The short answer: Game Theory.
Consensus networks like the Ethereum blockchain that record data and transactions as a single source of truth seems like it would be a great fit for multiplayer games. A natural suggestion would be putting games involving chance, such as poker, backed by the block chain and guaranteed to be fair by the blockchain network and cryptographic hashes. Why hasn’t the blockchain disrupted online poker and similar games?
The reason is not a technical problem, but more of a game theory problem. Due to the pseudo-anonymous nature of the blockchain, it is impossible to perfectly know the identity of each user and prevent collusion among participants. To put it another way, if you were to join an online poker table, it would be impossible to know whether the other players were other individuals, or if it was one guy with a dozen accounts. The fact that someone could create multiple identities and use that to his or her advantage is a difficult problem to solve.
People are working on solutions to this. Anyone dealing with any kind of financial regulation typically must comply with Know Your Customer (KYC) laws. While KYC compliance is not new, the tradeoff is that these processes are expensive and time consuming to implement. There are blockchain systems being developed that will rely on KYC to operate fairly, such as with FirstBlood, and it will be interesting to see what solutions they come up with.