Another reason the estimates are so wrong as that we engineers will always estimate conservatively. We want it to perform better than we promise it is going to. Then managers come along, and lop 10% off of our numbers, and then some big systems guy comes along, and puts everybody's numbers together, and takes something off the top of that, etc.<br /><br />In short, the rovers were always capable of more than their "advertised" abilties. After all, if NASA had claimed they'd be available for a year, and they died in six months, then we'd have a mission failure! If NASA claimed they'd be around for 3 months, and they died in six months, then we'd have mission success!<br /><br />Having said that, however, I think we've also had some very good luck with both of them, such as solar panel cleaning events, and just general good health. Add to that the excellent people working at control of these rovers, and you've got a good formula. Remember that the rovers have not been without problems -- we had the system shutdown when they filled up too much of the Flash memory, we've had a balky wheel on Spirit, a power leak on Opportunity, instruments getting older, etc. But the very flexible architecture designed in by engineers and the problem solving abilities of the controllers have been able to overcome all of those problems.<br /><br />All hail to the engineers!<br />