Profile of Inverse

Jump back to quoted post18 Dec 2013, 21:42 PMtheking10

Yes Noun they did, but they did not complain about 3 seconds input lag between orders for all units in every single game. My router firmware is up to date and i have a 50 mb download speed. Sega suport were not helpful and could not solve this issue.

I know for a fact that this problem is fairly widespread as I have spoken to many people with the exact same issue and have seen numerous posts all over the forums as have you in regards to this. Now I dont know if you will totally ignore this or will look into it on behalf of myself and all the other people having this game breaking issue but I truly hope you will investigate this thoroughly as it is destroying my enjoyment and many other peoples of this excellent game.

@Wifi I am unsure if you are directing an uncalled for and personal insult towards me or something else..., I am speaking from what I have experienced and from what other people I have met in game and on the forums have also said to me. I will stress again , I have had no such issues with input command delay with CoH 1 only for CoH 2 . This is not normal lag, as the game is still running without freezing or skipping. Change the 0 to 0.001 its still not noticeable. The commands/ orders in CoH 2 just DO NOT register until 3 seconds after. It is plainly obvious that it has nothing to do with the animations as some people are not even having this issue while others are getting a horrendous 3 seconds.

@Jin not based on anything ? I am telling you having played CoH 1 from the very first day it was released this was never a problem. Just because you are not having issues do not make the incorrect assumption that others are perfectly fine as well. Do some research, read some posts about input lag and you can see for yourself about the numbers do this before making any stupid statements.

Jump back to quoted post18 Dec 2013, 01:53 AMtheking10

@Inverse

CoH INPUT LAG 0

CoH 2 Input LAG 3 SECONDS

You do the math

Because client-server in an RTS game is impractical. There isn't a single RTS game on the market right now that's client-server. The closest is SC2, and even then the server is only used to validate and route information. No calculations are being made server-side.

Why? Think about it. An RTS works by communicating commands between players. Each player has an exact copy of the game state, and simply processes the commands (move orders, build orders, etc.) it receives from the other player(s). In an RTS, though, that can mean processing potentially hundreds of individual commands at once, since there can be dozens of units on the screen and commands being made at any given time.

When all you're doing is sending basic commands, that's actually really easy to do. It only takes a few hundred kilobits a second to relay that information to a player. But if you try to get a server involved, everything increases exponentially. You have to send your commands to the server. Then the server has to process the game state with those commands and determine the state of the game at the next tick. Then the server has to send the entire game state to both players. Both players have to receive that game state so that they can continue playing. And because the server is running the game and the clients are just running shell instances of it, the information that has to be sent is exponentially larger, and grows rapidly the more units the server has to process.

That's why client-server doesn't work in RTS games. SC2 compromises nicely by routing messages through servers, which makes drop detection and some aspects of anti-cheat better, but doesn't do anything about latency. The only reason SC2 has better latency in general is because matchmaking is limited to the server you're on, and the player base is so much larger so the matchmaker can favour lower-ping matchups much easier.

On the other hand, client-server works great in MOBAs like LoL and Dota 2 because there are far less player-controlled units per game (usually 10-20, up to a max of about 50 or so in a 5v5 match) and far larger game sizes (pretty much every match is a 5v5). This means you're using less server CPU time and bandwidth for 10 players in a MOBA than you would for 2 players in an RTS. Economically and technically, it just doesn't make sense.

EDIT: For another perspective, read the first answer in this thread: http://gaming.stackexchange.com/questions/65049/how-long-is-starcraft-2s-typical-input-delay-compared-to-other-competitive-game
He explains it a lot better that I could.