</i> Lately I've been trying to dust off some of my notes, and start working &#111;n a game design project again, and &#111;ne of the things that I've thinking a fair bit about lately is resource attrition. The game idea that I'm working &#111;n is essentially an old-school dungeon crawler with tactical turn-based combat. It's not really a roguelike, as there's a seperation between combat and non-combat screens. In that sense, it's probably more like Ultima 3 or 4 (from what little I played of them, many years ago, so it's possible I might have that mixed up) Ideally, the combat should have enough tactical depth enough to be interesting in and of itself, and I want resource management to be meaningful within the scope of a single battle. (I mean resources here in the sense of mana, energy, items, or anything else that can fund actions taken by characters in combat) It should be possible to run low &#111;n resources within a battle, or be temporarily tapped out at a bad time (say when a crucial healing spell is needed), and choosing where and when to spend these resources should be tactically meaningful, without encouraging excessive conservation. In many older or more traditional RPGs, resources are strictly finite in the scope of a single dungeon. You have so much mana when you set out (possibly restorable with a similarly finite number of items), and you don't get it back until you return to town or some other place to rest. On the plus side, it gives strategic relevance to encounters that are individually too weak to have a credible chance of defeating a rested party. Even if they can't outright kill you, they can force you to consume resources in order to defeat them. These battles become a form of longer-term resource management and attrition. Moreover, the fact that meaningful resources are steadily eroded as you delve deeper into a dungeon provides a sense of dramatic tension. I think that it can contribute significantly to a sense of danger, and of being far away from a safe haven. However, they also tend to encourage gameplay that is (in my opinion) less interesting. Resource management is often irrelevant within any single (non-boss) battle, since in order to have enough to last for many battles, you have much more than you could possibly use in just a couple. So if you're in an area sufficiently close to town (or another place to rest), you have pretty much free reign to cast your most expensive stuff with impunity. And if battles are generally balanced around the assumption that you CAN'T do that, they also tend to be trivialized. On the flip side, if you're in an area that does NOT have easy access to rest (such as exploring a dungeon), it often encourages a boring degree of conservatism. Since the player does not usually know the depth of a dungeon or the exact dangers they can be expected to face deeper within it, it's sensible to avoiding wasting something which might be essential to have later, even if using it now would speed things up. Frequently (if resources are finite enough in the game) this can result in magic-users avoiding using any magic at all unless absolutely necessary. I know this has been my own experience in some games, and I personally think it's rather boring if you're effectively encouraged to avoid using most of the interesting abilities of entire archetypes. Quite a few newer games (especially MMOs, anyway), have effectively switched this around, making resources finite in the short-term, but infinite in the long-term. This is often done through making the maximum pool of a resource &#111;nly large enough to use a couple abilities before being emptied, but regenerate very quickly. I think this has very positive ramifications for tactical depth within a single combat. You're encouraged to actively use abilities, and short-term resource decisions can be meaningful (eg: if I cast X now, I won't have enough mana to cast Y for a couple more turns. Will I need it in that length of time?). Unfortunately, it also means that unless a given encounter has a non-trivial chance of wiping the group out, it becomes rather meaningless. You'll just recover whatever resources you used during the fight shortly after it concludes. Battles which are not a credible threat, and have no long-term consequences are tantamount to time-wasters, really. You could, of course, make sure that EVERY battle was designed to be difficult enough to pose a challenge individually, but I believe that, a game's difficulty curve should be punctuated with challenging encounters, rather than consist of entirely of challenging encounters. In fact, I think it could be rather exhausting of the player. I believe that systems without long-term resource management also lack most of that sense of dramatic tension you'd get from being deep in dangerous territory. After all, even after slogging through hordes of monsters to reach the deepest parts of a dungeon, you're not appreciably more exhausted or weakened than you were when you started. I'm not sure you could even replicate quite the same sense of being tired and vulnerable and far from home that older-school games tended to provide. Maybe not everyone felt that way, but personally, this is something I'd like to be able to capture in my own game. In essence, I want to make both short-term and long-term resource management meaningful in my game, despite these often seeming contradictory in many existing games that I've looked at. How can you make mana finite enough to be exhaustible (and regenerable) within a single battle, and yet slowly exhaustible over many battles? I've brainstormed a number of ways that my game's resource system could be constructed to attempt to facilitate some of these aims, but they all suffer from various different problems. I'm not going to go over all of them here (although I suppose I could, later, if anyone was interested), but here's &#111;ne of the more simple and 'obvious' &#111;nes: A caster would have enough mana that it would be possible to exhaust themselves in a taxing battle, or if their mana was poorly managed, but more than enough to use a range of spells in an average combat. After each battle, a certain amount of mana would be instantly recovered (perhaps proportional to the difficulty of the battle, and affected by equipment or abilities that provided regeneration bonuses) The goal of this is that you would be free to use a reasonable number of spells without worrying about wasting mana (since you'd get it back after the battle was over), but that it would still be possible to win a battle with a net loss of resources, and that this drain could be cumulative. (ie: each battle you'd end up with a little less mana than the battle before it). Also, it would technically be possible to end battles with a net gain of resources, so this isn't just a death spiral. Unfortunately, if mana was to be used as the limiting resource in this way, it would also seem necessary to disallow mana regeneration in-battle, through any reusable source, be it through passive natural regeneration, abilities which could convert health into mana or generate mana in some other way. Because, if you could regenerate mana while in-battle, but not out of it, it should be quite possible to 'stall' for time in-battle to regenerate an arbitrary amount of it. For example, by leaving a single weakened enemy alive, who doesn't pose much threat individually. This actually sort of reminds me of the situation in Final Fantasy: Tactics (which I consider a great game, actually) where &#111;ne of the most effective ways to gain experience points and JP (used to purchase abilities) was to leave a single enemy alive, and then have your characters either hit each other with wimpy attacks, or spam free support abilities. I think 'generating mana through abilities' is a rather large design space to restrict &#111;neself from entirely, and I'd really rather not do that. But I decidedly do not want to create any obvious ways to gain tangible advantage through tedium. So…. I was basically just wondering what some of your opinions are &#111;n the subject of resource management and attrition in RPGs (or even more generally). It doesn't even have to concern my own problem, specifically. If you play these sorts of games, how do you feel about systems with/without significant attrition? Are there any games that you think have done it unusually well?

