Era adjustment: what it fixes and what it cannot

What changes between periods

Equipment, playing surfaces, preparation methods, tactical understanding and the size of the population competing all shift across decades in different directions. Each of those changes affects the raw numbers, and they do not move together, so a single correction factor cannot capture the combined effect. Rule changes complicate matters further, because they can alter what counts as a scoring event or how often a particular situation arises at all.

The result is that a raw figure from a distant period and a modern one are not measurements of the same underlying quantity. Any comparison therefore requires translation, and translation always loses something regardless of how carefully it is performed.

Standardising against contemporaries

The most defensible method expresses a performance relative to the distribution of performances in the same period rather than in absolute terms. This automatically absorbs changes in scoring environment, since a shift affecting everybody equally moves the whole distribution and not the relative position. It converts the question from how much somebody produced into how far ahead of their contemporaries they stood, which is a more stable idea.

The method assumes the distribution is comparable across periods, which fails when the population competing has changed in size or in composition. A wide margin over a shallow field is not equivalent to a narrower margin over a deep one, and the standardised figure cannot see the difference.

Depth and the meaning of a percentile

When participation broadens, the standard required to be average rises, and the distance between the best and the typical competitor tends to narrow. A competitor who dominated a small field may therefore post a more extreme relative figure than a later competitor facing far stronger opposition. Adjusting for depth requires estimating how good the field was, which cannot be done from within the same field without circularity.

External anchors such as measured performance in comparable conditions help in some sports and are unavailable in most team disciplines. The honest position is that depth adjustments are educated guesses whose uncertainty is larger than the differences they are asked to resolve.

Untranslatable differences

Some changes are not quantitative at all, since a role may have existed in one period and simply not existed in another. Where the demands of a position have been redefined, asking how a competitor would have performed in the other era is a question about adaptation rather than ability. Adaptation is unobservable, and confident claims about it are assertions dressed as analysis regardless of how much data accompanies them.

The same applies to preparation, since a competitor from an earlier period would have trained under later methods had they been available. That thought experiment is reasonable and unfalsifiable, which places it outside what any ranking can legitimately settle.

Presenting cross-era comparisons responsibly

The most useful format is to rank within periods and then discuss the relationship between periods separately and explicitly. This gives readers a defensible order in each era and an honest account of why the orders cannot simply be concatenated into one. Where a combined list is required, showing the adjustment applied and the uncertainty attached to it is the minimum reasonable standard.

Small gaps in a cross-era list carry no information at all, and presenting them as ordered positions overstates the evidence considerably. Grouping rather than ordering communicates the state of knowledge accurately, which is the only durable basis for this kind of comparison.