RE: Magicdice game over!
It's not fishy because you know exactly what the odds are and since they use a blockchain there is no way to fraud. There is not a way to make a casino more transparent than this.
However, their whole business model is based on having an edge over players. If you play enough games you will lose eventually because the law of large numbers comes into play. For the same reason martingale strategy doesn't work. https://en.wikipedia.org/wiki/Martingale_(betting_system)
The only way to make consistent money with a casino is by being the casino. You can invest into the bankroll on moonsteem. It's practically guaranteed to make you money for the reason that the house has an edge.
On magicdice you can double your money with a 49% chance to win. That means you win 49% of the time and lose 51% of the time long term. Therefore you have a negative expected payout ratio when gambling. Since there is no skill involved in this game you can't overcome the house edge and will lose money over time.
The only casino games that are an exception to this are poker and other games where some type of skill is involved.
Martingale (betting system)
A martingale is any of a class of betting strategies that originated from and were popular in 18th century France. The simplest of these strategies was designed for a game in which the gambler wins his stake if a coin comes up heads and loses it if the coin comes up tails. The strategy had the gambler double his bet after every loss, so that the first win would recover all previous losses plus win a profit equal to the original stake. The martingale strategy has been applied to roulette as well, as the probability of hitting either red or black is close to 50%.
If you lose a bet with a 94% win rate two times in a row is .36% which is very unlucky and unlikely. It's just math but I'm not a coder so Idk how it works.
It's not unlikely at all over thousands of games, more like the contrary. If you play enough games it's very likely to happen because law of big numbers kicks in.