[For the more complete version, see above post]

Many traditional RPGs have finite non-regenerating resources that are restored only at particular points (such as inns, etc.). Resource management is of long-term relevance, but rarely relevant within a single regular combat. This allows for less challenging battles to have meaningful consequences (consumption of resources), but often encourages a (subjectively) boring degree of resource conservation (eg: magic-users who avoid using any magic at all if they can help it)

In RPGs with rapidly regenerating (but smaller) resource pools, resource management is of short-term relevance only. This may allow for greater tactical depth within a single combat since the player can be expected to use a range of abilities, rather than conserve them, and running dry during a single fight is plausible and must be managed. However, if resources regenerate between battles, any fight which does not present a reasonable chance of outright defeating the player is effectively just taking up time. Attrition is not possible. Moreover, I think you lose a significant degree of dramatic tension that can come from being deep in dangerous territory, and running low on resources. In such a system, no matter how many battles you slog through, you're still effectively as fresh as when you set out.

Is it possible to design an RPG where resource management is relevant both in the short-term and long-term? Where excessive conservatism is not encouraged, but moderate-difficulty battles are not made irrelevant?

What are some of your opinions are on the subject of resource management and attrition in RPGs (or even more generally)? How do you feel about systems with/without significant attrition? Are there any games that you think have done it unusually well?

I suggest that the interesting part of dungeon crawling is the enjoyment of long-haul resource management and constant, steady progress. In your example of long-haul resource management, you made the suggestion that the developer could provide a finite amount of resources during a delve into the dungeon, which can then be replenished when the player enters the next town. You suggest, however, that this type of game encourages conservation which you deem is boring or bad gameplay. I disagree that conservation is unfun. I also disagree that the long-haul resource manage necessarily implies conservation.

As a developer, you can mitigate the player's desire to conserve by placing restocking points at regular, anticipated intervals. The truly conservative player will play conservatively no matter what restocking you give them, so ignore this type of player for your design. An average player, however, will quickly learn to trust that you will supply them with a marginal amount of resources in each area. Take, for example, Half-Life. The only resources that are managed in Half-Life are Health, Armor (another form of health, really), and Ammo. The level designers also placed copious amounts of health/armor rechargers in each level. Ammunition for various weapons was also available in adequate amounts. While I might not always be able to have an amount of ammo for any particular weapon that I would want, I always had some ammo to keep me going. For example, I might run out of rockets or SMG bullets, and have to switch some other weapon. (Note to FPS designers: I think you do your game a disservice by consolidating ammunition in this way. Different ammo types enhance the resource management aspect of your game!) Looking back at some older dungeon crawling rpgs, you will notice that the map designers placed treasure chests that often contained potions and restoratives to supplement what the player was expected to have consumed to reach that point. My favorite example of restocking points was in Dungeon & Dragons Online. True to D&D, casters could only cast a finite number of spells before resting. Casters could only rest at certain 'safe' places. The number of safe places in each level was limited, and once used, could no longer be used. This essentially gave the casters a maximum number of spells per level, which required the party to act in a very efficient manner.

