You all know the concept of Russian roulette. Take a revolver with six chambers, put a bullet in one chamber and leave the rest empty. Spin the chambers. There's a one and six chance that if you put the gun to your head and pull the trigger, you'll die. Would you do it? Almost certainly no. Indeed, we'd say that it's immoral to engage in such a risky activity for no reason.It strikes me that the problem with this line of argument is that it splits off the question of risk from what you are actually doing, and thus acts as if when you choose to do some thing which involves risk, you're simply playing Russian roulette.
However, we all engage in risky activities all the time. When you go swimming, you risk drowning. When you go cycling, you risk having a crash (or being hit) and dying. When you drive to work you risk dying in a crash. The list goes on. Even doing nothing carries a risk since a sedentary lifestyle increases your risk of various health problems. And yet everything you could do instead of sitting on the couch carries risk as well.
So, how many chambers does the revolver have to have for you to be willing to play Russian roulette? At what odds does engaging in a dangerous activity just for fun become moral? At what odds does engaging in a dangerous activity for a practical reason become moral? Is there a difference between these two?
Googling around for an example, I find a pair of statistics that you have a 1/2,000,000 chance of a fatal accident for each ride you take on a bike and a 1/2,800,000 chance of a fatal accident for each drive you take I a car. These are activities we feel pretty good about. Would it be moral to play Russian roulette with a two million chamber gun? How about a one million chamber gun? A one thousand chamber gun?
When I hop in the car and drive, I'm not doing so with the object of facing down a 1/2,800,000 risk of death, I'm doing so to pick up groceries at the store, or go to work, or visit family. My object is to perform that action. The means by which I perform that action (driving) carries a certain risk, but I consider the risk to be acceptable. If the risk was much different (say, if there were a 1/100 chance of dying every time you drove) I might decide that some of those things weren't worth the risk.
The problem with comparing this to Russian roulette is that there is not, so far as I can tell, any particular purpose to Russian roulette other than to face down a 1 in 6 risk of dying. In this sense, it doesn't even compare well to highly risky activities, because there isn't really any activity involved in Russian roulette other than risking death.
Say that your favorite activity is a really dangerous extreme sport. Let's say your chance of dying each time you participate in this sport is 1/5,000. Is participating in the sport the same as playing Russian roulette with a 5,000 chamber gun? Unless you're participating in the sport solely with the intention of experiencing a 1/5,000 chance of death, I would assume not. You participate in the sport because you enjoy the activity itself.
I'd argue that it's never moral to play Russian roulette no matter how good your odds. Risking your life as an end unto itself is wrong regardless of the degree of the risk. However, that doesn't tell us anything about the morality of doing some particular thing, with a purpose other than risking ones life, which nonetheless carries with it a certain predictable risk to one's life as a foreseen side effect. There it's necessary to make some sort of judgment balancing the purpose one has with the risk one incurs -- and weighting that risk, in turn, against the likely effects of one's death or injury on others as well as on oneself.