One of those cold-calculus-of-war things is that if you’re out-producing the enemy by (say) 5x, you can suffer 4:1 costs on an operation and still be “winning”

One of those cold-calculus-of-war things is that if you’re out-producing the enemy by (say) 5x, you can suffer 4:1 costs on an operation and still be “winning” in terms of causing more harm to their war effort than to yours. That is, you can spend four abstract units of war-fighting-capability to destroy one unit of theirs, and that’s still a net-loss for them. If you don’t have more effective ways to spend that (abstract) productivity, it may still be in your best interest to use it that way. It can make evaluating things like this tricky. So a raid that lost dozens of bombers and cost weeks in training time only put some part of their industry out of action for a couple weeks and diverted a week or so of other factory productivity for the rebuild and retooling effort—that looks bad, but could still actually be a deal you should take every time under certain circumstances.