I suppose what I am suggesting is that you divide your levels up into sections between restocking points. The suspense of the level is in getting through the whole thing intact. The immediate goal is to reach the next restock(/save) point, where you can recover some of the resources that have been lost to attrition. Players don't need to know the length or difficulty of the dungeon to play regularly. They only need to know that that they can trust you, the designer, to have placed enough restocking points to make the level challenging, yet fun.

Interesting topic. I'd love to see discussion about this, but here are my two cents.

If I remember correctly, Star Wars Galaxies (at least at the time I played it) had a system of wounds and damage, where wounds would decrease your stats, which determined the available amount of health. Health would then be effected by damage. Wounds would only be healable in town, or by a doctor whose ability was limited by consumable items. This might help with your mana ideas.

You didn't mention the positive aspects of "trivial" encounters. In typical RPGs, you get loot and experience/skill advances, which make it justified to spend your resources on them.

Personally, I think one of the key issues is making all the involved resources relevant. For example, I loved the Baldur's Gate games, but never really bothered with potions (collecting or using), because it was just too tedious to use my time on, and I kind of figured that they'd have to balance the game around a player not having any; this means that in those games, this particular resource wasn't relevant (they had to include it because it was D&D though). If you want to discourage over the top conservatism, make it necessary or just overly useful to e.g. use magic on a particular monster. A more extreme example would be to base the magic system on creature vulnerabilities and potions that allow a mage to strike a creature where it's vulnerable. Often (i.e. in D&D), the situation is that the opponent only has a single pool of hit points you have to overcome, and the mage differs from the fighter only in the average effectiveness of their attacks, and the scarcity affecting their use. In pen-and-paper RPGs, this is often to cut down record keeping, but there is no reason to make such a monotonous system using the computer.

If you want to make a good computer RPG, take the ideas from traditional games, not the implementation.

Quote:

Original post by Zarion I've brainstormed a number of ways that my game's resource system could be constructed to attempt to facilitate some of these aims, but they all suffer from various different problems. I'm not going to go over all of them here (although I suppose I could, later, if anyone was interested), but here's one of the more simple and 'obvious' ones: A caster would have enough mana that it would be possible to exhaust themselves in a taxing battle, or if their mana was poorly managed, but more than enough to use a range of spells in an average combat. After each battle, a certain amount of mana would be instantly recovered (perhaps proportional to the difficulty of the battle, and affected by equipment or abilities that provided regeneration bonuses) ... Unfortunately, if mana was to be used as the limiting resource in this way, it would also seem necessary to disallow mana regeneration in-battle, through any reusable source, be it through passive natural regeneration, abilities which could convert health into mana or generate mana in some other way. Because, if you could regenerate mana while in-battle, but not out of it, it should be quite possible to 'stall' for time in-battle to regenerate an arbitrary amount of it. For example, by leaving a single weakened enemy alive, who doesn't pose much threat individually.

If you could also regenerate mana by idling on the map (out of battle), then why wouldn't the player choose to idle on the map rather than in the near end of a battle where they must take constant damage from remaining enemies? If you are able to regenerate mana by idling on the map as well as in battle, then the problem you are describing here wouldn't be a problem at all, except that the flow of gameplay would be riddled with constant idling breaks (especially bad for the conservative player). The use of potions and your end-of-combat mana recovery mechanic however, sort of combats that.

Another way to persuade players to resolve battles as fast as possible is to make the amount of mana that is restored at the end of the battle be dependent on how fast they resolved the battle – quick encounters will restore more of your mana, while long, drawn-out battles will be more draining. This will also persuade players to use more of their higher-level skills/abilities to quash their opponents quickly.

[Edited by - Tangireon on April 30, 2009 9:20:29 AM]

It strikes me that being overly conservative in long-haul dungeons is largely because you've no idea what's else you're going to have to fight. You don't want to be using your best spells on an enemy that's moderately hard just in case his bigger, meaner brother is hiding around the corner.

