Chris was hired 2 1/2 months ago there so it's not exactly news. The real issue is that last week Ubisoft delayed Farcry because the multiplayer was in such a sorry state. To make this look like a good thing, they are trying to make it seem like they hired Chris to make the multiplayer better when in fact he was already working there and the publisher simply rejected their milestone.
The real problem with multiplayer is that the original multiplayer programmer had no multiplayer programming experience. While he is a very smart guy, the system he designed is fundamentally flawed. It tries to approach the problem as "Send the entire game world x times a second and simply set the client to that state." In a 0 ping enviroment that would work. However, experienced multiplayer programmers know that you have to use interpolation and extrapolation, as well as sending data contextually rather than simply a world dump. As a result, the net code has no mechanism to hide lag and requires server feedback on all actions (even though where it is not necessary, such as aiming). Also, because so much of the multiplayer is scripted you lack the kind of low level control necessary to implement this. This is all part of the core design, which unfortunately is so pervasive that the multiplayer is unfixable short of throwing it all away and starting over from scratch.
Another problem with the core design is that it takes inherently so much processing power that at 6 players you generate 60 ping. Past 60 ping the game is unplayable. So even on a LAN with an effective 0 ping, you have a 6 player limit.
This may not have been so bad had the management been better. If they had hired an experienced multiplayer programmer as an consultant to assist with the original multiplayer programmer, for example, perhaps things could have been turned around. Unfortunately, (among other things), they pissed off the original MP programmer off to the point where he quit (about 4 months ago). To make up for this, other programmers were diverted from other parts of the code. None of them had any multiplayer programming experience. Those of you who are experienced in the industry are probably nodding your heads now. As we all know, dumping programmers into an already late project, ESPECIALLY into foreign code they are not qualified to work on, isn't going to get the project done sooner. The Crytek management took exactly this route with predictable results.
This comment was edited on Nov 5, 21:34.