I buy lots of Indies. Basically, if a game is available for 1.50 € or less at GamersGate, and sounds at least somewhat interesting, I buy it. I also buy some games at Indievania, but only if they aren't available on GG and look really good (like Climb To The Top Of The Castle), or if they are much cheaper (I got the Milkstone bundle at Indievania for 10% of what I would have had to pay at GG). I don't buy Indie games elsewhere, especially not at Steam or Desura since I don't like their being client-based.
Regarding the original question: I tend to give most Indies a whirl. I'd say I install 80% of the Indies I buy, but only 30% of the non-Indie games - Indies have the advantage here because they are usually small, quick and easy to install, no need to worry about limited activations or similar crap - I just start the installer, and usually I can try the game about two minutes later. I then fiddle around with it for something between 30 minutes and 3 hours. Basically, the game needs to impress me during that timespan or I may never start it again (because there are so many other games to try).
This means the game should be easy to get into (not necessarily simple), and offer some gratifying moments or "Wow!" experiences right away. It's a bonus if the game hints at more greatness down the road.
Indies that I recently spent more than a day with include:
Fortix 2: Very easy to get into, difficulty level just right, great way of introducing more gameplay elements every other level. I played this game a lot during two very stressful weeks.
Osmos: Interesting idea, unique atmosphere and gameplay. I was intrigued right from the start. Good presentation (graphics as well as music) - not terribly elaborate, but well done. There's not much _beyond_ the novelty though, so my fascination only lasted a couple of days.
Fate of the World: Interesting concept, challenging tasks. The game undoubtedly has its flaws (the UI is horribly inefficient, and the gameplay is rather limited), but the setting is interesting enough for me to keep returning to the game from time to time. The distinctively different missions also keep the game fresh.
Braid: I ignored that game for a long time because (a) I dodn't like the art style, and (b) I just couldn't understand what the reviewers were raving about. I thought: It's a jump'n'run that allows you to spin time backwards to escape death, so what? Then I actually _played_ it and got blown away. I would have never imagined how original, thoughtful and challenging the "time manipulation" concept was worked into puzzles. It was an amazing experience to feel how I somehow intuitively solved many puzzles even before I analyzed them and "deduced" a solution. The game gets pretty hard after a while though, and I'm not sure if I'll pick it up again. But it was truly great.
Trine: Strikingly beautiful graphics, original (or at least rarely used) idea, also pretty easy. Actually a bit _too_ easy and repetitive to pick it up again after finishing it once, but it was fun to get there.
MotorHEAT: Simple, yet challenging racing game with good graphics and music. The achievements and the global hiscore table provide additional motivation.
Dwarf Fortress: Extremely deep complex game - so much so that it makes me put up with one of the worst UIs ever designed for a game. About once per year, I sink two to three weeks into this game.
Tomes of Maj'Eyal: I play lots of roguelikes, but this one stood out lately, since it's in active development, provides a lot of variation, and tells an epic story.
I also sunk a considerable amount of time into Mount & Blade, spend three days with Penumbra (but I'm not sure if I'll pick it up again, the dogs become frustrating quickly), and tried lots of other game which didn't impress me enough to mention them here ...
If you want some more details, ask away. :)
Post edited March 08, 2012 by Psyringe