Civ 5 Multiplayer: Some Simple, Important Changes Are Needed

ArcadicGamer said:
Assuming your all friends, all it takes is a save game and a reload and anyone can host the game again. Not a major issue.
Click to expand...

Have you tried to do so with 15 players ? It would take ages to try and only to find out that it's impossible !!!! Even with 5 ppl that shouldn't be a case to be forced to reload the game. There should be simple and common methods implemeted to avoid such situations.
 
For me the OOS issue in the underlying design of the code should really change. In Civ-IV the entire network game of multiple PC's transpires as an identical state machine on each machine, with only user interface changes trafficing across the network connection to alter the state machine identically on all computers. The only mechanism to detect problems is a checksum algorithm, but that is it. The OOS is quite fickle because the code has no protection for data inconsistencies at all. These inconsistencies slip in for a variety of reasons, sometimes modders make mistakes in their code, sometimes pro coders make mistakes. Some times there are network issues. Sometimes people's installation of resources cause issues. The days of low bandwidth connections are over, and so now I dream that the MP code in Civ5 should be able to exchange more data flow to keep the game state consistent. Data structures across the network should be self repairing (because of redundancy). Code execution and state should be restorable to a known consistent point in the flow. If resources and infrastructure do not match on game load, they should be brought back to a reference standard and corrected during load across the network. These are tricky issues.

The really difficult bit will be how to help amateur coders from causing MP problems by bad coding. Functions should be inherently MP safe, but I suspect that even in a perfect world, modders will still cause MP issues. Maybe there is a clever way.

Cheers
 
Yes I'm sure the the system can be improved even if we stay with a GS powered p2p system.

But the problem is always going to be that data is critical in a TBS game, unlike a FPS the game can't guess at missing data it has to ask to have it retransmitted.

That is always going to present a problem to the programmer that program the MP code, how do we handle data loss from lagging players? When do we call it quits and drop the player?

Now the problem may be that GS is always the invisable player and also adds to the data being transmitted each turn, I mean lobby games are listed even while in progress, and this is good because it allows us to rejoin games without having to know the hosts IP. But we have to optimize this code as well to minimize the effect it has on the amount of data being transmitted.

In the end If I do get the chance to help firaxis beta test this game you can be assured that I will identify all the issues with MP in the past and do what I can to change then for the better.

CS
 
I was unaware of how bad some of these multiplayer issues were for the bigger games. I'm a big fan of the Civ MP, but mostly just play 1 v 1, although I plan to play more FFA in the future. One thing that struck me is that u guys have to really love Civ MP a lot to go through all that waiting and OOS, and restarting, just to play the game.
Do u guys who play the large player number games use Comrade or some other IM program to keep track of everyone when u do play LAN games? I would think that would help some with the not knowing which game to rejoin problem. Also, do u ever try to play games more with dedicated ladder players on LANs, but to keep the flavor of the larger games use AI to make up the difference. What I mean is, having a LAN game with like 6 to 8 dedicated players, but use the AI to fill the other 8 or so slots. Now I don't know if this would be acceptable to the gameplay for u guys, jus throwing it out as a suggestion/wondering if u guys did it.
Sounds like u guys who play the large number of people games dont play much 1 v1, lower number FFA, etc. Is this true? If so, why don't u play these type of games more? I always liked that the 1 v 1 , while true can't have all the dynamics of a huge map 14 person game, still eliminated alot of the randomness and luck that would have to come into play when u are in a game that large. And I never found my 1 v 1 games to lack depth or interest (it did get a little tiring playing the tiny map 1 v 1 on ladders, as people were soo good that the games would end soo fast and it didn't seem as fun to me, so I would like to play 1 v 1 on some maps a bit bigger and wiht different start eras to mix it up. Anyway if my rambling has no relevance to what u guys were complaining about, then I am sorry for wasting ur time.
 
King Lewie XIV said:
I was unaware of how bad some of these multiplayer issues were for the bigger games. I'm a big fan of the Civ MP, but mostly just play 1 v 1, although I plan to play more FFA in the future. One thing that struck me is that u guys have to really love Civ MP a lot to go through all that waiting and OOS, and restarting, just to play the game.
Do u guys who play the large player number games use Comrade or some other IM program to keep track of everyone when u do play LAN games? I would think that would help some with the not knowing which game to rejoin problem. Also, do u ever try to play games more with dedicated ladder players on LANs, but to keep the flavor of the larger games use AI to make up the difference. What I mean is, having a LAN game with like 6 to 8 dedicated players, but use the AI to fill the other 8 or so slots. Now I don't know if this would be acceptable to the gameplay for u guys, jus throwing it out as a suggestion/wondering if u guys did it.
Sounds like u guys who play the large number of people games dont play much 1 v1, lower number FFA, etc. Is this true? If so, why don't u play these type of games more? I always liked that the 1 v 1 , while true can't have all the dynamics of a huge map 14 person game, still eliminated alot of the randomness and luck that would have to come into play when u are in a game that large. And I never found my 1 v 1 games to lack depth or interest (it did get a little tiring playing the tiny map 1 v 1 on ladders, as people were soo good that the games would end soo fast and it didn't seem as fun to me, so I would like to play 1 v 1 on some maps a bit bigger and wiht different start eras to mix it up. Anyway if my rambling has no relevance to what u guys were complaining about, then I am sorry for wasting ur time.
Click to expand...

