Comparison to your own past

People engaged in long improvement generally report feeling that they are not improving, while the same people would be startled by evidence of where they started.

Why progress is invisible

The reference point moves. Current ability is assessed against current standards, which rise as ability does.

Which means the gap between what you can do and what you want to do stays roughly constant, and the gap is what you perceive.

Memory reconstructs. Recollection of past ability is filtered through current knowledge, so previous work is remembered as better than it was, and previous difficulty as less than it was.

Improvement is gradual. Daily change is below the threshold of perception, and the aggregate is only visible across a long interval.

Attention goes to the errors. A more developed skill makes flaws more visible, so the experience of improving includes seeing more that is wrong.

That last effect is genuinely disorienting: increasing competence produces increasing dissatisfaction with your own work, which is easily misread as a lack of progress.

The remedy is records

Not memory, which is systematically wrong here.

Keep artefacts. Early drafts, early recordings, early photographs, old code, old work.

Looking at something from two years ago is the most direct evidence available and it is generally startling.

Keep numbers where they exist. Times, weights, output, whatever the domain offers.

Keep a note of what was difficult. The most useful record and the least kept.

Recording what you found hard at a particular point is more informative later than recording what you achieved, because the things that were once difficult have become invisible.

The review

Where the records get used, and without it they are just storage.

Quarterly is a reasonable interval — long enough for change to be visible, short enough to be actionable.

The questions worth asking: what can I do now that I could not do then? What was hard then that is routine now? What am I still stuck on?

That last one is the useful complement, because it identifies what has not moved, which is where the effort should go.

Why this beats social comparison

Comparison against others is confounded by everything you cannot see: their starting point, their resources, their time, their circumstances.

Comparison against your own past is confounded by nothing, because both data points are yours and complete.

It also has the property of being available regardless of circumstance. Someone in a difficult period can still have improved relative to themselves, which social comparison will not show.

The plateau

Worth addressing, because it is where records become most valuable.

Skill acquisition is generally not linear. Periods of rapid gain alternate with periods where nothing appears to change.

Some plateaus reflect consolidation — the underlying capability is reorganising in a way that produces no visible improvement until it does.

Others reflect a genuine ceiling on the current approach, where more of the same practice will not help and something must change.

Distinguishing them is difficult from inside. Records help by showing whether the plateau has lasted weeks or a year, which is the main evidence available.

The general principle from expertise research is that improvement requires practice at the edge of current ability with feedback, and that comfortable repetition of what you can already do produces no gain regardless of volume.

Which means a long plateau is frequently a sign that practice has become comfortable.

The honest caveat

Not all effort produces improvement, and the framing that progress is always occurring invisibly can be a way of avoiding an accurate assessment.

Sometimes the approach is wrong, the practice is unfocused, or the goal does not suit.

Records help here too, in the other direction: they distinguish a period of invisible progress from a period of no progress, and only one of those calls for patience.

Where it also misleads

The honest caveat, since this comparison is usually recommended without one.

Circumstances change. Comparing current output to a period when you had fewer responsibilities, better health or more support produces the same discouragement as comparing to someone else.

The useful version accounts for conditions: not whether you are doing more than before, but whether you are doing well given what this year actually contains.