[1] https://onlinelibrary.wiley.com/doi/abs/10.1111/1745-9133.12... [2] https://www.fbi.gov/contact-us/field-offices/dallas/news/pre...
But here in Europe cops rarely even draw their guns (and in some countries don't even carry them)
That job Y is higher risk than job X has no relevance to whether or not steps should be taken to mitigate risks on job X, unless either for some reason mitigating risks on job X would prevent mitigating risks on job Y and so those resources would be more effective going toward mitigating Y's risks, or you want to argue that there is some threshold and people in jobs whose risks are below that threshold should not try to mitigate those risks.
When a given fact supports both sides equally well in an argument that is pretty much the definition of an irrelevant fact.
Crossing guard death rates on the job appear to be around 19 per 100k. For police in the US the rate varies quite a bit year to year, ranging from around 14 per 100k to around double that. The average depends on over what time period you look, so it is not clear if it is actually higher or lower than that of crossing guards, but whichever it is it is not very far from that of crossing guards.
But that is the rate for police with guns. The comparison you want if you are arguing against police guns is the rate without guns.
Using the rate with guns to argue against guns is like arguing that golf tournaments should not shut down when a lightning storm passes over because the rate of people being killed at golf tournaments that shut down during lightning storms is so low.