If the question is whether or not there is a high chance of someone committing a violent crime caused by playing video games, then you are correct. If the question is whether there is any causation link between violent video games and committing real life violence, then you need more data than what you are using.

We have roughly 380 violent crimes per 100k people each year in the US. Just to keep it simple, let's assume they are committed by unique people. Half of adults play video games and about half of those play violent video games. (Very rough numbers but the right ballpark.) So in 100k people, you have 25,000 who play violent games of some sort. Of those violent video games, if only a certain type can create violence, you may be looking at only 10,000 people or fewer (per 100k people) who are playing the games that could cause violence. It also seems likely that the ability to cause violence would be tied to genetic and environmental issues in the person playing the video games. On top of that, for the violence to manifest itself, the person would need to find themselves in a circumstance where it could manifest. (Being bullied at school, etc.)

It seems very likely that by the time everything that might have an impact on violence is considered, you might be only dealing with 1,000 people out of every 100k who would be in the at risk group. If the effect for the at risk group is very high, say 1 out of 1,000, you are talking about one more violent crime per 100,000 people.

It would seem that an increase like this could easily be hidden due to other programs to reduce crime like cleaning up lead pollution, or better training teachers to to identify at risk students.

(And please understand, I'm not saying that there is a link. Just pointing out how it would be likely to show up if there was.)