wheel of Fortune. Also, you're required to disclose whether or not AI wrote your post.
Goblin Feb 10 @ 7:11am 
Originally posted by ruler2k2k2:
This post breaks down what happens when you use a Wheel of Fortune. Also, you're required to disclose whether or not AI wrote your post.
This is a joke thread, bucko. Lots of intentionally grandiose language that says very little, only to end with "it's actually very slightly rigged in your favor" to subvert expectations.
Moray Eel Of Fortune, more like it.
Wheel of fortune is

Wheel of fortune does
Goblin Feb 10 @ 7:36am 
Originally posted by Moray Eel:
Moray Eel Of Fortune, more like it.
When you get hits on Wheel, and it gives a good feel...
That's-a moray~
If gamblers fallacy isn't real then can we really say wheel is 1/4? Checkmate liberals.
Originally posted by RazzberryMocha:
If gamblers fallacy isn't real then can we really say wheel is 1/4? Checkmate liberals.
Datamining. We can literally look at what the code says is supposed to happen, and have. (I can hope you're joking, but the Flat Earth Society exists)
Originally posted by HeraldOfOpera:
Originally posted by RazzberryMocha:
If gamblers fallacy isn't real then can we really say wheel is 1/4? Checkmate liberals.
Datamining. We can literally look at what the code says is supposed to happen, and have. (I can hope you're joking, but the Flat Earth Society exists)
Wheel is listed as a 1/4 chance which means that for we should expect one out of four wheels to work. Hypothetically, if I spun three and all of them were nothing then the fourth one is still a 1/4 chance, meaning that statistically the odds say it shouldn't work but the sample size says statistically it should.
In conclusion, wheel isn't actually a 1/4 chance. It's actually a, uh, a 1/4 chance.
I'm deadly serious about it not being 25%

I'm pretty confident that if you could gather the data on every wheel ever spun across all games played by all players you'd find it would come out a little above that.
Originally posted by x_equals_speed:
I'm deadly serious about it not being 25%

I'm pretty confident that if you could gather the data on every wheel ever spun across all games played by all players you'd find it would come out a little above that.
In an experiment with 400 wheels spun they actually had the opposite, with a 75.5% fail rate. Of course, literally all statistics have a margin for error in them (even the census, something that is literally just counting people, something without any meaningful variance, has a margin for error).
https://www.reddit.com/r/balatro/comments/1gwkol9/a_study_on_the_probability_of_wheel_of_fortune/
I believe you, but I'm still pretty confident that if you could gather the data on every wheel ever spun across all games played by all players you'd find the success rate would come out a little above 25%.
Last edited by x_equals_speed; Feb 10 @ 12:07pm
srn347 Feb 10 @ 12:30pm 
Some of those WoFs used would be in runs where the player had oops all 6, so yes that's correct.
That is exactly the reason for my confidence, good job :D
Originally posted by srn347:
Some of those WoFs used would be in runs where the player had oops all 6, so yes that's correct.
There's probably one specific run that had some way of getting Oops All 6s twice and also Perkeo screwing up the math even harder.
Adam Heron Feb 11 @ 11:34am 
Originally posted by HeraldOfOpera:
Originally posted by srn347:
Some of those WoFs used would be in runs where the player had oops all 6, so yes that's correct.
There's probably one specific run that had some way of getting Oops All 6s twice and also Perkeo screwing up the math even harder.
AFAIK Oops All 6s isn't an incrementing buff, it just sets a value. Multiples don't do anything.
< >
Showing 1-15 of 21 comments
Per page: 1530 50

Date Posted: Feb 10 @ 5:28am
Posts: 21