Some video poker machines pay out better than 100% return on max bet. That is, if you sit there and stick money in it endlessly, you'll come out ahead, in the long run. If A then B. But problems come up with the "stick money in endlessly" and "long run" parts. Even though you're coming out ahead in the long run, you don't have enough money in your pocket to finance the long run. It amuses me to think that people still lose, and casinos still win, when paying out better than 100% return. The casino wins not by stacking odds against you, but by letting simple randomness eventually run you out of money, and by having a large enough bankroll themselves that the chances of them running out of money first are astronomically small.
Some online player's guides document this and give good numbers on how much it'll cost and how long it'll take. Another thing with potentially well over 100% return is progressive payout linked slots games, where a jackpot goes up as people play a bank of machines until someone finally wins it.
Reading a blurb (lost the link, sorry) about the development of the new MegaMan game for the Wii, a passage struck me, where they said that it took longer than they expected, even though the project was relatively technically unambitious, but luckily they were privately funded.
I've been railing against risk aversion lately. Other times, I rail against doomed software projects. There's no easy way to decide when to cut your loses and away from a software project or a video poker machine. In both cases, whether you should walk away depends on future events that are impossible to predict. Most funded software projects are failures. One survey put the number at around 90% being considered failures. That's probably about the same percentage of people who walk out of a casino down money.
I propose that these online video poker player's guides be used as a rough sort of manual for software development project management. Specifically:
Don't start a project you can't afford or aren't willing to bankroll, or else you'll wind up with nothing -- not the cash you spent or a finished product. Don't let your ego trick you into thinking you're a high roller when you're not. There's no shame in playing low stakes.
Time to complete the project, like time to hit a jackpot, is a function of probability and has a wide window. If you have a fixed window of time, you're in the wrong business.
Stick money into a machine does not translate directly into a payout, and putting labor onto a software project does not directly translate into a finished product. In the first case, you have expected payout. In the latter case, you have a basic efficiency factor. In the casino, you can usually find some book or site that tells you the expected payout, and video poker machines tell your their paytables (how much is paid for each hand). The randoms are not rigged, despite popular perception. In software, nothing tells you how efficient your effort is, but that doesn't mean that you shouldn't always seek to increase efficiency. And obviously, increasing efficiency doesn't mean cutting out bathroom breaks -- it means getting rid of barriers to getting work done.