The deviation from the expected profit is crazy. I could understand it if the site has only a few thousands of bets, but it has over 2.8 millions bets already. How unlikely is it for a site to have an expected profit of 0.98% and to have an actual profit of -1.91% over so many bets?
It's impossible to answer that question if you don't tell me the distribution of the bet sizes.
If all the bets were the same size, it's very unlikely. If 1% of the bets were significantly bigger than the other 99% then it's much more likely, because only 1% of the bets would be significant to the total site profit.
When Just-Dice was running for BTC we had billions of bets and ended up with an actual profit of around 0.35% when the expected profit was 1%. That's because a small percentage of the bets were huge, and they did better than expected. The billions of tiny bets have little to no impact on the bottom line.
On this app there's been made:
572102 bets @ 0.0001 BTC (on 2x). 282271 have won, and 289831 lost resulting in a net player loss of 0.756 BTC.
543899 bets @ 0.0002 BTC (on 2x). 269188 have won, and 274711 lost resulting in a net player loss of 1.1046 BTC
277403 bets @ 0.0004 BTC (on 2x). 137664 have won, and 139739 lost resulting in a net player loss of 0.83 BTC
140016 bets @ 0.0008 BTC (on 2x). 69327 have won, and 70689 lost resulting in a net player loss of 1.08 BTC
This sort of general trend continues, unless you get to the bets that have only been done a few times, where the player(s) have got rather lucky
e.g.
82 bets @ 4 BTC (on 2x). 45 won, and 37 lost resulting in a net player win of 32 btc
8 bets @ 6 BTC (on 2x). 5 won, and 3 lost resulting in a net player win of 12 btc
so actually, the small bets haven't really done much but added a lot of noise. The investors would be far worse off without them
