The PvP problem, and how to remedy it
PvP gambling can be a headache for both casinos and players. What’s the PvP problem, and how can we figure out ways around it?
The problem for players
A traditional off-chain casino will find it hard to create a perfect PvP (player-versus-player) system. This is because of a systemic risk: that the player is actually playing against a malicious house. You can join a PvP website and find various open games ready to join, opened by what seem to be normal players. Many of them might be normal players, but many of them might also be the house itself: you wouldn’t know, since there is a low barrier to entry in terms of creating a PvP match against someone.
What is so wrong with playing against the house? The answer is both obvious and important. If you’re playing against the house posing as a normal player, and the house determines the outcome of the game, then the house can rig the game against your favour. You have no guarantee that the game is fair or that the odds are actually as advertised. As a player, you’re essentially going in blind, and the entire game is setup against you. That’s why it’s not very smart to play on a PvP gambling site where the house itself is able to determine the outcome of the game. It usually doesn’t end well.
Players have two main ways around this. The first is to play against someone known and trusted, like a friend or a community figure. This works to an extent in close-knit communities, but it falls apart when you try to scale such an operation to a higher level. With a high-volume, fast-paced PvP operation, most players won’t be able to get matched up with a friend or trusted individual whilst playing. Some (if not most) of them will end up playing aganist total strangers, and we once again encounter the malicious house problem. So, if you can’t play against a friend, the only way for a player to be safe in a PvP game is to use an external source of randomness. This means using a method to determine a winner that cannot possibly be controlled by the house itself. It has to be bigger, and beyond the scope of just the website: the best example is block hashes. Block hashes are random(-ish) and can’t be controlled by any one entity (unless you’re Vitalik Buterin). An external source of randomness like a bitcoin block hash can ensure that the outcome of the game is decided in a truly fair manner. The one caveat is that the match prize can’t be significantly larger than the block reward, but that’s a pretty rare circumstance. Another way of making a PvP game provably fair(-ish, once again) is to use the classic server-client seed setup (further explained in this article). The game’s outcome is predetermined, and is pulled from the usual set of games that the casino has, which the house knows but cannot feasibly change without breaking the provably fair system. This is only fair for one party, and cannot be fair for both: it depends on who is deciding the outcome to bet on. Another very simple way to make PvP games fair for both parties is for one party to commit to their pick beforehand with a one-way hash function (like SHA256), and for the other party to verify the provided string after making their choice.
The problem for casinos
One word: volume. Compared to versus-house games, PvP gaming is virtually dead. You need to find a set of players with the willingness and ability to bet money against each other, which is pretty rare to find in sustainable quantities. Sure, a casino may find bursts of PvP wagering, but it usually doesn’t last very long. The numbers show that versus-house gambling is far more popular. BTCArena, a very popular PvP gambling site, only has around $150,000 in wagers, whereas flashflip (which is both new and not as well known) has over $600,000 in wagers. PvP games come with the inherent problem of being slow (in that you need to take the effort to find a willing opponent, or wait for one to drop in), and if the site is using an external source of randomness like a block hash, waiting for the game result to resolve can be painfully slow. It’s just logstically easier and faster to play an instant game against the house.
There’s a way to remedy this, which may not have been extensively tested yet, but is about to. If you want someone to make use of a product, you usually need to provide a compelling financial incentive. flashflip is about to, for a 3-month pilot, transition into free PvP gambling: zero house edge, zero bot commission, and zero fees. For what might be the first time, players can wager bitcoins against each other without having any edge whatsoever and without the players losing money as a whole unit. This has never been done before on a large scale, and it’ll be very interesting to see how it plays out. Most probably, PvP wagered volume will see a significant upwards push: it’s a fun game theory experiment to see how much people value money over time. Are players willing to accept a 1% house edge if it means they can play instantly, whilst ignoring the opportunity to play with a 0% edge if it means waiting a little bit? The answer remains to be seen.