Playing in an online game and noticed some interesting stats. Going through 200 hands, there's several hundred more but feel this is sufficient.

Playing 6 handed and two players not once made it to the river together. And they only made it to the turn together 5 times. Those turn pots are relatively small (max 25bb) for how deep the game plays where average stack is 250bb+.

They often get online and offline within minutes of each other if not seconds as well.

Is this strong enough proof to argue they are colluding?

No. A few hundred hands may seem like a lot, but it is in fact a fairly small sample. I cannot definitively say they are not colluding, but what you described is not really evidence that they are.

I personally have had stretches where I am a bit card dead where I have only gone to the river a small handful of times over that size sample. I would not be all that surprised that another player likewise only went to the river a few times during the same stretch. If player A only gets to the river 5% of the time and likewise with player B, it would not be particularly surprising if they never went to the same river over a few hundred hands. (Plus not all hands get to the river, so that only makes it less surprising).

Actually makes it very slightly less likely they are colluding(small sample size). Better to have your partner in the pot so you can control the action better.

200 hands is a very small sample size. And if they are colluding as mentioned above they would probably be involved more with each rather than just trying to avoid each other actually.

Too small of a sample