When Valve moved the game to F2P, the game was out for almost 4 years. By that point, basically everyone that was ever going to buy the game had purchased it. Plus, Valve already had the loot chase built into the game(random drops were added in 2009, IIRC). And while it's popularity wasn't waning, it was still usually the third place game on Steam's total users list behind CS:Source and CS 1.3.
So they basically had full market saturation for their product, and probably weren't seeing much new revenue even when they did major content updates. By making the game free, they opened the door to a LOT more people, and allowed them to get a lot more revenue from both new and long time players. I've spent probably somewhere around $40 in hats and weapons since the conversion.
The truly brilliant thing that Valve did was make the transition absolutely seamless for the veteran players. Outside of a few new buttons on the main menu and a very unobtrusive news section that advertised changes, the overall experience didn't change at all for them.
I am a little miffed that they never upgraded the console versions and I've never really played it on PC. But I'm glad they're still able to get life out of the game, considering they've been upgrading it (the PC/Mac versions) well past its shelf life. It's an inspiring example for other developers with games that aren't commercially viable yet still have a well-knit community.
The thing is, all these big ideas (Spy, Engineer) who are implemented are balanced and working. This is just possible because it is not a dead line driven development.
Most of the game industry is still working with hype and big releases, but the most successful ones are incrementally developed. World of Warcraft and Minecraft for example. Community feedback, frequently new content and constant balancing/bug fixing creates the best games.
TF2 showed this is not just possible for MMORPG's, the best genre games are yet to come.
Tribes is also free to play and I'm curious to see how things evolve compared to TF2. Currently there aren't any things such as 'hats'. In fact, you spend your experience to just unlock the basic classes. I doubt that strategy would have worked well with TF2 since a lot of their balancing comes from the various classes.
The massive increase in poorly balanced weapons and items and the flood of newbie players who were not that interested in the 'team' aspect of 'team fortress' really killed the fun for me.
There was a similar effect with major releases before it became free to play. I particularly remember the engineer update - for about a week that's the only class 1/2 the team would want to play, so good luck being Blue (attacking) on a pipeline map (for example).
I'll try it again one day soon so see if things have settled down, but I think that the commercial success came at the cost of game quality.
Personally I liked TF2 when it first came out (before achievement based upgrades came out) and felt that the updates and new weapons and especially after hats and such came out were all too much. My problem is probably that I was a hardcore counter-strike player and TF2 really turned into a pub oriented game.
It is still fun to play a week every few months if you can find a decent server to play in.
Probably this is the habits I learned on mediocre pubbie servers coming back to bite me, and I need to learn to be less aggressive all the time, but it still rankles.
The question is, how will they balance a hardcore gaming community which hates paying to get better (stats-wise) items vs monetizing F2P? Will they charge only for cosmetic items, or will they drive away their core Dota fanbase?
Let's consider:
* Replay system
* Training system
* Coaching System
* Microtransactions
* Trading interface
* Community Submissions + submission royalties
Who knows what else? At this point, I think tf2's success is going to influence DOTA2, and I wouldn't be surprised at all if DOTA2 is released free to play.
I probably won't ever play DOTA2, but I'm thankful for all the improvements to tf2 over the last few years. I'm just hoping for some mainly bug fix patches to be released soon as this point.
The Degreaser flamethrower in TF2, for example, is a non-stock item that lets you switch weapons faster, but also reduces the burn duration of enemies you set on fire.
Hell, you don't even need to be trying to monetize to screw this up. See: Modern Warfare multiplayer and its propensity for handing strictly better weapons to veteran players, resulting in even more lopsidedness in matches.
edit: If in game, player driven markets are the sort of thing that interests you: keep an eye out for Diablo 3's real money auction house.
Also: hats.
http://www.joystiq.com/2010/10/21/user-created-tf2-items-bri...
"Each content update started with a teaser trailer that hinted at several possible new items or features, and Valve developers would monitor the community reaction in the forums to determine which aspects caught the players' attention."
I had seen these, and had always thought Valve already planned out what was going to be included. Like magicians, they set themselves up for victory - no matter what the player ends up wanting.