Game Mechanics

[TSMI]

So right now I am working on my timing system. I am going about it in a really, really ad-hoc way.

Is this going to bite me in the ass? Are there any decent guides out there to a good roguelike mechanic system? I am not just talking about a time system, but stuff that ties into it like combat.

[Quendus]

There are two potential problems with an ad-hoc timing system - first, if you really weren't thinking when you were making it then you might get the game slowing down with a large number of actors. I figure you probably won't have let that be a problem.

The bigger risk is that it might not be flexible. If you want features like changing an actor's speed (eg. through boots of speed), or especially changing an actors speed during another actor's turn (eg. through slowing/hasting effects), then you need to have a time system which can accommodate that. There's a nice algorithm on roguebasin which queues each actor's turn in a dictionary in order to avoid having to iterate over every actor - this system is especially problematic with slowing spells, because you have to dig into the turn schedule data and move an actor's turn whenever an external agent changes their speed!

Other potential issues are with timed effects - if you have a poison or burning status effect which does damage over time, do you want it to do a consistent amount of total damage to the victim? Do you want it to last for a constant length of game time, or should the speed of the damage scale with the actor's speed? What about confusion effects? Should a fast actor be confused for half as long as a normal actor, or should a fast actor suffer the effects for twice as many turns? Is it OK for some effects to scale with speed and some effects to remain constant?

These aren't trivial considerations - they can mean life or death for characters. http://angband.oook.cz/forum/showthread.php?t=5635

Still, if you don't intend to add complicated things like that then an ad-hoc system will work just fine.

[TSMI]

The system I am crafting now involves "fatigure points". So essentially actions cost different number of fatigure points, when you perform the action it gets added to your total, and every gametick your total is decremented. when it reaches 0 you can act again. works well so far.

But I guess I more mean stuff like stats. Do I just copy D&D like everyone else? I mean I've played it like twice. Right now I have some weird mixture of Warhammer and Morrowind...

[guest509]

  Making a stat system can be complex. It's best to just go with something, then tweak it over time. People can obsess over it for far too long I think. I really like the system in Demon Hunt. The stats are Attack, Defense, Speed and Magic. Also HP and Mana. Slots are Melee, Ranged and various passive 'Wear' slots like head, body, ring, etc...

  Each item in each slot alters the 4 basic stats. And that's it. The rest is balancing of gear and monsters.

http://www.demonhunt.org/

[Quendus]

When it comes to stats, do whatever you want. Some people like familiar stat systems like D&D, others complain about unoriginality; some people think the best stat systems are carefully balanced to make stable gameplay, others think the best stat systems are the ones with easy formulae that let the player know exactly how likely a given action is to succeed and how much benefit they'll get from an item (without getting out a calculator).

Like Jo said - no matter whether you use an original or unoriginal stat system, don't get hung up on it. Figure out what you want from your stat system and go from there.

As for the fatigue points, that's a pretty standard tried-and-tested system. Normally it's the other way round, with points called energy and an actor's energy total incrementing each turn. Same difference. Certain optimisations of that system (such as calculating in advance which tick an actor will take their next turn) can lead to the problems described in the last post.

[guest509]

  Looking outside the genre can be valuable as well.

  Turn based strategy games tend to have a point system as well as a turn passing system. Think XCOM, the original. You have so many action points, once you use them up you are frozen, pass the turn and the bad guys get their go. I like this a lot, very discrete. Sometimes with complex timing in a roguelike it seems like I sorta get extra turns sometimes.

  Another way to do it is how wargames do it. You can move and take an action each turn. You can move a certain amount depending on your speed, movement mode (flying, wheels, etc) and terrain (flying ignores terrain, wheels take extra points to move over rough, etc). Before or after moving you can take an action. Drink, attack, that sort of thing. Generally attacking uses up your movepoints as well. Attacking without moving can get a bonus, depending. Nearly every board wargame handles the turn this way.

  Some tabletop  RPGs also do the Move + Action method as well. It's easy peasy and everyone gets it. No record keeping.

  So there are just tons of tried and true methods. I just played through the new XCOM. It's a TBS game but uses wargame turn flow, I really liked it. It frees up the mind from worrying about counting points and lets you strategize more. Opinions differ.

[TSMI]

So it seems like my timing system strategy is sound and you can largely make up stats as you go along. Good to hear.

[requerent]

So it seems like my timing system strategy is sound and you can largely make up stats as you go along. Good to hear.

You might find yourself facing some pretty funky time economy if you invest a lot of arbitrary differences into ticks. Making a distinction between Turns and Ticks isn't always helpful-- you run the risk quantising Ticks in a manner that is confusing to the player. Although the difference between that and Fatigue Points is trivial; the paradigm, I think, is important. Design it so that every Turn is explicitly meaningful to the player. The measuring stick will always be the cost of the most basic actions-- usually Movement and Attack.

The selection of an action, in regards to how much time it will take, is always relative to the comparative value of the most basic actions. It's the Fiat of time. A spell that takes 3 turns to cast has an opportunity cost of 3 attacks-- the significance is easy to visualize. If you get too liberal with fatigue points, however, the player will have difficulty conceptualizing how meaningful each action is relative to time. It's simple in a real-time game, but in a turn-based game it can get confusing. Extra actions pop up in a manner that either demand too much extra thought from the player or they get lazy and randomly die.

If you do go the Fatigue Point route, KISS.

[guest509]

  If I remember correctly, Rogue timing was just an I-go-U-go sort of thing, but some monsters would go twice (like bats) and if you were hasted you could go twice too. Did some monsters go only every other turn?

[kraflab]

Previous discussions of timing that I recall:
http://roguetemple.com/forums/index.php?topic=2813.0
http://roguetemple.com/forums/index.php?topic=2493.0

As for myself, I just used a "the player goes, everyone else goes" loop in Epilogue, with modifiers just either making you lose a turn or gain an extra turn.