Would the many-worlds interpretation (MWI) of quantum mechanics [1] mean we should do anything differently? Say a pair of strangers set up the following quantum suicide lottery:

Each puts $N into a common pot and then set up a machine that will on the basis of a random quantum event instantly kill one or the other of them. The one that survives gets the pot and is $N richer.

Under the MWI an outside observer will see one die at random, but each participant will experience winning and not losing. This differs from the Copenhagen interpretation where the world does not branch and only one of the people will experience winning this lottery.

Assuming that the MWI is correct, this is still faulty reasoning. It is based on the idea that only worlds you are around to experience count. Consider an alternate lottery, completely the same except the machine is set not to kill. Someone still wins the money, but the loser remains alive. Whether you would prefer this lottery over the other should depend almost entirely on whether, should you happen to lose the money, you would rather be dead.

(This is the same confusion which leads people to average utilitarianism instead of total.)

[1] I don't know enough about the theory to know whether it's true, but I'm interested in the consequences.