There are a few things to consider for a good aerobatic airfoil. The transition between attached and separated airflow has a big influence for spins or snap rolls, and a good aerobatic airfoil needs only a small change in AoA (angle of attack) to fully transition between the two. To achieve this, it helps to shape the forward part similar to an ellipse (a lemniscatic function, to be precise), and to use a straight line for the last 70% to 80% of chord. The goal is to have a Stratford pressure rise on those 70% to 80% which gives the maximum possible pressure gradient without separation. Increase AoA just a little, and you have fully separated flow over three quarters of the wing chord.
When all works as planned, you can get the plane out of a spin by applying rudder at, say, 45° before the fuselage points in the desired direction and get out of the spin within 10° of that direction. Same goes for snap rolls, they get really snappy. It is a real pleasure to fly such planes, but you need some discipline. There are no aerodynamic warnings before you stall, and stalling comes immediately. On the other side, with the limited lift loss of such an airfoil and the powerful engines of aerobatic airplanes, stalling is no big issue.
The next thing to consider is relative thickness. Modern aerobatic aircraft are rated ±10g (note that FAR/JAR only requires ±6g in the aerobatic category), so a thick root airfoil helps to keep the structural mass down. On the other hand, maximum lift is possible with a thickness around 12%, so the compromise is to taper the wing from 15% at the root to 12% at the tips. The airfoil numbers given by mins refer to the relative thickness as percent of chord.
None of the airfoils in your collection (which looks awfully like an older version of Michael S. Selig's UUIC database) will fit well. Go ahead and design one with XFOIL and use the calculated polar.