Companies, exchanges and services have nothing to do with Joe's full node in his basement, but rather with people putting money in bitcoin.  As another joker said in another thread, he's firing up 4 nodes, and is going to fire up 4 more nodes, in order to help decentralization in bitcoin  

Miners ARE sensitive to people buying their coins.  They want, logically, to optimize the extraction of value from the eco-system.  That can be by selling few coins at high value, or selling a lot of coins at lower value.  But on top of that, miners are locked up into a Nash equilibrium of steady-state.  It is not simple for them to leave the existing consensus unless they can be in a cartel and decide upon something.


Core was the central authority until a few years ago, writing the only code out there.  So there was no game theory, it was a totally centralized system until recently.    The only "game-theoretical" aspect that was in there for miners, was the signalling system, which was essentially Core taking the lead in the cartel formation.  And ONLY miners voted on signalling.  In other words, Core being the software monopolist, had total power over the rules.  

On top of that, they only used soft forks, of which it is known that a majority (> 50%) of hash rate IMPOSES the rule onto everyone.  They decided about the possible soft forks, because they had the code monopoly, that miners used too.  They were the moral authority, inherited from Satoshi.   The changes they wanted to implement, were proposed as soft forks on which a MINER signalling was applied ; when they reached 95%, the soft fork "fired", and even if a significant fraction of miners would try to step back, they were then locked into the Nash equilibrium of the new rule set.

So essentially, Core always got what it wanted ; except for ONCE: segwit.  Their tactics didn't work, the MINERS didn't signal 95%.  The former core leader, Gavin, broke the software monopoly with Classic.   These two events signalled the start of the end of the code monopoly.  From that moment on, bitcoin started to decentralize, and game theory starts to set in.  Until the block debate, bitcoin was an entirely centralized entity.


Nope.  Core decided on what to signal on, and miners voted.  No "negociations" with anybody else.  Core was used to miners just run their software and accept everything they put in there.  But the block debate showed them the end of their central power to an extend.


On the contrary, it is exactly part of the design of the system.  Consensus by miners alone, under the proposition of the central power, Core, has ALWAYS been the sole path to protocol modifications.  The "boss of the miner cartel" was Core.  The cartel voted by signalling, and that was a smart system, because their vote engaged them ; they couldn't just vote and not act, because the sole software that was out there was also the software that counted the miner votes.  Soft forks imposed the majority wish on all.  What happens now is that the miner cartel sees that it doesn't need the Core to be the boss of the cartel.

From the moment the miners don't need to run the sole Core software, core loses its central authority and bitcoin becomes decentralized ; the game theory kicks in.  Until now, it was a parliamentary monarchy, where only the King could propose laws, and the parliament (the miners) voted.  There's not much game theory in that.  From the moment the King is gone, it starts.  It is only starting now in fact.


It is maybe not the purpose, but it is the consequence.  Yes, I think bitcoin is a gigantic pyramid game because it was designed that way, even if that was not its "purpose".  It could have been different if it were designed differently.   Again the confusion between purpose and consequence.


They are "waiting" to perform a hard fork because:

1) core hasn't yet completely lost its dominance on software.  In reality, most miners wouldn't trust software on another basis for the moment.  Only Gavin was smart enough to write modifications of the code.  

2) but the principal reason is simply that miners are locked up in the Nash equilibrium of steady state.  Individually, forking off is a losing proposition.  As long as they don't form a strong cartel, they cannot change individually.

At no point, they need Joe's node in his basement for that.  The validators of miner action, are other miners.  They keep one another in check, and this is the Nash equilibrium I'm talking about.  It is only when there's a concerted effort to fork with large majority, together, that this equilibrium can be left for another one.


Indeed, the "system stops" and this is not acceptable, but it is MUCH LESS acceptable for the myriads of economic actors than for the miners.  This is why the economic actors will comply much much faster than the miners.  If the miners can find a majority agreement (cartel formation) and they ALL switch to another Nash equilibrium, the economic actors are all INDIVIDUALLY pushed to switch too.  The economic actors are NOT locked into a Nash equilibrium on the old protocol ; on the contrary, their individual switch to the new protocol is highly advantageous to them.

In fact, the FIRST SWITCHERS (of economic actors) will make HUGE BENEFITS.    The first exchange that "starts working again" will get the lion's part of bitcoin trading.  The exchange that stubbornly refuses to switch and hence cannot allow deposits or withdrawals, because it refuses to upgrade its node to the miner's protocol, will kill its business.  
The whale that has a lot of coins, will be able to transact if he upgrades to the miner's protocol ; he can only lock up his holdings if he refuses.

So the game theoretical motivations for miners and for economic actors are ENTIRELY DIFFERENT.   Miners, keeping one another in check, are locked into a Nash equilibrium ; only a concerted action (formerly done by Core, the software monopolist and central decider of bitcoin) can move them all  together to another equilibrium (change of protocol).   Economic actors, individually, have everything to gain in following the miner's protocol.  Their Nash equilibrium follows that of the miners.  They are out of equilibrium if they decide upon anything else.


Yes, that's game-theoretically to be expected IF miners can form a cartel and decide together, and IF they have sufficient code independence from Core.

This is not "me day-dreaming".  It is simple game theory.  You know that a Nash equilibrium is when every actor is adopting such strategy, that if all others were to keep their strategy, and only this actor changed, that he'd be worse off for every choice he'd make.  This is why economic actors will always follow miners and not vice versa.  A miner cannot make an individual deviation from his peers (fork off with small minority).  He has to agree with a large number of peers.  But an economic actor can switch easily.  He's in the most profitable individual position when he follows the miners protocol.  All his other choices are less profitable to him.


If a lot of miners leave the system fast, bitcoin dies.  Because of the slow difficulty adjustment.

There are many subtle ways in which miners could apply covert soft forks, getting new users to buy coins, and killing off old users.  They are not yet doing such sophisticated strategies, but it would not be difficult for them.

For instance, a soft fork could be, to always include transactions that are only, say, 20 transactions away from recent block rewards, and require higher and higher fees for transactions that are "older".  This would give a very fluid impression of bitcoin for all those buyers of newly mined coins (including fees !) ; and it would only piss off older whales or older bitcoin users, who most probably are NOT buying new coins.   As such, that strategy would induce newcomers to buy coins, and would put a lot of friction on older coins, decreasing bitcoin's velocity, and hence pumping up its price.
If you buy "new miner coins" you would have a great experience with bitcoin, and hence buy more of them ; if you have old coins, you would almost never get a transaction through, and have to pay huge fees.

In the beginning, miners could just apply those rules to their own block building, but they could even start orphaning blocks that are not complying to their rules -> soft fork, imposed by majority.  No "code change", just block selection.