Pricing is platform dependant. When your competetors are $0.99, you need to have a very superior game to charge $4.99 on an iPhone. But on the other hand, if you price at $4.99 on Steam you are sending a signal to potential buyers about the quality of the game you are offering... a signal that your game is very short/low on content or quality. Different channels have different pricing expectations.
In general you should price your game at a spot that you feel is fair, but offer a variety of promotions and offers to increase visibility of the game. Valve has discussed this on their blog, that developers oftenalways make more money when they sell their game at 75% off than they do at full price, because promotions get people interested and raise visibility. But that's not the same as PRICING your game permanently at the 75% off price point... people love deals, and when your game is priced at the lower point it's not a deal, it's just a 'cheap' game.
So find similar games to yours in the channel where you are selling and match their prices, but also offer frequent sales and promotions as this increases visibility and allows you to reach price points that raise your potential audience as well as increasing word-of-mouth about the game.
Finally, here is an interesting point from Gabe about game pricing: Now we did something where we decided to look at price elasticity. Without making announcements, we varied the price of one of our products. We have Steam so we can watch user behavior in real time. That gives us a useful tool for making experiments which you can’t really do through a lot of other distribution mechanisms. What we saw was that pricing was perfectly elastic. In other words, our gross revenue would remain constant. We thought, hooray, we understand this really well. There’s no way to use price to increase or decrease the size of your business.
So take what anyone says here with a grain of salt, because it's quite possible we're completely wrong. Your price point may really do almost nothing to your revenue, good or bad. Real world experiments trump any quantity of theory.