Is Hearthstone is locally random? That is: for a given time/SR band, does the deck archetype that any given player is likely to face when they press the play button distribute randomly. I think you’re being a bit glib by claiming that it isn’t, but that’s probably a less interesting discussion.
Is Hearthstone globally random? That is - should deck archetypes distribute across the ranks in some random way. Your claim, based on your simulation, is that it shouldn’t.
Your simulation definitely shows a structure that appears over time, but that’s not mutually exclusive with randomness. It’s also not necessarily indicative of how the game works.
The reason you’re seeing these structures in your model is that a rank distribution composed of bands where each band is followed by a band that beats it (paper < scissors < rock < paper < …) and each band is sufficiently wide is stable under the rules that you introduced. A randomly evolving system will tend to end up in a stable state.
There’s a lot of randomness here though, because there are many stable configurations. How many bands should there be? How should their widths distribute? The stable configuration you end up with will heavily depend on random choices early in the process.
The model’s also not indicative of something you see in the game. You don’t really see bands of Deck rock < Deck paper < Deck scissors < Deck rock… It’s more of an artifact of the model than reality, and a lot of this structure will disappear when you tweak the parameters.
As an aside - we still don’t seem to be using the word random in the same way, and we both agree on quite a bit, including that the game isn’t rigged and that the laws of probability and the existence of a matchmaker will force structures on a meta.
Well, I do not want to be mean, but if you are playing quest murloc with shudderwock or similar decks to attempt your first time Legend, you’re not doing it the right way.
I get what you’re saying, but people claim their decks are scanned and they are matched based on card choices. So every time they queue up is another scan and another opponent selected to beat that specific deck. Which is fine I guess, but clashes with the other claim of a forced 50% win rate IMO.
