I'd rather have MM work on AI as opposed to TCP/IP - Page 3

igoblin · #21 January 10th, 2001, 06:24 PM

I would have to disagree, TCP/IP would be great for SE4 because unlike Master of Orion 2 it has a simultaneous mode. This way everybody takes their turns at the same time so the only waiting would be if someone takes exceptionally long for their turn. You can also play with more than 2 people in simultaneous unlike turn based games.
I will probably still do play by email though because its hard to schedule times for everyone to play together. I haven't tried PBEM since the demo, I hope the bugs were fixed.
-igoblin

Daynarr · #22 January 10th, 2001, 09:45 PM

quote:
Originally posted by igoblin:
I would have to disagree, TCP/IP would be great for SE4 because unlike Master of Orion 2 it has a simultaneous mode. This way everybody takes their turns at the same time so the only waiting would be if someone takes exceptionally long for their turn. You can also play with more than 2 people in simultaneous unlike turn based games.
I will probably still do play by email though because its hard to schedule times for everyone to play together. I haven't tried PBEM since the demo, I hope the bugs were fixed.
-igoblin

IIRC MOO2 had ONLY simultaneous mode. You would give orders and when you click end turn button, they would be carried out, just like SE4 simultaneous mode. I have played several games of MOO2 on Heat.net but they would never be finished. It would just get too long and boring to wait for all of the players to finish your turns (they would get really long if a player had some large tactical combat). That is a problem with turn-based games, AFAIK.
However, I do think that SE4 should have TCP/IP multiplayer, but I believe that the majority of players will play it SE4 in singleplayer or hotseat, so AI should be addressed first. Personally, I don't have any preference, as I like to play both multi and single player games, so I would vote for both.

dmm · #23 January 10th, 2001, 10:04 PM

My $0.02: First improve the AI.

Selfish reason: If I'm going to play humans, it will be members of my own family, and for that I can use hot-seat. So I personally don't care about TCP/IP.

Unselfish reason: Improved AI will also improve TCP/IP games (and hotseat games, and PBEM games), unless you're going to play with only human players. No, even then, it helps, because strategic combat will be better. So improved AI benefits everyone, whereas putting in TCP/IP only benefits one segment of the market.

Londo · #24 January 10th, 2001, 10:35 PM

Consider me a tcp/ip voter. I love the challenge that only a human opponent can provide. For my own use, I can only do hotseat with a friend of mine one day a week, but I could do tcp/ip daily and therefor get more bang for my buck with this game. IMO, improving the AI is a gradual process that we modders can participate in along with MM. Tcp/ip is something we can not do on our own, and it represents a revolutionary step as far as game mechanics. I am certain that it will be difficult to code, but it will be well worth it based upon the demand that I have seen for it on this site.

------------------
Vir! You have made a mess of this post once again, now give me a large cup of brivari so that I may begin to forget the bad times and remember the glory of the great Centauri Republic.

Socialist · #25 January 11th, 2001, 02:44 AM

Lets put it another way, people who use Email
spend much more time to finish games..almost rediculous to set up a game and play it for a year (patchs themselves will goof up games with so much time spent)

With tcp..hardcore..and i mean hardcore wargamers.in smaller Groups could actually play and save. There are ways of allowing people to look around while others do their turn and do it right. AOE had some workablity in that area. resonable games..
mid level galaxy 60-70 systems or so..we all know huge 200+ would be rediculous. even solo i dont do anything that big..100 systms is huge in my opinion. The point being, people would play each other and can get back to one another. Wont take long to pick out the losers who dont come back and finish what they started. Keeping games at 4 or so players is workable..and reasonable..finish working on the AI, im not against that. What does bother me is if they dont make a
good "Stable" tcp ip connections..turn based game work great in TCP IP, just needs a little time and effort and most will be satisfied.

sorry but playing a AI is meaningless long term, they are just "extras" to move aside ...variables in the mix..in between Humans.

An AI doesnt care if it wins or loses.

------------------
Waves his Red flag

Socialist

Voidhawk · #26 January 11th, 2001, 07:02 AM

Just to elaborate a little on what dmm touched on: I think he's very correct.

