Hmm, here's a thought. Actually, here's a few of them:
-Most games get returned shortly after they go on sale because of the lack of content. Increase the singleplayer content from a 9-hour ordeal to something that can last a bit longer, or give players an incentive to actually play the game again--don't just sell us a nine-hour long hallway filled with cutscenes.And treat multiplayer as a way to pad the singleplayer, not the other way around.
-Find a way to cut development costs. Games do not need the most realistic graphics to sell well, so license an engine that's a little behind rather than the cutting-edge polygon renderer. Keep team sizes small until you absolutely need to go bigger, and then reduce as you need to. Get a solid prototype working early, and maintain a strong design focus to reduce the amount of work you'll eventually waste on the game. Sell game for a fraction of the cost.
-Make people actually want to keep the game. Use expansion packs to greatly increase the game's content (not just "lol weapon reskinz!"), create an engrossing multiplayer mode, something. Make them want to keep the game for some reason, even if it's only for the next quarter.
Or, alternatively,
-Go an all digital route. Release your game to Steam, Xbox Live, PSN, WiiWare, something. Cut out the middleman, and let your game prices actually reflect that.
Evolve or die, console developers. Treat your customers like customers instead of wallets.