When to lose units

[void_genesis]

I have been getting into warfare more recently and finding the changes in Vox populi to be excellent. The way the different units mesh together with the landscape is working really well now. One thing I have started to wonder about is the impression I get that one should never lose units (especially on higher difficulties) if you are playing the game well. Is this still always true? As an example I was playing Assyria recently, had cities with excellent production but poor gold (no trade routes possible since everyone close by wanted to kill me) and my science was great. This meant it was easier to build new, highly promoted and up to date units than it was to bring my half dead ones back from the front line to upgrade and heal. Sometimes I will also send a scout to get a ruin even if there is a chance they will be mobbed by barbs and killed. Any other thoughts on when it is good strategy to lose your units?

 

[tu_79]

Letting your units for slaughter is never a good idea. It increases your enemy war score. It is a better idea to produce new units than to upgrade, specially when the newest units begin with the same level. But you don't need to sacrifice your obsolete units. Gift them to city states (only at full health, mind you) or disband them in your homeland for a few bucks.

An exception could be made if you think that sacrificing some old units could allow you to lure a city garrison or capturing an enemy great general, or breaking an enemy citadel, that kind of things.

 

[Tekamthi]

breaking a citadel, or a key road, is probably the only time I intentionally lose units vs AI... though even then I'll try to have at least one +movement unit to get in and out.. vs other humans I'll accept losing units on a mathematical basis, making a rough assessment at what's "in theatre"

 

[void_genesis]

Are "on conquest" bonus civs like Assyria and France more of an exception? I also felt it was worth risking them to take cities for those civs. The huns might also be a case where they have so many trash level barbarian units that they can afford to lose a few. Are there any remaining bonuses for losing units? IIRC there used to be a religious belief that gave faith for lost units but it may have been removed.

 

[tu_79]

Are "on conquest" bonus civs like Assyria and France more of an exception? I also felt it was worth risking them to take cities for those civs. The huns might also be a case where they have so many trash level barbarian units that they can afford to lose a few. Are there any remaining bonuses for losing units? IIRC there used to be a religious belief that gave faith for lost units but it may have been removed.

It's still there. I never take it, though.

 

[Noob Fanboy]

Only if you win more losing that unit, for example, using a cheap unit (specially scouts) to bait a archer or a city focus, breaking roads, citadels, etc.

 

[ElliotS]

It all depends. Sometimes losing a unit can move the whole front forward. I know that I've had plenty of times where I'm happy to throw a knight into trebuches and archers, knowing it will get murdered, because the overall outcome will lead to like 6 of their units dying instead of 2 and a bunch more turns healing.

If you want to win the war faster, sometimes sacrifices need to be made.

 

[pineappledan]

This rarely happens, but if I have sacrificed mounted units by placing them behind a unit line in hopes of either denying them ZOC or forcing them to lose a turn attacking. Even if the unit ends up dying it can deny the AI an effective retreat, so I have another turn to move the rest of my army in to take another 2-4 units of theirs

 

[Grabbl]

In my experience, "sacrificing" units (that means, leaving them in a position where they will most probably die) can make sense in certain cases.
This includes, for me:
- maintaining a city blockade also at the expense of one or two of the blockading units dying
- pillage strategic tiles (especially citadels and road connections, sometimes strategic or monopoly resources)
- trade key units (especially ranged and siege units, great generals, units much better attacking than defending or single melee units that can otherwise take a city next turn).
- distracting enemy units (to get a turn or two to get my defenders to a threatened city), also with the smarter AI this is possible
- protect my key units with ZOC (especially vulnerable siege units, highly-promoted damaged units or great generals for citadel placing) or protect important tiles (strategic resources, citadels)

And even against the AI, trading units can be beneficial. You may be at long-term production disadvantage, but often the local power increase for a few turns is all that is needed to win a war, especially when you manage to create a difficult supply situation by blocking roads etc.).