This is a fair point, but I think it gets more complicated if betting becomes involved, because the players may use the same dice, but the player who understands how the dice work has a betting advantage.

Differing beliefs in how games of chance work may still be an arbitrage that works to the advantage of casinos and lotteries.

There was a post on here awhile ago[0] about using loaded dice to cheat at Settler of Catan. All the benefit goes to the person who knows the bias.

[0] https://news.ycombinator.com/item?id=15921322

This makes intuitive sense with just a basic level of understanding.

A biased dice may appear random, but over time youre bound to roll a given number or set of numbers more often. Over time, it adds up to a big advantage.

In many games greater familiarity with the biases of the dice would be no more an advantage (sometimes considerably less) than greater familiarity with the game rules and strategy. It's not clear what games were played, or whether they were played repeatedly amongst groups or by strangers competing against a "house". The article does suggest the biases were visible enough for such games to not hinge upon a particular participant's false believe in the uniformity of the dice though, as well as hint most people's betting strategy probably wasn't sophisticated enough for them to care

This depends on the rules of the game. To use a classic example a coin flip that simply requires HT vs TH (TT or HH get new flips) is unbiased even with a biased coin.

Assuming people had less trust in these dice you end up with different games.