Multiplayer Skyward Testing!

[x4000]

Hello!

I have surprising news. I've been working this whole time on keeping skyward mp-safe as I coded, but have not tested it in mp even a tiny bit. No point until sp settled down, right? Keith graciously offered to take a bit of time away from he next ai war expansion, and spent a day or so getting mp set up as you may have noticed. He found and fixed a few things in he actual gameplay actions related to mp, but mainly his work was on new code -- connections working at all; a manage players screen, some underlying logic, switch to a host-client model instead of a client-server model, etc.

After all that he was able to connect and place some tiles and such. Fine, I figured. Probably still a million bugs, and I'll get to those later. Well, today, I was planning on starting my own testing and seeing what I could see.

But last night Keith and his wife started playing a co-op game of skyward, and got to turn 10 without it having any mp-specific issues. Saving your game seems to kick the clients, which is bad. But apparently that was the only thing noticed thus far.  I'll fix that early this afternoon (EST).

So! With that in mind, I guess it is actually time for real testing of the co-op by more people than just me, if it seems that clean! I'm sure there are other bugs in there, and I definitely want to root those out.

We also need to think about what counter force there should be, if any, to the fact that people get 1x the action points per player. Faster woes? More frequent bandits? The factional stuff will balance itself out, I'd think. My inclination is just to increase the speed of bandit spawns.

Thanks as always for your help!

[PokerChen]

so you didn't just assign the two players to a colour each? O.o

Faster woes required, I think. 2 x actions equates to more than twice the woe countering ability, imo...

I'll be back home on Monday so no testing for a while for me.

[Mick]

Having each player control a different color makes sense to me.

It naturally increases the difficulty because each player has to fight the urge to destroy the other.

[Tridus]

Having each player control a different color makes sense to me.

It naturally increases the difficulty because each player has to fight the urge to destroy the other.

That sounds like it could be really neat, from a player coordinate POV. Having both players control everything could be confusing, too.

[keith.lamothe]

Bear in mind you can have up to 8 players, so the one-color-per-player thing doesn't really work past 2

[Tridus]

aaw. Ah well.  Could just do it by player house rules I suppose in a two player game: "you only use those guys and I'll only use these guys!"

[PokerChen]

But Keith... the prospects of 8 town colours battling on the same. board....

Well, I think when one gets past 4 players, 2D geometries isn't going to cut it anymore.

[keith.lamothe]

But Keith... the prospects of 8 town colours battling on the same. board....

Well, I think when one gets past 4 players, 2D geometries isn't going to cut it anymore.

The word you're looking for is "Expansion"

Quite literally, in this case.

[Pepisolo]

Just make it like a co-op/competitive footy (soccer) game. You can choose either side or you can choose both on the same side. Or 3 players controlling one team with 1 player on the other team... etc. You're covered then for when numerous other factions are added as a result of the game's phenomenal success... possibly.

[LaughingThesaurus]

I'm into the idea of increasing woe frequency. I've been giving the game more testing now that finals are finally over. It's come a particularly long way, and it seems like woes are a big central part of gameplay. Granted, I've just got a flood coming in a little while, but I do have to keep lakes away from my precious town centers with some pretty judicious smiting, but that would practically be a free action with double the action points across the two players. The major things that seem to come into play that will really ruin the balance are bandits and woes. You might play with mixing an increase of both for multiplayer.
I'm curious if you can make this game work for hotseat multiplayer as well. It'd just be more cycles of red/blue before it goes on to the next turn, with a bit of a different dynamic resource-wise.

[Pepisolo]

Post1.0 idea incoming...

Thinking some more on this, it seems like you may be able to have a unified co-op/competitive multi-player system. Let's say that there are four factions, Norse, Greeks, Romans, and the Cyberbots. Now let's say that there are only 3 players, just to make things a little more complicated.

All players enter the multi-player lobby. There would be two main options for the host to select, Omnipresent or Squads. Omnipresent is how I understand the
game works at the moment with every player having control of all factions. Choosing Squads would allow the player to customize which factions they control
as well as which players are on each individual faction, in the manner of a soccer game such as Pro Evolution soccer.

Norse   ---   Greek   ---   Romans   ---   Cyberbots

Player 2      P2            P1 + P3        P1 + P3

This setup could be either co-operative or competitive, or a third sneaky rat-in-the-house scenario whereby one of the players is secretly trying to sabotage
the other team, either for kicks or because they want their faction as the last standing. It's entirely up to the player. This sounds cool to me!

Other ways to play, 4 player husband/wife teams:

Couple 1      Couple 2      Couple 1      Couple 2

Aside from the issue of programming feasibility, which I pretty much have no clue on, the only situation that would be strange is one such as:

P1 + P2 + P3 trying to be on the Norse side and nobody else trying to control any of the other factions. If there was a CPU AI to take control of the other factions that would be great, but since that can't really happen (at least not without a lot of hard work), you would just have to instruct the player that all factions must be under human control.

Yes, this is a little outlandish, and has no chance of happening for 1.0, but it struck me as cool, so I just thought I'd get the idea out there.

[Pepisolo]

We also need to think about what counter force there should be, if any, to the fact that people get 1x the action points per player. Faster woes? More frequent bandits? The factional stuff will balance itself out, I'd think. My inclination is just to increase the speed of bandit spawns.

I don't like the idea of any of the in-game factors changing. Seems like it would mess with the balance of the game. Couldn't we manipulate the actions points somehow to split them between however many players there are? What if you alternated between each player once an action point is spent? This way you would keep the same amount of action points, just spread between the players. Or how about a shared AP pool? Somebody might try hogging all the points themselves, but that's what ear-bashings on comms are for. Hmmm, it's a tricky one...

[x4000]

As an FYI, I really really don't ever see us having more than red, blue, and yellow in the game.  I could wind up eating those words, but we'd have to recode an enormous amount of stuff for that, and the interface wouldn't support that cleanly, either.  I also don't think it's remotely needed.

My goal with co-op is not "let's make a different game that is related to the first."  It's "that looks like a cool game, I want to play too."  None of the suggestions tossed around here really fit with the second thing.  This game is predicated around the premise that you control all sides except for the bandits.  Changing that is a completely different game, right?  Choosing sides in MP was something Josh and I tossed around for a while, but we discarded it because it just doesn't fit with the game.

For now I'm making it increase bandit frequency when you have more players.  But I also think that some of the other changes that I have in mind that are coming up will make this implicitly balance itself a bit more.