That time range "conveniently" includes the 2011 Breivik massacre, which skews the result mightily. Mass shootings are sufficiently rare, particularly in Europe, that single events have an outsize effect on the stats.
Also, the gun homicide rate is around 10x that of other rich industrial nations. If you think that there's a substitution effect you're right: if all homicides regardless of weapon are taken into account the ratio drops to around 4-5x.
There were 27% more casualties per capita from mass public shootings in EU than US from 2009-15. That's across the entire EU. So the low population skew effect is removed at that scale.
If you'd like, we could compare European Nations to individual US states. But then if you get to discount Breveik in Norway, Connecticut gets to discount Sandy Hook.
According to most other stats, the US has more "school shootings" than most of the rest of the world combined, see for example https://qz.com/37015/how-school-killings-in-the-us-stack-up-...
Also, the gun homicide rate is around 10x that of other rich industrial nations. If you think that there's a substitution effect you're right: if all homicides regardless of weapon are taken into account the ratio drops to around 4-5x.