I'll list some data here: I played with payout 1.5-1.6, win chance 66-61.87%. The overall win chance of my lastest 2000 rolls is 65.3%, well, this is fine. But the win chance decrease while the bet amount is doubled:
Bet amount 0.001, win chance 65.5% (>65.3%)
Bet amount 0.002, win chance 66.8% (>65.3%),
---here is where the trick comes in---
Bet amount 0.004, win chance 65.0% (<65.3%),
Bet amount 0.008, win chance 63.4% (<65.3%),
Bet amount 0.032, win chance 64.3% (<65.3%),
Bet amount 0.064, win chance 55.6% (<65.3%).
See? The higher the bet, the lower the win chance!
We can see clearly the trend from 2000 rolls, the trend will be more clearer if the Fortunejack team can provide all my history.
If you insist the data is not enough, the Fortunejack should be obliged to provide enough data against my proof and conclusion.
This is not proof. This is just some data about some rolls that were made in the last 2000 bets.
If (for example) you bet with lower risk at smaller amounts, and higher risk at larger amounts, it makes sense that the higher bet amount as a a lower win percentage in the end.
Also, I would assume that less bets were made with the high value, so it is a tiny sample. You would need a way bigger sample to provide proof.
Lastly, Kiritsugu already explained that the fact that you can verify the rolls already makes it impossible to be rigged.