Not sure what level OP is playing at, but non-specialization is the killer on my Emperor games.
You should always have some hybrid cities with both hammers and commerce, but the fact of the matter is without specialization, you nullify national wonders such as Oxford, HE, Ironworks, and Wall Street (you can't always go with corps or have a far spread shrine).
The second issue with this is the sheer # of
that are REQUIRED to devote to multiplier buildings. Since you have only so many
coming in from all sources, you NEED to have all multipliers working in all cities so that you're properly boosting ALL functions of your empire. You have to build library/market/grocer/univ/bank/observatory/forge/courthouse/factory in ALL ur cities.
Of course, once you dominate an entire continent or overtake 3 civs, this becomes less needed. But if you're doing so that early on, perhaps a diff level increase is in order.
And just how would one of these legs magically fall off of the Stool of Power? Generally speaking, a heavy CE is fueled mostly by a cottage spam bureau capital which better be able to defend itself. SE is nearly impossible for AI to dismantle the science sources since you can swiftly replant farms. So it's not very easy for the AI to snag that leg. Military should always be a priority and with a specialized Heroic Epic focusing entirely on food and hammers, you should have plenty of military. Not to mention the hybrid cities spitting out units sporadically and any farm heavy whip/draft cities for the icing on the military cake. Hows about GP Farm? I've never had a game in my 5 years of CIV where my GP farm was in danger of being pillaged or captured.
Really, the only way the AI can damage one of your "stool legs" is via military. If you know how to play CIV, then you know you need military. Also, the loss of 1 city in a non-specialized civ is the loss of more
than you want to count (imagine losing courthouse/market/grocer/bank/library/univ). All those
coulda gone to military and not even lose the city in the first place.
NWP20 said:
Now perhaps this was "sub-optimal" and I could have won faster/better but I only had three "specialised" cities (in an empire, on 15 cities before the final push), and they were only specialised because I got two shrines in one (therefore wall st), and settled my scientists in my GL city (therefore oxford)
If you've specialized 3 cities, then you didn't do what the OP posted (originally). I'm pretty sure most players only specialize 3-5 cities and the rest are hybrids. Specialization is only optimal for Oxford city, Wall Street City (specialization not even required in many cases), HE city, Ironworks City, and maybe another cottage spammed city or 2. After that, hybrids are generally the best way to go in most situations. The ability to support itself, provide a bit of
, and produce units/spies infrequently is all I ask of 2/3 of my empire.
That's still specialization. Try playing the game again without specializing Oxford, Wall Street, HE, Ironworks and you'll see the story changes significantly. Also, your neighbors greatly affect the ability to spam multiplier buildings in all cities. If you've got Mansa Musa, Gandhi, and Roosevelt as neighbors, then you can easily devote
to all those buildings. But try that strategy when you've got Shaka, Monty, and Catherine as neighbors. The lack of units due to devoting
to buildings will spell almost certain doom even on diffs as low as Prince.
And you need some way to get to rifling first. So all the discussion about specializing definitely figures into that 
