If you are having trouble understanding the economic mechanisms that create self-balancing behaviour within the Saito network, pay attention to the slides in our presentation on proof-of-transactions starting on page 18. These slides can be downloaded from the sidebar on the right.
One thing many people miss when trying to understand Saito is why increasing the amount of money we pay miners increases network security. Blockchainer people have a rough sense that paying for more mining *should* secure the network, but they tend to take this on faith instead of understanding the underlying mechanisms.
And there are a few ways that mining secures the network, but the primary mechanism is that when miners propagate their solutions back into the network, they do so in the form of regular fee-paying transactions. This raises the question of what fee miners should attach to their transactions. Miner shares of golden tickets constitute an impressive chunk of the entire network payment, so any pressure that can induce miners to part with a greater portion of the haul should speed up block production measurably. And what increases this except for competition between miners? In large part, we expect miners to pay fees in proportion to the amount of competition they expect. A miner who believes it is likely another miner will find a competing solution will pay a very high fee in order to ensure his own is included, while one who expects zero competition will pay only enough to ensure the transaction is broadcast deep into the network. Increasing the paysplit encourages miners to pool together to support the pro-miner block, encouraging more competition and — through the higher fees generated — a faster blocktime.
Another important insight into the economic forces at play in Saito is the realization that full-nodes and miners are pitted in constant battle with each other for the fees collected by the network. Over the long-term, members of both groups wish to collude with their peers to shift the paysplit in their direction, yet in the short-term they may be willing to support a block that votes in the opposite direction because it offers an opportunity for an immediate payout. Full-nodes who support pro-miner blocks are essentially defecting from group-consensus. Miners who agree to hash on pro-node blocks face less competition from their peers and thus earn higher short-term income. The economic pressures push the network to the point where defection and collusion is equally profitable on both sides.
In the presentation above, we explain this in the abstract terms of when colluding versus defecting from group consensus makes sense. The primary point is that miners will naturally adjust the difficulty to the point there is one solution per block. This difficulty will reflect the highest possible difficulty possible at the . In order to overcome the longest-chain attackers must be able to produce the longest-chain at this level of difficulty. They need not only , but equal hashpower to the rest of the network. In other words, paying miners more incentivizes increases in the network difficulty that make it harder to change consensus variables and require greater hashpower (in addition to a greater flow of transaction fees, and network presence routing them) to attack the network.