Sanders, Clarke, Stewart and Whiteley (2011) simulate the 2010 election under the Alternative Vote (AV). Their results show large gains for the Liberal Democrats. I alter their methodology to make the estimation of voters first preferences under AV more accurate. With this change, the Liberal Democrats gains are considerably smaller, on the order of 10 seats or less. Bootstrap estimation shows that a Lib-Lab majority coalition would probably not have been feasible under AV.