Multiplayer strategy games

[Margalis]

So I've always liked strategy games, especially "SRPG" type games. (Tactics Ogre being by far my favorite) I have been thinking of how to expand strategy games into the multiplayer arena in a way that doesn't suck.

So I wrote up a little something I thought I would share. It's very specific, it's not a "game design" or anything like that, it's just a brief treatment of what I see as the fundemental problem: timing. (Or, damn does playing a strategy game online take forever or what)

Here it is, horrible mis-spellings and all. The style is how I often ruminate on things, esentially writing notes to myself.
----

Goal:

The end goal here is a multi-player online tactics game, somewhat similar to a game like Tactics Ogre or Warhammer.


The problem:
Nearly very online turn-based game suffers from similar issues. The time it takes to play a game can be too long, and/or involve too much waiting. We have to solve this problem while still offering some engaging gameplay. It's important to note that we want the solution to scale to N players, not just 2.


Existing Systems and Their Issues:

I go/you go (AKA Igo/ugo): Igo/ugo dates back to the days of seat swapping. One player takes a turn, then another, then another. Even with time limits per turn, this does not scale with number of players. (10 players = 10 times as long a wait every turn). In addition, it's one of the least strategically interesting sytems as well.

Simultaneous Turn Resolution (STR): In this model, each player enters their moves for each turn at the same time, and then the turn is played out. (Examples include the old Mac game Robot (Robo?) Wars and Konami's Vandal Hearts. This system scales time-wise to N players, but I don't think it would scale logically. Having 4 sides all resolving at once would be very confusing. STR games often have weird gameplay artifacts, like two characters walking right past each other without doing anything. I don't think it is a pleasing system, design wise.


Wait Time System (Tactics Ogre): Taken from a FAQ:

"Each TO battle starts with WT counting down, one point at a time, for all characters.  WT = 550 - AGI + Weight of all equipped items + class WT penalty. The first character whose WT reaches zero, that is, the character with the
lowest WT will get the first turn.  During a turn, a character can choose to move and/or act, in any order, or if he so chooses do neither.  If the player does both of these things, his WT will reset and start counting down again from its full value until his next turn.  If he does only one, it will count down from 3/4 of its value.  If he chooses to wait and do neither, it will count down from 1/2 of its value."

Translation: Different units act at different relative speeds (depending on their class, encumberance, etc), and depending on what you do on a turn the wait until that unit's next turn will vary.

This system is GREAT from a gameplay perspective. (For reasons I won't get into) It is also described in language that suggests real-time, although the system itself is turn based.



Proposal: A Wait Time System based on Real Time.

This is nearly identical to the "Active Time Battle" present in Final Fantasy games 4-6. Each unit can act every so often, in *real time.* Depending on what that unit does, the next turn's timing will vary. Under this system, there is an upper limit on how fast you can enter commands. Unlike in a typical RTS game, faster speed makes a difference only up to a certain threshold. That threshold is how fast you are ordering commands versus how fast units are recycling.

This scales well to N players because the timing of your sides moves are totally independent of other players. In fact, your opponents could all be AFK and you could still continue to play your normal game.

One issue here is a fixed pace for the game may not work. When you have a lot of units it may be too fast, but as units dwindle the wait time becomes too long. That can be fixed by picking speeds based on an average wait time, and scaling the speed of time to meet that objective. For example, with 10 units on the board time would count down 1/10th the speed of when a single unit was on the board. The aim would be an average fixed turn rate.

How is this better than time limits? Time limits don't scale to N players, and this does. In addition, you can't end up in a scenario where a fast player is perpetually waiting for a slower one. The speeds of the various players do not interact at all. I suspect that would *feel* better, even if the total time taken ended up being about the same.

How is this better than a fixed OVERALL limit as in speed Chess: Again, that doesn't scale to N players. In addition, a lot of speed Chess is meta-gaming the timer vs. actually playing. One of the strengths of this system is that being fast helps, but only up to a point. Whereas in speed Chess, the faster you are the better, always.

This system doesn't reward you for going quickly, it just discourages you from going TOO slowly.

[Trippy]

Your proposal doesn't fully solve the STR "passing in the night" problem -- it reduces it to a certain extent but doesn't eliminate it completely. There will be many times where a player will be able to move a unit simultaneous with another. E.g. Player A has a melee attacker and is trying to move into range to attack a Player B unit but that unit is able to move at the same time as well. You also get corner cases like if you have two equal speed units, both melee attackers, you could get into a stalemate (a la Chess endgames) if one player is actively avoiding the other.

[Margalis]

Yeah, that's something I didn't quite get into. My thinking was when you actually finish inputting your commands for a unit the actual execution does lock everyone else out. So if I make my wizard move 10 spaces and cast a spell everyone has to watch him do that, they can't make their own guy move AS my wizard is walking. Nor does time count off during that animation.

That doesn't quite scale to N players for large N, but I think it's acceptable. My theory is you are probably interested in what's happening anyway so being forced to wait for it to happen isn't a huge inconvenience. As soon as the walking + casting animation is done I go back to whatever I was in the process of inputting.

As far as the stalemate is concerned, that can happen with any strategy game no? It can certainly happen with Igo/Ugo. Matches would need some sort of time limit but I didn't want to get into anything about the game other than the mechanics of battle.

Thanks for the input.

[Margalis]

Come on, there must be someone here who wants to call me a retard.

[Viin]

Sorry, I can't read that much text in a single post. Maybe if you broke it into little bitty posts, with one post put up every 3 or 4 days ...

