That's what Help prioritizing games with alternative qdisc design is all about. But it's a thread about experimentation and development of a method. If you just want to know what's the answer, you'll have to check back once we've found it 
Mean time, if you go to the github https://github.com/dlakelan/routerperf it has instructions how to try it out in the README, so if you try it and want to jump into the main thread and provide
we would be happy to see your experimental data.
In particular I'm sure @moeller0 would like to see systematic comparisons between SQM and the script we develop in that thread.
Also, unfortunately while we can show the objective reality of how fast the packets go through... it turns out that isn't necessarily the same thing as smooth game play. People find better more playable games with moderate controlled lag and jitter for some games