Pitch ratings are ranked opinions that get treated as measurements

An official scale running from good to unfit is ordinal data, and everyone handles it as though the gaps between grades were equal.

Good, average, below average, poor, unfit. Five words in a row.

Assign them one to five and you have created a number that supports arithmetic it was never built for. That is the whole problem with match official pitch ratings, and it runs through every league table of venue quality anyone has ever compiled. The scale is ordinal. It tells you the order of the categories and nothing about the distance between them. The step from good to average is not obviously the same size as the step from poor to unfit, and there is no reason it should be.

Yet the moment those grades enter a spreadsheet, people average them. A venue with two goods and two poors comes out level with a venue rated average four times. Those are not the same venue and no amount of decimal places will make them equivalent.

Then there is who is doing the rating. One official, watching one match, assigning a grade after an outcome he has already seen. That last part matters more than anything else here. A low scoring match makes a surface look worse, and referees are human, so the rating is partly a function of the result rather than of the surface. You cannot untangle the two after the fact because the grade was recorded once, at the end, by somebody who watched the whole thing.

Consistency across officials is the other unknown. Different referees apply the descriptors differently, and rotations are not random, so a venue's record partly reflects which officials happened to be assigned to it.

If I wanted a defensible system, I would keep the descriptive grades for sanction purposes and separate them completely from any comparative measure of venues. For comparison I would use physical readings and match outcome distributions, both of which are recorded automatically and neither of which requires a person to convert a judgement into a rank.

And I would report grades as counts, never as means. Two goods and two poors, written exactly like that.

The pattern here is a general one and it is worth naming. Ordinal categories, averaged, then ranked, then used to make decisions with financial consequences. Sport does this constantly with survey scales, coaching assessments and referee grades.

Words got turned into numbers for filing convenience, and everyone downstream forgot that a conversion took place.