I'm not sure if you understand the real point of precommitment. The idea is that your strategic position may be stronger if you are conditionally committed to act in ways that are irrational if these conditions are actually realized. Such precommitment is rational on the whole because it eliminates the opponent's incentives to create these conditions, so if the strategy works, you don't actually have to perform the irrational act, which remains just a counterfactual threat.

In particular, if you enter confrontations only when it is cost-effective to do so, this may leave you vulnerable to a strategy that maneuvers you into a situation where surrender is less costly than fighting. However, if you're precommitted to fight even irrationally (i.e. if the cost of fighting is higher than the prize defended), this makes such strategies ineffective, so the opponent won't even try them.

So for example, suppose you're negotiating the price you'll charge for some work, and given the straightforward cost-benefit calculations, it would be profitable for you to get anything over $10K, while it would be profitable for the other party to pay anything under $20K, so the possible deals are in that range. Now, if your potential client resolutely refuses to pay more than $11K, and if it's really impossible for you to get more, it is still rational for you to take that price rather than give up on the deal. However, if you are actually ready to accept this price given no other options, this gives the other party the incentive to insist with utter stubbornness that no higher price is possible. On the other hand, if you signal credibly that you'd respond to such a low offer by getting indignant that your work is valued so little and leaving angrily, then this strategy won't work, and you have improved your strategic position -- even though getting angry and leaving is irrational assuming that $11K really is the final offer.

(Clearly, the strategy goes both ways, and the buyer is also better off if he gets "irrationally" indignant at high prices that still leave him with a net plus. Real-life negotiations are complicated by countless other factors as well. Still, this is a practically relevant example of the basic principle.)

Now of course, an ideally rational agent with a perfect control of his external behavior would play the double game of signaling such precommitment convincingly but falsely and yielding if the bluff is called (or perhaps not if there would be consequences on his reputation). This however is normally impossible for humans, so you're better off with real precommitment that your emotional propensity to anger provides. Of course, if your emotional propensities are miscalibrated in any way, this can lead to strategic blunders instead of benefits -- and the quality of this calibration is a very significant part of what differentiates successful from unsuccessful people.