The dates given by AOS for past events may actually disprove evolution entirely

[pocaracas]

The error bars come from the dating methods.
The layers of ground then show that the ones below are older than the ones above, provided there's no evidence of any geological activity which may suggest that those layers have been disturbed.

So, for the provided layers:

Tuff 8: 3.46 +- 0.12 Ma
Tuff 7: 3.56 +- 0.2 Ma

If, you, somehow, come to the conclusion that Tuff 8's age is closer to 3.46 + 0.12 = 3.58Ma, then Tuff 7 must be, at least, as old, 3.58Ma is included in the error for Tuff 7, so it's ok.

But there is a possibility, which is not statistically insignificant, that the layers are out of order. Furthermore, the actual error ranges do not include an error analysis. Nor do the dates show what measurement technique was used.

There is a method called isochron dating. It is supposed to eliminate errors due to initial conditions. And the isochron method can be also done with different isotopes.

So what method was used to date these layers and what would be the results of other measurement techniques.

I don't know the method used. If you want to know, I remember I linked the whole paper, so you can read it there, or just follow the references at the end.