If you think about it, coordinating an ongoing TCP game with 2 or 3 people shouldn't be a problem. Coordinating a TCP game with 5 or more people could get hairy. Add even more players and the problems increase exponentially.

"Well, so-and-so can spare an hour saturday between 8 and 11, but what's-his-name has to take his wife out. Meanwhile who's-his-face has a wedding to attend sunday, so he can't make it saturday after 10. Etc., etc..." Schedule problems will also be compounded further by different time zones.

With this in mind, it seems that the most enjoyable TCP games will be played with around 3 or 4 people. And a large quadrant with only 3 players could get... well, boring quickly. Now, fill the remaining slots with a few of the current Version AI's and you've got basically just a few minor hindrances until you can get to one of the other human players. Fill the empty slots with a few good, solid AI's and the game would really take on some color. It would be really cool to have to seek out alliances with your human buddies in order to survive a vicious AI onslaught. I guess you could accomplish this easily enough with the current AI by using Team Mode, but then diplomacy with the AI would be rendered useless.

Then again, all of this is pure speculation as we have yet to see how well TCP would work. So maybe I'd like to see the TCP first so we can start seeing how it plays. Honestly, I started writing this as an argument to improve the AI first, but now I find that I'm completely torn on the issue.

Sorry for wasting everyone's time with this absurd post of mine hehe

...

Derek · #27 January 11th, 2001, 07:18 AM

Also, a good AI would make it difficult to tell whether the person you are facing is an AI or another human. (That is, without relying on the content of text Messages they send you...)

Derek

sogard · #28 January 11th, 2001, 09:22 AM

One feature which is common in several tcp/ip enabled games is a time limit on moves. If you are playing in simultaneous mode, you just set a time limit that the players are permitted to make their moves and I don't see why the same concept could not be used for a turn based game. This simple device will keep most games moving along. The only problem would be IF tcp/ip would permit tactical combats to be played. At present, that only works with the turn based mode. The nice thing about hotseat play, at present, is that you get to do the tactical combats -- which can be a bLast.

My current hotseat game is on a large galaxy map with the maximum number of players (two of which are human) with the AI set at the most difficult levels and bonuses. So far, the game has really been a bLast.

Atrocities · #29 January 11th, 2001, 10:53 AM

Yes, time limits on moves would be a must for TCI/IP play. A real benifit to the game. However, in some cases, that does cause crashes as the packets get lost. (Rebellion, BOTF)

Overal, I think its much more important to support TCI/IP than anything else at this point. Make the game universally accessibable to players over TCI/IP, and the game will grow in leaps and bounds.

Look how long BOTF was active at the Zone. Dispite its flawed MP, it was still a great game to play, and poeple played it there for over a year and a half. Rebellion was also a success in that regards although limited to only two players. (that was its down fall)

SEIV has the potential to be a "long term" game. You set up in the morning, and play all day. Of course, you would want to play with about 5 to 8 players, with a medium galaxy (100 to 125) systems.

Otherwise, you'll be saving and coming back to it again and again.

The one BIG problem with TCI/IP games is the fact that people can not dedicate the needed time to play a full game. They will often, more than not, drop out resulting in a CRASH. AOL players get shafted here as AOL will regularly shut down there connection. The people with Cable and DSL will get spiked every two or three hours as there provider randomized there IP's on a DHCP system. (Same as getting disconnected for a minute or two)

Over all, TCI/IP will be boost to the game, but it does have its draw backs. But the players of this game are dedicated types, and I would hope are into a game for the long hull.

Well those were just my thought.

------------------
"We've made too many compromises already, too many retreats! They invade our space and we fall back -- they assimilate entire worlds and we fall back! Not again! The line must be drawn here -- this far, no further! And I will make them pay for what they've done!" -- Patric Stewart as Captain Picard
UCP/TCO Ship Yards

imperishable · #30 January 11th, 2001, 11:39 AM

I wouldn't suggest getting 5-8 people into an IP game. That's cause once you save the game, you will not open it for several weeks, if not months - that's how hard it is to get everybody back together again. I'm speaking from my experience with Alpha Centauri. If you want 4-20 people, pbem is gonna end up as faster way to play. IP games are best for 1 on 1 playing.