Format fragmentation and why one cricket ranking cannot hold

One sport, several optimisation problems

The constraints that define each format change what a competitor should be trying to maximise, and that changes which abilities are valuable. Where the resource limiting a side is time, preserving wickets matters most, and the correct approach is patient accumulation with low risk. Where the resource is a fixed number of deliveries, the scarce commodity is scoring opportunity, and controlled risk becomes the rational strategy.

A competitor optimised for one problem is not automatically suited to the other, because the two reward opposite responses to the same situation. Describing them as the same sport is true historically and misleading analytically. The shared equipment and vocabulary disguise the fact that success in each version is defined against a completely different scarce resource.

Why a combined rating misleads

Averaging performance across formats produces a number that corresponds to no actual competition and describes no actual role. A specialist excellent in one format and unused in another receives a diluted figure that understates their value in the setting where they play. A competitor active in all formats is rewarded for availability and versatility, which are real virtues but different ones from peak quality.

Because the mix of formats played varies enormously between competitors, the combined figure is partly a measure of selection policy. That makes it a poor basis for any comparison, since two figures may be composed of quite different ingredients.

Separate ratings and their relationship

The workable approach is to compute a rating for each format and to report them alongside each other without collapsing them. Readers can then apply their own weighting, which is appropriate because the correct weighting depends entirely on the question being asked. It is also useful to report how strongly the ratings correlate across formats, since that correlation is itself an interesting property of the sport.

A high correlation would suggest the underlying skills transfer well, while a low one confirms that specialisation is genuine rather than fashionable. Publishing the relationship gives readers something the combined number destroys.

Selection feedback and its distortions

Ratings influence selection and selection determines who accumulates further evidence, which creates a loop that reinforces existing assessments. A competitor rated highly receives more opportunities, and more opportunities produce a more stable rating regardless of the underlying level. A competitor rated modestly may be moved to a format where they are considered a better fit, which removes them from the sample entirely.

The observed distribution of ratings therefore reflects selection decisions as much as it reflects ability across the whole eligible population. That loop cannot be broken by better arithmetic, and acknowledging it is the only honest response available.

Comparing across formats when you must

Sometimes a comparison across formats is unavoidable, for instance when a single award is offered for a whole year of competition. The defensible approach is to state the weighting in advance, apply it identically and publish the component figures alongside the total. A weighting chosen after the candidates are known will always be suspected of being reverse-engineered, and often it will have been.

Publishing components allows disagreement to focus on the weighting rather than on the outcome, which is a much more productive argument. It also lets a reader recompute the result under their own assumptions, which is the strongest possible form of transparency.