Dissecting turns

Posted in think

When we write games we are often forced to create this arbitrary division of time called “a turn”. A turn is a chance for everyone who is allowed to impact the scene to have their chance to impact the scene. There are three interesting ways to manage the turn (though an infinite number of variations). I’m going to think out loud about these now because I have three new game projects on the table at the moment and Diaspora uses two of these structures while my favourite game I didn’t write, Reign, uses the third.

The most common turn system is to derive an order and then poll each player in order and determine first if they can act and then second what their action will be. Their action is resolved and we move to the next player in order. This system requires that a mechanism for “initiative” be established, even if this is just arbitrary (referee picks the starting player and we go clockwise). This lacks “realism”, obviously, but also arguably embodies it (for pretty much the same reason it lacks it) given a good initiative system. That tells me it’s actually just fine.

A variant we use in the Diaspora space combat system is to invert this. For each possible type of action, request actors. Whoever declares they are performing the action get to do so, resolving in the order that they declare. This de-mechanizes initiative and makes it a function of player skill (do I act or react?) rather than character skill. For example, our caller might declare, “Beam combat phase. Are any vessels firing beam systems?” Firing beams first may destroy an enemy early, but not firing saves them for defensive use in the next round. There is a tension to who declares when that is fun.

Finally there is (maybe obviously) simultaneous action. Actions are planned, dice are rolled, and results for the whole table calculated. Unfortunately, this is not merely a procedural gimmick like the difference between the first two, but rather relies on deeper integration with the resolution mechanism. ORE is a great example — everyone says what they want to do with their dice, everyone rolls their dice, and the dice determine what the order of resolution is and what the results per resolution are, including impact on following actions in the same “simultaneous” round. The least elegant (though hugely fun!) version of this are games that follow more traditional sequential action turn construction but embed it in a “planning” phase — everyone writes down exactly what they intend to do and reveal all at once, and the turn is resolved by the referee. This was a common sight in may wargames back when I was playing them a lot, most memorably in GDW’s Harpoon, where the players may never even share a common map (the referee plotting real positions secretly while the players acquired a graph paper map full of detection symbols and position references).

In Diaspora we refused to choose one. We used two. I’m not sure that Fate lends itself to the third, though I note that FUDGE does, so there’s something to be learned in there. I guess in order to proceed we need to think about what these things actually deliver.

Per actor: traditional form, well understood (modeled by chess, for example). This would seem to be at the heart of a really “traditional” feel for a game.

Per action: social tension between players. Less traditional more genuinely emotional. Good for a highly competitive feel?

Simultaneous: great sense of verisimilitude. Leverages clever dice systems that can involve multiple axes of information as well as move dice into and out of the result.

Sorry this post doesn’t come to conclusions — it’s really an ongoing thought. And so I invite commentary with the understanding that this is not a completed idea for debate but rather a kernel of thought for exploration. Can we list the cool things about each so we can find a way to match game design intentions with each of these systems?

–BMurray

Posted by halfjack   @   30 November 2009

Related Posts

Like this post? Share it!

RSS Digg Twitter StumbleUpon Delicious Technorati Facebook

8 Comments

Comments
Nov 30, 2009
20:28
#1 Lon Sarver :

So why do you think FATE couldn’t use a simultaneous resolution system. I’m going to try one in an implementation for replacing D&D 4e soon. I’ll probably be letting folks know how it works over on the FATE list.

Nov 30, 2009
21:07
#2 halfjack :

I think it can, Lon, in fact I was playing around with it this afternoon stealing some ORE ideas. It just doesn’t, as written.

I was thinking, declare (or wrote down) actions, all roll. Highest roll resolves and reduces his target’s roll by one. Fate points apply per resolution, not per roll. I think. Maybe declare after the rolls even. Hmm. Not sure what game this suits, but it’s interesting.

Dec 1, 2009
07:34
#3 Toph :

I’m not sure the advantage of the ORE system is verisimilitude (or if it is, it doesn’t come because of simultaneous actions). Its elegance for me emerges because of the act of interpretation that follows all the rolls: dice hit the table, but we still don’t know what happens until the table narrates what happens. I might roll a great kick to your shins, but a gobble die from one of your allies means that I don’t in fact kick you. That’s a rich storytelling mechanism.

Compare this with the social initiative for Diaspora’s space combat, where the decision to act is the determinative feature: every act is the result of a conscious choice, and there are costs to acting first. I love this structure, because, when combat can be separated into discrete actions in this way, the result remains very tense, like submarine combat, as we have said.

So ORE takes mechanics that lead to an interpretation game, and Space Combat takes mechanics that lead to a decision game. Ore gives a flurry of actions that resolve into focus yielding perhaps only a few that do anything, and Space Combat makes every action significant and with consequences, even if it doesn’t hit.

Both of these offer me something that “my turn is up” doesn’t, where the mechanics provide a regular chance to act.

