There's a problem with the recent trend to emphasise online multiplayer in games, and it is this: nobody's playing.
On XBox Live today if you want a multiplayer game of Halo 3 or Gears of War, you'll be cocking your weapons within minutes. But if it's The Darkness or Crackdown you're after your chances aren't so good. Heaven help you if you were hoping for a ranked match of Catan or Doom.
Catan, it should be noted, has no local multiplayer. So if you've just shelled out your $10 or whatever on Live Arcade and discovered that all you can do with it is challenge some rather lame AIs you'd probably be right to feel more than a little cheated.
These things have critical mass. If no-one is playing Mario Strikers Charged Football today, then no-one is going to be playing it tomorrow. Waiting in an empty lobby in the hope of someone showing up is not fun; and if no one is willing to wait for an adversary to arrive, then no matches are ever going to be made.
Multiplayer content takes time and money to develop. That's a cost that's explicit, in that it's passed on to the consumer by way of the purchase price, or implicit, in that it diverts time and resources away from the single player experience. When something with a reasonably well fleshed-out online mode like The Darkness simply has no multiplayer following, that's a massive waste.
Multiplayer content also decays. What may have a reasonable community at release may well have no-one less than a year later. While the Halos and Counter-Strikes of the world may stay strong year after year, they're the small minority. Try getting anyone to play The Ship or DefJam Icon. Achievements in these games that require multiplayer action grow more unreachable with every passing day.
One solution is passive matchmaking. This means that matchmaking occurs while players are engaged in other content. In Crackdown it works reasonably well, where you can be on the look-out for players who want a co-op game while you're working your way through the single player story. In World of Warcraft, though, it works poorly, where the Looking for Group interface is badly supported and ignored by the majority of players.
Before committing to developing online multiplayer content, developers should seriously ask themselves who, exactly, will be playing. Where will the boundless hordes of players come from, and why will they keep coming back for more? How will you stop them from having to wait around in empty lobbies? And what will you do, eventually, when it turns out nobody's playing?