But if it's just an average, and it's ultimately, probabilistic in the end, then the black object occasionally and momentarily becoming red, isn't any true violation. At best you can call it a statistical impossibility, which is not at true impossibility. It may only be impossible in the same sense that throwing dice and landing snake eyes a googolplex times in a row is impossible. In all practicality, it's a statistical impossibility, which really means it's only highly improbable. The only true impossibilities are the physically impossible, logically impossible, and mathematically impossible. But this is only mathematically improbable, which isn't a physical impossibility.
Also with the case of of your hypothetical mass with the red and black particles, maybe the perpetual pitch-blackness only seems that way to the naked eye. But let's say it's being observed with highly precised instruments that chart it's redness and blackness on a line graph. It may chart the most minute spikes in the red spectrum that we couldn't possibly see with the naked eye, but the fact that there's spikes at all on the chart would suggest that things aren't completely determinate. And of course, it would show spikes on the graph in the black "spectrum" (I know there's no black spectrum, but you get what I mean lol) that shows it dipping below its average blackness.
If the machine is precise enough, then there's no way the ratio is completely uniform throughout time. Because that would suggest that the amount of particles being red and black at any given time, is always that exact value all the time. e.g. there's always exactly 33,728,388,026 red particles out of a total of 100 trillion particles. There's never 33,728,388,027 or 33,728,388,025. I think what you're saying though is that you're fine with those minute fluctuations, but that a mass going from having roughly 33 billion red particles to suddenly having nearly 100 trillion particles is far too unlikely to ever happen. But maybe that doesn't have to happen. The universe is known to have chaotic systems. Meaning they are systems that are sensitive to minute changes, such that when the changes happen, the system will have a dramatic behavior.
In other words, maybe a single particle turning red inside of a person is enough of a fluctuation for the person to decide take on a path of professional race car driving instead of a life of a chief. Okay maybe it's not that chaotic. Maybe it's enough for the person to decide to race cars instead of motorcycles, and maybe it's a thousand red particles instead of one.