You're probably thinking games are larger than they are.

A typical FFA is 5-8 players. Less is not too common, and significantly more than that is usually considered very risky due to lag and crashing problems.

A typical teamer will either be 3v3, 4v4, or 5v5.

Personally, I've played more MP games than I could even count but have only played on a Large map a handful of times. I've never even played on a huge map.
 
I think some of you guiding this topic in a wrong direction. The basic message is we need a completly new multiplayer engine to perform as in other games on the market. Solid, reliable, manageble. In which only problem you might have would be when your ISP let you down - not the game itself. We don't want them to patch old system - they've failed to do it for last five years. And there is a reason. The old engine is causing all the problems due its nature-architecture. To fix it they would have to rewrite it all + choose other strategy than peer to peer + udp. Just don't try the old cheap tricks.

If a new system requires dedicated servers - I don't mind - I can even fork out couple of quids monthly to stop wasting my time waiting / fixing- and I'm not talking about minutes every game-day - these are hours - repeat - hours wasted every session. I'd rather pay monthly for having these hours saved to do other things than to keep fixing the game.


BTW - I can easily live without great graphics/all the eye candy stuff. I'm not excited at all about all new animated leaders and other gfx bollox. Ok - It's important for product commercial success - all kids will love it + buy it. That's why you did the Civ Revolution and Colonisation - kids are happy playing it on xbox. I agree the product should be graphically appealing. But if the gameplay sucks the whole game sucks. After all - this is a board game - not a crysis. Please do a great board game this time.
 
Your right I would be great to have a brand new MP engine with a brand new 2K Server at it's core. But I think, and probably most people here as well, expect that they will still use Gamespy.

So the best we can hope for is to use the GS p2p system to it's maximium performance and ensure that the games MP code is as optimized and effiecent as possible. And of course add all the game side MP features we can get.

Neither 2K or Firaxis have any history of running their own game servers, to expect that they will invent all that infrastructure for just Civ5 is unrealistic.

So that is the jist of all our ideas here, add all the features that we want and make the best p2p system possible.

CS
 
They sell milions of copies of the game - don't see anything stopping them from doing it as good as other much less popular titles successfully do.

Involving Intel in their marketing affair makes me thinking they cannot release the game not matching today's standards.

Perhaps some internals will speek out here (or wherever) about multiplayer engine to cut down all speculations and make us confident to preorder the game.
 
Well the involvement of intel and the GPA 3.0 engine, seems to me to indicate that Civ5 will be able to run on a wider range of hardware than Civ4 did in it's era.

But of course if you have a lower end PC you can expect to have a much drabber looking game, and probably wait longer for the AI code to run each turn etc.

Yes other games have good MP, but to be fair we should only compare Civ to other p2p games using GS, and see how they fare as well.

I have no doubt that civ5 will come in at or below average system requirements for a 2010 game. The problem is that Civer have got used to running former Civ versions on very weak hardware and despite Civ5 being a lot more system friendly than most other new games, these people will have to upgrade.

CS
 
bossconian said:
They sell milions of copies of the game - don't see anything stopping them from doing it as good as other much less popular titles successfully do.
Click to expand...
The Battle for Wesnoth wargame is freeware yet i had no problems playing online on their server. So, if a non-profit TBS game can have a stable working multiplayer and online lobby, Firaxis should be able to pull the same miracle with Civ 5 too. Well, of course they flop it, but that's another story :lol:
 
Yes it certainly is within the resources of a company as large as Take2 to build a server system for Civ5, but no doubt they will simply take the cheaper route and continue to contract to Gamespy.

CS
 
CanuckSoldier said:
Yes it certainly is within the resources of a company as large as Take2 to build a server system for Civ5, but no doubt they will simply take the cheaper route and continue to contract to Gamespy.

CS
Click to expand...

No doubt I'll be contracting Al Qaeda to suicide bomb Gamespy's headquarters then...
 
mrt144 said:
Most these are related to the :):):):) quality of the people you find on Gamespy.
Click to expand...

And while you can't change peoples online behavour, Firaxis could pay to have the lobby moderated by GS......or whatever other service they might use.

CS
 
I would like a bit more transparency into the multiple alias that people have, the ability to easily change your name encourages bad behavior.
 
You must log in or register to reply here.

Similar threads

H
Civ 5 load time taking too long
Replies
1
Views
422
Naokaukodem
Naokaukodem
B
Civ 5 is disappointing
Replies
5
Views
4K
Bonyduck Campersang
Bonyduck Campersang
The_J
PCGamer: Sid Meier was apparently baffled by the decision to launch Civilization 5 without Genghis Khan: 'Sid said, What?'
Replies
1
Views
798
Quintillus
Quintillus
D
Hi gang - it's been awhile 😊
Replies
3
Views
234
raider980
raider980
P
[G&K] Difference in difficulty from Emperor to immortal
Replies
1
Views
648
Kriogen
K
Back
Top Bottom