In the type of experiment they were running, an almost gambling type situation with small amounts of money at stake and some element of random chance, I know that I personally wouldn't always follow the "correct" strategy just because if there isn't enough at stake it might be more fun to go for the "worse" option. Although if there were real high stakes then I don't know how I would behave. I think @hpoe's hypothesis that it has to do with signaling the environment has changed makes a lot of sense. It's hard to know what you would do in an experiment without being in it of course since you can rationalize anything, but I do know that in prior psych/econ studies I've been a subject in I've found it sometimes fun to deliberately gamble the riskier option when it's only a few pennies out of ten bucks for the session at stake.