What gives a mining pool power to maliciously dominate the system if they're mining over the majority of the coins? Profits are disseminated to miners individually. I really don't understand how this enables a bunch of individuals to conspire to warp the market. I think I'm fundamentally misunderstanding the problem. Can someone explain how having a 51% share enables a pool to prevent, reverse, or double spend transactions?
Conceptually it isn't difficult to understand conflicts of interests (COI) principles. If a pool has over 50% of the computing power required to demonstrate adequate "proof of work" to add entries to the Blockchain Ledger, this absolute centralized power will eventually "absolutely corrupt".
That centralized pool could be used to change the history of what constitutes the longest chain in the Blockchain by computationally orphaning blocks that were legitimately mined by other pools or solo miners. When a pool gets very large, do you honestly think it will not be tempted to modify the nominal Bitcoin transaction priority scheme?
If a pool has roughly a third of the computing power used to populate a Blockchain, they can start to statistically change the destiny of the Ledger, especially if all of its miners are using STM mining protocol. This is because miners are effectively slaved to the set of transactions the pool decides is important to mine, whether legitimate or not. When STM is used, the pool sets the mining block agenda, not the miners. Can you now see where the COI enters into the picture?
If miners use GBT, opposed to STM, they have more choice in the grouping of candidate transactions they bundle together for mining as long as they follow a template. As long as those miners are not colluding with the pool, this will help keep the large pools honest. The problem is GBT coordination sucks more network bandwidth, effectively ever so slightly lowering the profit margins miners can achieve. Most miners are interested in maximizing their profits, and are thus likely not to adopt GBT.
Knowing this information, and looking at the two big to fail pools is very disheartening. A barometer for when the big pools have gotten too large is when numerous small pools suddenly have higher orphan block rates, and their mining profits start to go south of what is expected for known difficulty factors for a protracted period of time.
To keep large pools honest, for the good of the public, the large pools that mine 25 to 30% or more of the Bitcoins need to support transparency with above the board full disclosure policy of what they are currently mining in advance of advertising their lucky block strikes. Such transparency, helps to ensure they are not violating the trust of good Internet netizen miners competitively mining at other pools.
This is a very hot topic. From the flack that I received (i.e., negative Bitcoin Stack Exchange passive-aggressive votes) when asking a prior related question, I believe this is something that must be watched very closely...
FYI - I'm a security consultant and a guy that did wargaming for over 10 years.