upgrade units?

[Addled Platypus]

did you by chancehit the "upgrade all" ??

[Nikomakkos]

Well, isn't this mostly a question of circumstances? If your land is such that you have plenty of production, then disband old units and use the gold for beakers. If you have limited production then hold on to the units you've already built. Isn't this sort of obvious?

But one thing: Remember that units cost maintenance. Sometimes it's better to disband a good portion of them as soon as you don't need them anymore instead of saving them up for a rainy day. This could mean 10 or 20 extra % on the science slider. It's good to actually have the technology required for the choice of either upgrading or building.

I also think it's a good idea, if the situation allows it, to keep limited forces to avoid maintenance, but have hammers in the bank for units in case of emergencies. What I mean is, in a few cities you build units until there's 1 turn left and then switch to something else, so that you can at any time go back to building the unit and have it the next turn. This of course means you loose the hammers when the upgrade technology comes along, unless you switch back to it right before (thus getting a brand new obsolete unit). But it is a good way of being ready without paying maintenance. I imagine this would be cool in multiplayer. Someone invading you, thinking you have a lot smaller army, and then the next turn you suddenly have all these units. And then you whip and have even more. You could double the size of your army in two turns! (Don't tell anyone about this, I wanna use it myself if I ever play online).

[Blaupanzer]

Not a good idea to disband a lot of units all at once in most circumstances. The AI will interpret the drop in your power score as a sign of weakness. This can bring generally unwanted invaders into your land in 3 or four turns. Hardly worth the maintainance savings if it triggers a war (possibly a multifront war) which will dramatically affect your build queues. Taking the units down at a rate of four (or fewer) per turn while building the new units will not so dramatically shift the power graph while still saving you a little maintainance each turn.

Of course if you are running in the red, that's a different matter. Sometimes the risks are worth it. Also, if you are looking for a fight, it is better to have the AI attack you from the perspective of war weariness.

[wodan11]

Lately I've been experimenting with small, tight empires where I prioritize getting a tech lead. No so much going on a conquering spree, even if I can. I've been looking closely at the expenses in conquering a city or three, and the benefits thereof (and how long before the "break even" point).

As a spin-off of that I've been experimenting with the Great Wall, Imperialistic, Protective, and variations of the theme. Let them come to me, then whomp them, is the idea. This naturally results in a lot of GGs which you can settle in your military city or make a ton of warlord untis. If the former then you'll be disbanding a lot of old units, which as implied above, can convince AIs to come attack. Which is a good thing.

If the latter then you have lots of uber warlord units, which snowballs quickly. A very fun strategy.

In such cases, getting AIs to DOW you is often a good idea. Not that you want to go conquer them, but that you want them to have negative dip modifiers with your friends, and you want them to send their pissant units in so you can kill them.