Two numbers sit next to each other in the same small Police UK extract for West Yorkshire, and they don't sit comfortably. The first is 66: the number of records classified as violent crime, out of 161 in total. That's 41 per cent, the single largest category by a distance. The second is 24: the number filed under a category called "Other Crime". (Source: Police UK, crime-west-yorkshire)

One of those numbers gets argued about constantly. The other is the second most common thing in the file, and it has no meaning at all to anyone reading it from outside.

Think about what that does to a conversation about crime. Violence dominates the extract, and that matters. But the runner-up isn't burglary, isn't shoplifting, isn't criminal damage. It's a residual bucket. Fifteen per cent of everything recorded here is, functionally, unlabelled. You cannot campaign on it, budget for it, or hold anyone to account for it, because you don't know what it is.

Below those two lines, the picture gets specific again. Drugs offences account for 18 records and public order 17, close enough that they're effectively level. "Other Theft" contributes 11. That last label deserves care: it's a narrow sub-category in the Police UK taxonomy, not a count of all theft. Burglary, vehicle crime and shoplifting are separate categories in the national classification and aren't broken out in this extract, so 11 tells you nothing about how much property crime West Yorkshire is dealing with.

Which leads to the third number worth noticing. The five categories listed here add up to 136. The total is 161. So 25 records, roughly one in six, sit in categories that didn't make the top five and aren't shown. They exist. They're counted. They're just invisible in the summary anyone is likely to actually read.

My view, and it is a view: this is why crime arguments go nowhere. The headline figure is a broad label, "Violent Crime", covering everything from harassment through to the most serious offences with no severity split visible. The second-biggest figure is definitionally a leftover. And a sixth of the file is off-screen. You can pick almost any story you like out of that and never be provably wrong.

The honest counter-argument, and it's a decent one. This is a snapshot of 161 records, not an annual force-wide total, so proportions will swing hard month to month and area to area. Residual categories are a normal, defensible feature of any classification system: you need somewhere to put the things that don't fit, and "Other" is what that looks like on a chart. The full Police UK dataset does carry finer detail than this extract shows, and a 24-record bucket in a file this small is not evidence of a systemic problem.

Still, the summary is what travels. And in this one, the number two crime in West Yorkshire is a shrug. If you want a serious public argument about violence, and 41 per cent of this file says you should, start by insisting the categories mean something.