[HaemishM]

Ideally, how many players do you want in the game? That's one thing that real-time has over non-real time. Real-time takes the relative same amount of time for all players, regardless of number of players. It doesn't reward deep-thinking, however, at least not during the game. The deep thinking is really the meta-gaming aspect of build orders.

I think if you try to add too many players, strategy games just really bog down into unfun and unplayable. I liken it to miniature games. 1 v 1 or 2 v 2 games are decent fun. You start talking about 4 v 4 and 8 v 8 scenario games and while they can be fun, what invariably ends up happening is you get bored waiting for your turn. Efforts to remedy this (and other things) such as the ruleset Piquet really either work for you or suck horribly, depending on your point of view. I think just about all non-real-time systems can break down into unfun if you just keep trying to throw too many people at it.

[tazelbain]

Turn-based, but everyone takes their turn at once.  Each turn takes X amount of time, unless all Players complete their turn.  If a certain precentage of players fail to fully complete their turn, X + 15secs.   

[Margalis]

I think 4 players in a game would be the max I am talking about.

About simultaneous turns: I don't think they work well in tactical games. You get too many odd issues where my guy tried to attack your guy and your guy tried to attack my guy and they end up walking past each other or something like that. I've never seen a tactical game that had that done well. (Can you name one?)

In my system 4 players would scale and be only a bit slower than 1 player. The slowness would come from when a player inputs a move you have to watch the resolution of that move. But 4 players would not be 4x slower, it would be 4 * K slower, where K was the average animation length. Again, what I mean is if I am going to walk 3 spaces and cast a spell, once I've finalized that everyone would have to watch my guy actually walk and cast, but they don't have to wait while I input those steps.

Edit: It's worth noting that I'm working on a Java-based mockup of the system, I'll let people know when it's ready.

[MaceVanHoffen]

About simultaneous turns: I don't think they work well in tactical games. You get too many odd issues where my guy tried to attack your guy and your guy tried to attack my guy and they end up walking past each other or something like that. I've never seen a tactical game that had that done well. (Can you name one?)

I always thought Deadlock did that very well.  It was turn-based, with the turns running concurrently.  There was an optional timelimit on each turn to make the pace of the game more hectic.  A very fun little game with 8 (IIRC) alien races to choose from.

[HaemishM]

I think the best way for 4-player strategy games are going to be either Combat Mission style or X-Com style.

Combat Mission: Everyone inputs moves simultaneously, then watch then acted out on screen. The only bugaboo is as you said, when two actions overlap or conditions executed in the moves contradict each other.

The X-Com/Fallout:Tactics/Mage Knight/Incubation style is that all players act in one turn, but with action points and initiative rolls determining the order of actions. Each unit/soldier has a set number of action points, and all actions require a different amount of points. An initiative roll is made after each unit's action points are used up, with modifiers based on stats or whatever conditions you wish. Action points can be spent to get "interrupt" or "overwatch" style actions that can interrupt another player if conditions are met, but are wasted otherwise. When all units have spent all their action points, the turn resets.

[Samwise]

About simultaneous turns: I don't think they work well in tactical games. You get too many odd issues where my guy tried to attack your guy and your guy tried to attack my guy and they end up walking past each other or something like that. I've never seen a tactical game that had that done well. (Can you name one?)

Ship-to-ship combat in Puzzle Pirates works like that.  Each side gets something like fifteen seconds to lay down four moves (which are things like "sail one space forward", "turn left", "fire cannons to the left", "fire grappling hook to the right") and at the end of that time, the moves get played out one after another.  It sounds simple, but there's a surprising amount of strategy in it.

Usually ship combat is 1v1, but during "blockades" there'll be something like twenty ships on the board, all player-controlled (and each ship holding maybe twenty players).

[Murgos]

There are a plethora of multiplayer turn based strategy games out there, some with several hundred players.  Mostly they are browser games and they have several methods for dealing with turns.  Some give you X action points every few minutes and then you spend them performing a few actions or save them up and do a bunch of things all at once, these are fun because once a day or so you can go do a bunch of things and then plan out your next set of moves in between.  Or plan to make moves after your opponents initial reaction when you know he wont be able to respond for a while or you have the option to pretty much be doing something continuously.  A strong static Defense and excellent responding ability is key in these games.  This style is currently my favorite.

Some let you take a turn once a time period, if you miss a turn you miss a turn, some let you queue up moves in advance though, I haven't found any that require you to move or the game stops.  All the ones I've played like this were simultainious and yes it makes it either very easy to evade your opponent or nearly impossible to get away depending on how the mechanics are set up.  I've played a few like this and I don't really care for it.

I'll post some links when I get home, I've forgotten the URLs as I haven't played any of these for a long time. 

edit:  Here have fun:

http://dmoz.org/Games/Video_Games/Strategy/Turn-Based/Browser_Based/

[Margalis]

The X-Com/Fallout:Tactics/Mage Knight/Incubation style is that all players act in one turn, but with action points and initiative rolls determining the order of actions.

How does that scale to more players? One thing I want to avoid is a game with 4 people taking twice as long as a game with 2 people.

[HaemishM]

I'm not sure. Incubation had the entire team taking all of their turn for each player and it was hot seat, but I never saw that 4-players were all that bad off. I think if you alternate initiative by individual unit (whether that's an individual soldier, or a platoon/squad/whatever) so that you're generally intermingling players acting, it should scale fine.

You know, there's also FPS with Bullet Time type action, where everyone acts like a real-time FPS until they perform an action within LOS of another player or within a "Sphere of Action" such as within pistol range or something. Then the game slows down to a quasi-real-time bullet-time form of action for those involved in the action. I suppose that really depends on your game's Point of view as to whether it would work or not.