I think everyone aware that the management servers in OpsMgr 2012 are combined into pools. But not everyone understands how it works. Here, I briefly describe the principles of the pool.

So, first you need to know that the pool operates like a Failover Clustering, a pool is considered "online" yet functional for more than half of its members.

Cases

You have a single management server

If it is unavailable \ failure etc. - There is no pool. The entire pool's functionality is lost.

You have two management servers

In the event of failure of one of the members of the pool is exactly half and in this case, the database acts as the "vote", allowing to form a majority.

You have three or more servers in the pool

The rule of the majority. That is, until more than half of the servers in the pool are available - the pool is "online" and working. Database has no vote. But.  If your pool has an even number of members, when exactly a half of pool members are down a database will have a "vote" and it allows your pool to form a majority.


UPDATE (Feb 8, 2017): https://blogs.technet.microsoft.com/kevinholman/2016/11/21/understanding-scom-resource-pools/