Originally Posted by
Brodie Butland
Failing a set is a helpful lesson; intentionally failing a set is counterproductive.
It is not uncommon for lifters to bail from a set early, either by re-racking the bar before reps are done or by giving up on the lift once it gets hard. You get stronger--mentally and physically--by forcing yourself to try to accomplish something that you don't think you can do. If you don't permit yourself to fail in the short term, you will fail in the long term. Failing a set is a great teaching moment. It teaches you that failing won't kill or seriously injure you. It teaches you to get up, dust your pants off, and try again next time. It teaches you the importance of focusing on the process. It teaches you to never give up on a lift, no matter how hard it seems...and that, in turn, is the only way for you to learn what you truly are capable of.
That being said, failure must be kept in context. In a well-designed program, failure is a signal. At minimum, it means that you didn't accomplish what you were supposed to on a particular training day. This may be because you just had a bad day, as everyone does from time to time. But if you start failing frequently--and especially if you fail in consecutive training sessions--that signals that something systemic is wrong. Maybe it's your recovery (eating, sleeping, stress levels, etc.), maybe it's your rest periods (The First Three Questions), maybe it's a programming issue. But the point is, something is rotten in Denmark, and it's time for honest reflection to figure out what it might be and to fix the issue.
This is why training to failure makes no sense to me, even on a theoretical level. You're intentionally overstressing your body rather than letting it accomplish a pre-defined goal based on a specific programming choice. Empirically this does not work long term, and it is not even very effective in the short- or medium-term compared to well-engineered programming that does not rely on failure. Theoretically, you take what should be a warning signal and turn it into a deliberate part of the program, which a priori renders failure far less effective as a diagnostic tool. And then there's the safety issue...failure means you couldn't accomplish the lift, which often means there is a form breakdown somewhere. Why would you want to trigger a form breakdown if you don't have to?
tl;dr - Accidental failure happens, and that's okay...use it as the powerful learning experience that it is. Intentional failure is nuts.