There was a game back in the 80s (you’ll remember which one) where one’s initiative determined how many times you could attack in a given hundred-unit period. If you had initiative 5, you acted on 80, 60, 40, and 20 (and zero? I forget), but if you had Init 3, you acted on 66 and 33 (and zero?) in any case, the first character would get a turn (80), then the second (66) then the first twice (60, 40), then the second (33), then the first again (20). (Something like that; it’s been almost 25 years of memories I’m sifting through here). In any case, that kept things “per actor” but still provided a material reward deriving from mechanics that (arguably) going around in initiative order (and less so, simply going clockwise) does not.

Dec 1, 2009
07:41
#4 halfjack :

Toph, “Car Wars” ran that way and around the same time “Starfleet Battles” also used a similar system. More recently the White Wolf product, “Exalted” also does this. I see it as a kind of branch on per-actor as well — more a specialized initiative system — but a divergent enough branch to warrant separate consideration. Thanks for reminding me of these!

Dec 1, 2009
08:19
#5 Chris A. (CodexArcanum) :

After playing Diaspora (and being a huge, huge ORE fan) I was immediately taken to the idea that it could be run simultaneously (talking about personal or social combat here). The resolution mechanic is already fairly simple: roll and figure result, compare to opponent’s result for margin of result, apply the margin as either stress, spin, or to resolve a target number. If you rolled a defense, you keep it on deck until your next action.

So the seeds are there, it’s just ordering it I think. As I picture it, you’d poll the table for actions (using whatever ordering mechanism, maybe high-stat-order Awareness then Agility) and have everyone roll at once. Record your “action value” for that roll, then roll defenses if anyone was targeting you. So it’s a bit of a reverse, action stands and defense rolls. Defense can stand too though, once it gets rolled.

That just leaves resolution order, since we already have declaration order. We could use stats,but as Toph says, emergent narratives are what make it fun. So maybe we can pull a second vector of information out of the dice? I’m thinking you look at just signs one set at a time. Most pluses goes first, then most blanks. You use whatever you have more of, and pick the higher if you tie. That means a person with 0 result (++–) would go before a person with +2 (++00) but he would go before a person with -1 (000-). If someone else had +2 (+++-) then they would go first.

That gives a pretty good ordering that is not completely determined by just the roll’s total, but does reward better rolls.

I’m not 100% sure if you need the ORE’s flinch rules or not. That was put into Godlike as a simulation element, and just kind of stuck around as a useful work-around to not getting a defense roll. It means a good attack will protect you and keeps the system more active. If you allow a defense roll, it’s not needed. If you don’t allow a defense roll, and do flinches instead… I don’t know. It could work with FATE. It makes spin a little more likely for defense actions, and make damage a little more potent for attacks.

If a person doesn’t roll defense, do they just have an automatic -4 + Stat as a target for enemies to hit? Those are interesting questions for a play test I think.

Dec 1, 2009
08:55
#6 halfjack :

Thanks, Chris — these are all things I’ve been playing with at the table over the past couple of nights. I’m liking, so far:

1. Declare actions
2. Roll
3. Highest roll (not total!) resolves her declared action, using extant rolls to be interpreted as defensive actions (in which case they are note degraded by flinch but cannot now be used offensively) or no defense (keep the roll, but using the offensive skill, and degrade it regardless of success or failure). Add fate and spin as desired but remove spent tokens after resolution.
4. Next highest resolves with her extant roll, possibly degraded, unless she used it defensively once already. Add fate and spin as desired but remove spent tokens after resolution.
5. until everyone’s had a whack at it

So you have tactical choices to make — do I change my action to a defensive one to take advantage of it better? Do I have a lot of attackers and how does that change things? Do I have a lot of allies? Is my roll good enough to still win when degraded?

As you point out, it needs play at the table to be sure. But it smells good.

Dec 1, 2009
11:40
#7 Lon Sarver :

@Brad:
No, it wouldn’t work as stands, because the “attack roll, defense roll” structure would make simultaneous resolution take forever. Here’s what I did:

Everybody declares actions. Compels are offered. Dice roll. Invocations and tags can be made at this point.

Purely defensive actions take effect immediately, because these rolls set target numbers for other actions.

All other actions are resolved in descending order of shifts, from most shifts to least. More invocations and tags can be made at this point, if people are willing to spend the Fate points. Characters engaged in actions directly related to the main conflict defend at the level of their most relevant skill. Characters engaged in other actions defend at 0. Successful attacks disrupt the action of the defender. Anyone who is attacked can elect to abort their action and roll a defense, instead.

Any necessary bookkeeping is done last.

Repeat as needed.

I adapted this from Ron Edwards’ Sorcerer, a game I like a great deal. It worked fine there, and I’m hoping my FATE implementation of the idea will work as well.

I’m hoping this will give a more organic feel to conflicts in play, and more of the sense of the unexpected shifts that happen in real time. I also find that reducing the number of die rolls speeds up the resolution of the conflict, and makes the mechanics less intrusive.

Can’t wait to play it and see if I’m right.

Lon

Dec 4, 2009
16:23
#8 dr_mitch :

I think the second edition of FATE used simultaneous actions as the standard method of resolution.

Leave a Comment

Name

Email

Website

Previous Post
«
Next Post
»
Powered by Wordpress   |   Lunated designed by ZenVerse

Bad Behavior has blocked 84 access attempts in the last 7 days.