[The Qwerty One]

I'm having annoying problems playing Team Fortress 2. I have Roadrunner cable and I keep getting bad lagspikes. The closest I could find that matched my same problem is this video. http://www.youtube.com/watch?v=aJECcRpqPZ4

I'll be playing and I just suddenly can't move and if there are other people on the screen, they'll be frozen too (animations still work such as flame or legs moving if someone gets frozen midair), but I can still look around, see text entered in the chat box and even hear people on the voice comm.

When this happens, netgraph suddenly displays no bars at all for the duration of the lag (except when someone uses their mic) followed by a really tall bar. I don't really know what that means but it's probably not good. I noticed that the bar is made mostly of "other players" and "entities" though.

I ran a ping test in cmd and noticed that I get "request timed out" fairly often. Another thing that might be related is that I always randomly get disconnected on my instant messengers.

So it'd be cool if someone could help me diagnose my problem. It's happened ever since I started playing, though it use to be much worse but it still happens about 20 times in one round.

[BG-0]

It's a Valve issue. Whine to them, give financial support so they can stop wasting time making the game into an MMORPG and fix all the ****ed-up-ness of the game. They've failed to fix several bugs for the duration of 13 updates now, and yet they keep adding only useless crap to the game. There's no fix.

[The Qwerty One]

I'd like to think it's just a problem with the game but I played at my cousin's house yesterday and it was a little bit like heaven compared to my connection. I take it some people get it and some don't but what's wrong with my computer that makes me prone to the issue?

[BG-0]

Could be just about anything. It's pretty mysterious how other source games may run on the same computer just perfectly and then this freeziness happens in TF2 only.
From what I've observed, it seems a bit like that the problem's magnitude correlates with the value of choke(netgraph 4 shows it at least, packages that could not be sent at an interval, prolly an average of a second's time) that some of the updates added quite a bit earlier. Some machines suffered a lot, some less. Could be that some hardware components have been somehow made less compatible, and they don't let the engine run properly... I really dunno.