If you knew beforehand that there was 10 kobalds, 10 orcs and one humongous dragon before stepping into the dungeon, you'd be much better placed to plan out your resource usage. You wouldn't find yourself either completely depleted of your good stuff for the big enemies and similarly you wouldn't complete the dungeon with all your big spells still held in reserve.

I'm not sure how you'd integrate that into a game though. Just telling the player before they go in would seem to be a little fake and immersion breaking. Perhaps you could gather intelligence on the dungeon by quizing local NPCs (and collate that into an in-game notebook which crossed off enemies as you find them). This would only occur for major or semi-major enemies, grunt enemies which shouldn't be much trouble wouldn't be included so you retain some sense of danger and unpredictability.

Another option could be to give the ability to sense creatures in the same region as your party. Almost like infrared goggles you would have enemies displayed as vague heat traces on a map, showing you both the density but also how tough they are likely to be but without explicitly telling you what enemy type they actually are. Again, smaller grunt enemies would be excluded so you don't have perfect foresight and can't just walk around with your pants around your ankles.

This is interesting topic and I've been thinking about the same goals. Generally I lean towards the long term conservation type designs, for example the D&D games with restricted resting. Although I think they do require quite careful level design to balance them properly, particularly when there will be large variation in the players party build by the time they encounter the dungeon. The Neverwinter Nights games are examples of games that barfed this a bit, and made up for careless level design by letting you rest anywhere which took away the challenge. (Or at least left determining the level of challenge too much to the players discretion.)

One suggestion I have is to divide your spells into two types. One type is more powerful, but requires preparation of the incantation and perhaps the reagents, this preparation is assumed to be done automatically while the player rests. The other type is more about the immediate manipulation of energy. So the latter type depends on a mana bar with recovers in a short period, and the first type is based on once per day type usage.

Another suggestion would be to effectively have two mana bars. The first one has a short recovery time and is a limiting factor for all spells. The second you'd need a different name for, perhaps something like fatigue or concentration. Your concentration bar depletes slowly, but as is gets low it starts to decrease your rate of mana regeneration and also increases the chance of a high level spell fizzling out on you. So you can still attempt a high level spell when heavily fatigued but with some risk. High level spells would have more effect on your fatigue bar, so as a wizard gains levels he becomes able to use progressively more spells routinely.

Definitely agree with making scouting and reconnaissance useful in dungeons, with spells like invisibility, wizards eye etc helping. You also make so that characters from certain classes or with a high Lore skill can discern more facts about creatures and surroundings. (For example what their resistances are.)

Then I'd be classes that didn't have resource expenditure. I mean, if my character is useless (read: dead) after running out of arrows/mana then I'd avoid even coming close to being out of arrows/mana. It's less of a problem when you're controlling a party (since the tanks can often delay enough to flee) though.

Honestly though, going back to the inn/town or sitting around waiting for mana is not fun. 'Dramatic tension' as you get closer to death is not fun for me.

If you're making a dungeon crawler, then you want the player to dungeon crawl. Anything that pushes them towards 'recovery' and not explore/hack/slash is perhaps against the core game ideal.

I've thought about this a bit for a top-down action-RPG/shoot-em-up hybrid. My conclusion was that ammo type mechanics (i.e. a finite supply of something you could only replenish outside levels) offer nothing of real value, for pretty much the same reasons you have listed. Particularly in an action game (which doesn't really apply in your case), I think the inevitable conservatism would essentially just deny the player of extra toys with which to blow **** up (this, in my case, being a very fundamental part of the appeal).

A big issue for me is to make attacks non-spammable, even though they are infinite in supply. For instance, guns would have a finite clip-size to force the player to occasionally reload, and special weapons would have cooldowns (like, you can throw a grenade every 5 seconds and fire a missile salvo every 20 seconds). You could use the same idea in a turn based game. In other words, instead of having a Fireball cost 5 points from some finite supply, give it a 2 turn cooldown, to force the player to use a selection of moves rather than whatever is the most powerful (and hence avoid some of the monotony). If you have different rotation options that benefit shorter or longer fights (like one with instant damage attacks, and another including damage-over-times and buffs/debuffs) then you can make the boss fights feel fundamentally different from the regular fights (and even retain a sense of "pulling out all the stops" for the big fights).

I don't think difficulty is much of an issue; killing a boss is supposed to be harder than killing his henchmen. Let the player have a bit of a breather while he's making his way there.