One idea, badly explained

Progressive overload is the rule that muscle adapts to a demand it has met, and stops adapting once that demand stops rising. That is the whole of it. Everything else — periodisation, deloads, block programming — is bookkeeping on top of that one sentence.

It is also the part of training most likely to be skipped, because it asks for something people are bad at: remembering, accurately, what they did three weeks ago. Not roughly. Exactly. Which weight, how many reps, whether the last set moved or stalled.

Why memory is the bottleneck

Ask anyone mid-session what they benched a month ago and you will get a number that is confidently wrong, usually in the flattering direction. That is not a character flaw; it is how memory works. The consequence is that most people run the same weight for far longer than they meant to, then jump too far, miss, and conclude they have plateaued.

A log fixes half of that: it tells you what happened. What it does not do, on its own, is tell you what to do next — which is the half that decides whether the next twelve weeks go anywhere.

What GOPUSH does with it

Every exercise in a session carries a target worked out from your own history with that exercise: the loads you actually used, the reps you actually completed, and whether the last attempt moved. It is derived, not stored — so it cannot drift out of date, and correcting a set corrects the suggestion with it.

The rules are deliberately unexciting. Clear the top of the rep range on every working set and the load goes up by one step. Miss the bottom and it holds. There is no motivational language attached to either outcome, because a missed set is information, not a verdict.

What it will not do

It will not invent a target for an exercise it has never seen you do, and it will not guess a rep range you never set. A suggestion built on nothing is worse than no suggestion: it looks identical to one built on a year of training, and you cannot tell them apart at the moment you are deciding what to load.