I watched the rollover piece last night on MSNBC, and found it very interesting. They talked about two lists of rollover probability, one based on a mathematical formula based on the design of the vehicle, and one based on historic accident rates. Neither are public knowledge, and the lobbiests for the auto manufactures are working to keep this data hidden.

Based on the reference information on this table, this is data based on the mathematical formula. And this is not likely based on reality, which was acknowledged on the MSNBC program.

Look at the data. I recently bought a new Corvette, which is the best handling stock automobile for sale in the US, able to create lateral g force in excess of 1g. It also has very advance traction and yaw control. I've never felt more in control of an automobile. Yet the Vette is beat out in this study by a Crown Vic with soft suspension. Why? Maybe it's because the Vette is capable of 175 mph, and the Crown Vic is probably limited to 90 mph. Yet I believe that if applied to the exact same scenario, the Vette would be less likely to roll.

Volvos of late have become increasingly performance oriented. Without knowing the specific algorithm, I would guess that they took performance orientation into account, and penalized those cars because they're more likely to be driven closer to the edge.

I don't think Volvos are very likely to roll at all, and are less likely to roll than many other vehicles, including John O's Regal!

(BTW, I have a Suburban too, and was glad to see that it rated higher than other SUVs)