One good thing about pacifism is that the view is perfectly universalizable: if everyone were committed to avoiding war, that would be just fine. The problem, of course, is that pacifism is not actually universalized. And, given that some are entirely willing to achieve their ends through violence, it seems like others will need to stand up to them in order to protect the innocent. So goes the standard objection. Is there anything wrong with it?
Let "absolute pacifism" be the claim that engaging in warfare (or, in the strongest version, violence of any kind) is intrinsically wrong, always and everywhere, without exception. I don't know if anyone could defend such an extreme view, so I'll set it aside for now. We might do better to consider a more moderate position, let's call it "epistemic pacifism", which claims that in practice it is never advisable to go to war. The consequences are so predictably awful that we should always prefer to find some alternative instead. Perhaps we can point to a rare case where going to war turned out for the best (though can we really be sure there wasn't any better alternative?), but the epistemic pacifist will nevertheless insist that we couldn't have known this beforehand, so the original decision was criminally reckless, albeit lucky in the end.
I think the standard objection is enough to refute the absolute pacifist's attribution of intrinsic wrongness. As a matter of principle, fighting for the innocent seems a noble enough endeavour. The epistemic pacifist can grant all that though. They simply point out the problem of implementation, and worry that the cure will likely turn out worse than the disease.
It's no doubt a good rule of thumb to be very reluctant to go to war. But it's not so obvious that the rule should be exceptionless, as pacifists demand. There may be clear cases (Rwanda? Darfur?) where humanitarian intervention is justified, and would predictably do more good than harm. If we could set up trustworthy (preferably international and/or deliberative-democratic) institutions that reliably identified such cases, then following their recommendations would be a better approach than straight pacifism. Leaving such decisions to the executives of individual nation states is probably not, however. So we may do best to commit to pacifism in the meantime, and work to create the improved political institutions that are a precondition for making legitimate declarations of war. Sound plausible?