A lot of it seems to be tricky math. It's not MPG, it's MPGe, the 'e' is for 'equivalent'. However, marketing guys have been playing those games forever, so no surprises there.<p>They're claiming E85 has 1/3rd less energy than gas, so they're actually getting 110*0.66 = 72.5 mpg of E85.<p>72.5 mpg is not as good for marketing hype, but it's still impressive.<p>He seems to be playing some games with the acceleration profile as well. If you 'gun it', you're in 'power mode', not 'efficiency mode', so the mpg numbers no longer apply. I'm not certain if this is common practice or not in these kinds of tests. It probably is.<p>I think he is mainly just taking control of the throttle away from the driver, and doing a "throttle by wire' system, giving you the most efficient (and probably painfully unresponsive) acceleration profile possible, when you're in 'efficiency mode'. I'm not sure if that buys him the whole 72.5mpg or not, but it's in his toolbox. <EDIT: Actually I can imagine such a system that wouldn't be too bad. When the user makes relatively small acceleration addition requests with the throttle pedal, you use your efficiency optimized acceleration profile to give it to him. When the user "stomps on it", because a truck is about to hit him, then you give him the requested throttle amount. I can imagine such a system would yield efficiency gains without being too painful to drive.</EDIT><p>Whether there's something else in the engine that gives you the rest of the efficiency gains, or if that's all of it, remains to be seen.<p>Of course, we don't know if any of this is real, but if it is, it's a cross between "real" and "marketing real".