I examine the incentives of professional mixed martial artists and how they influence the strategies of the fighters in high profile competitions. Using data collected from hundreds of professional fights, I examine the effects of 1st and 2nd round performance together on 3rd round performance and on risky behavior, in bouts lasting three rounds. Using probability logistic regressions to calculate performance, and linear regressions to estimate risk taking, we find that 1st and 2nd round performance together have little effect on third round performance. However, a good (bad) performance in the first 2 rounds reduces (increases) risk-taking behavior in the third round or changes in risk-taking behavior in the 3rd round, relative to the first two rounds. We also examine the rare cases of rematches, and find that while performance in the original fight is a good predictor of the outcome and performance in a rematch, the actual outcome of the original fight is not. That is to say that fighters who did well, but lost (by bad luck) in the original match are more likely to win the second match.