Actually, that was a wrong statement & idek how i posted that. I was initially going to say that but the research I heard about spoke about psychometric testing and not those games. However those games kinda count as that.
Games based assessments do not "kinda" count as that. They are very different assessments designed in very different ways.
For clarity on this, some ability tests there is known potential adverse impact (verbal reasoning, WG, numerical reasoning, abstract reasoning) - but this is not seen in games based assessments.
Basically, there is a best practice rule called the 4/5th rule which monitors adverse impact. With ability tests, like the ones mentioned above, there is evidence to show there can be adverse impact
IF you set a very high percentile benchmark (ie the 4/5th rule is not met). This is typically against women for abstract and numerical reasoning, and against BAME applicants in verbal reasoning and assessments like the WG.
However, firms are acutely aware of this and part of the reason their percentile benchmarks are probably much lower than you expect is to minimise or eliminate adverse impact.
I literally could bore the hind legs off a donkey about these type of assessments. Happy to explain further if people want though....