FORUMS_

Ss78_ wrote:

 

Matchmaking and hosting and how it's been implemented currently?

 

That's my main issues with this game and the apparent lack of cut off points leading to the code being unable to efficiently handle the huge disparity it itself creates.

 

Or you could just do a pure sync video like we've recommended. Up to you.

I'll leave the sync to that other guy. I know how much work that is, and have no desire to increase my workload unnecessarily ^_^

Matchmaking is such a mess and such a popular topic, but it is incredibly easy. Request server list, sort by ping, wait, refresh servers below **ms ping, join lowest average ping server with open slot. Done. It's just automating what PC players have been doing manually for almost 2 decades.

Perhaps I'll discuss matchmaking/hosting in a commentary on networking itself, sometime in the (assume distant) future.

There will be a new commentary going up eventually, but I hesitate to give a timeframe. I will, invariably, find a way to not have it finished on time. No matter what time that is.

playerone2 wrote:

...

Didnt they use a new engine in bo1. is it passible that the mw2 engine is just much more simple. ..

 

I heard something about them using unreal 4 engine. Supposed to be easier to build a game with, but will it be optimised for internet?

I think everyone can agree that this is an online multiplayer. and the campaign is extra. With the last 2 games being as broken as they are. i havent even given campaign the time of day.

 

Cod should not even add campaign anymore. and just give us 5-10 more maps to begin with.

...

The MW2 engine is superior in every way, there's no questioning that. But they can't use it. If it were my decision, I would've paid the old IW team a million dollars each to stay and still made 1000% return on investment. Alas, the legal problems were merely a symptom of something far greater.

There are built-in online features in many engines, but in a game of this caliber it's up to the game developer to code their own. Those built-in features are rarely optimized and never specific for your game, no matter what engine you use.

The campaign is the only part of the game that justifies the $60+ they charge for it (in effort, not value). Don't expect it to disappear anytime soon, even though many gamers never play it.

@jessedart <3

playerone2 wrote:

...

 

Maybe all the advancements in the graphics dept. leads to problems in the networking dept. maybe just more data to sync?

...

The problem is strictly networking in this case. Rendering graphics is entirely separate and done on the client. In a COD game, where there are no destructable walls, even physics results need not be sent over the network.

That's not to say that increased demands are not a problem for consoles. But the calculations necessary for 'lagcomp' are such a small percentage of that, sacrificing it would not improve anything elsewhere.

playerone2 wrote:

I agree to that. But if they spent less time on campaign maybe theyd have enough time to get the networking right(prob not)

 

...
theres hundreds of games with storylines that are fun. but theres only one online muliplayer like COD. and thats cod.

They don't need time so much as they need motivation. It wouldn't take much time at all to fix the networking issues.

Perhaps this would be solved if they had some competition then.

RatedM wrote:

 

Ovenbakedmuffin, regarding the discovery of the circa 200ms advantage for players moving from cover - could you possibly test for this on an earlier COD like say MW2? I'm working on the assumption it is simply a characteristic of the games engine used in the COD series since COD4 as opposed to a network related issue/camera issue unique to BO2.

 

Just a quick test, nothing extensive, purely to verify the above.

I did. See 8:17, MW2. Only 5 points of data for each, as I had no intention of repeating the tests to the degree of the BO2 section.

>Baseline player delay: 111ms less than BO2.

Of note, BO2 differs from MW2 in at least the following aspects:

>BO2 delay increases equal to the player's (full) ping vs half ping (outgoing as player). This suggests a method similar to lockstep, where the most lagged player determines the delay.

>Synchronized game-state - player deaths occur at exactly the same time across all clients, regardless of ping. The same is true of player location / spawns / other actions, so two observers will see exactly the same gameplay. (This confirms a method similar to lockstep.)

>Authoritative interpolation occurs at >800ms in BO2 vs >200ms in MW2. (Rubberbanding above that ping, but players cannot "run ahead".)

The synchronization is what causes this form of networking to fail. They have compensated, if you will, for actions but not reactions. They would need to add a second layer of delay to allow for players to shoot back in a fair environment, thus producing a very noticeable delay and defeating the original purpose of networking. An excellent form for any gametype besides FPS, really.

Rest assured, I've checked and rechecked my data for everything I could think of (and annoyed my subs by delaying the entire commentary for that). But I welcome thoughtful responses such as this. No one is perfect.