Sunday 6 July 2008

More Hit Points

Shortly after I argued for hitpoints as the best way to represent damage because they makes the game shorter, Craig Perko of ProjectPerko has argued strongly against them because they make the game longer.

Hit points are one of those 'trendy whipping boys' of game concepts, like character classes and cut scenes that periodically are trotted out as an example of bad game design. But, like these other examples, hit points are the best mechanism that we currently have to solve a difficult design problem: in this instance to represent the process of damage. The alternatives suffer from over complexity and design weaknesses such as negative feedback loops (See my original article for further discussion.) [Edit: As Craig points out in the comments, I'm actually talking about positive feedback loops with negative consequences]

I don't see anything compelling in Craig's argument against hit points. I guess as a roguelike designer, I'm used to random damage, which seems to be the crux of his problem with them. Having hit points as a resource to manage is a strength, not a weakness, and even a card based game system such as Magic: The Gathering uses them to model damage.

Instead, Craig is using hit points as a straw man to address more fundamental problems around combat design and grind. I have no problem with trying something different with combat design but trying to call out hit points as the fundamental problem with current game systems is the wrong way to go about it.

13 comments:

Jonathan Stickles Fox said...

Dwarf Fortress's adventure mode is a nice sample of a complex combat system that mostly avoids hit points. I say "mostly" because behind the scenes it has "blood", and when the creature runs out of blood it dies. I have a pretty good guess how it works, because I've maintained Liberal Crime Squad, a previous game by bay12games, and in LCS blood is basically a hidden hit point mechanic. But on top of it is a complex system of body organs that can be punctured, limbs that can be severed, teeth that can be knocked out, eyes that can be gouged, and so on. In action, combat is much more exciting than a system in which the player manages hit points. Running at a wolf and swinging your mace only to have its left leg fly ten meters into a tree and it starts staggering around bleeding to death *is* fun.

The main problem is obvious if you play a weak adventurer, however: you're susceptible to the same negative feedback loop that the wolf is. It can latch onto you and start tearing you apart, and soon you're passing out from pain before you can stagger to safety.

I see two obvious ways to try to take the good of this damage system without making the game extremely hard. The first is to try to improve the survivability of the main character somehow, whether it's by giving them extra chances (maybe side hit point pools that have to be expended before a status effect is incurred), giving them resets, or even immunity to certain debilitating effects. The second is to simply add more heroes. Liberal Crime Squad uses a combination -- it has a very lethal combat system where one good hit to the head or chest can kill instantly, but with a squad of up to six people and the ability to train your people to be more evasive than your enemies, it's possible to still become very strong with a low probability of serious damage to your people. And when the enemy does cripple one of your good people, they're more likely to not be dead outright, and you'll be able to use your other characters to get them off the scene and return to base to get them patched up.

One valuable lesson you can take from both Dwarf Fortress and Liberal Crime Squad is that permanent debilitating damage to characters is extremely good at telling emergent stories. When my character is sliced in the face and loses his left eye, that shakes up gameplay, but it also tells a story. Battle scars are memorable in ways that hit points can't be.

Andrew Doull said...

Thanks for the detailed response: I was aware that Tarn was trying to get away from using hit points and the fact that there's a hidden mechanic that still mimics them nicely supports my argument that it is a solid design.

You've also pointed out the obvious workarounds for the negative feedback loop, which is that the individual dwarves, LCS members are not important (but still interesting) in the scheme of things. The fact that DF comes from a roguelike tradition of permadeath, and Tarn Adams has publicly stated that he wants to make losing fun, make negative feedback loops of a critical damage system much more tolerable as well. In the long run, DF is intended to design interesting worlds e.g. the ultimate sandbox, and in that scheme of things, horrific injury is much more interesting than a arbitrary number.

Brog said...

The problem with hit points is that one tends to use them without thinking, because "that's how damage works".
You are correct that in most cases it is the best solution we have; this doesn't mean it's the best in every case and should be used blindly. Game designers should carefully consider why they're doing things (I know from your previous articles that you do think about this, so I'm not trying to convince you of anything, just sharing some of my thoughts).

Consider Darwinia/Multiwinia. I think each Darwinian has a small hp count, but they don't really need to - there would be no fundamental difference in play if they were destroyed by a single hit. It makes more sense to think in terms of groups of Darwinians rather than the individuals; if you do this, the "hp" of a group is the number of members it has. There are "status effects" as the "hp" goes down - a smaller group has less firepower - but this is limited somewhat because only the units near the perimeter can attack the enemy.

More abstractly, if there's another value which does the job of hp (that is, you lose/die/suffer when it is reduced to zero), then it's more elegant to just use that than to tack on an hp system as well. Some possible values are: amount of territory, number of units, mass of sentient goo..

So:
- most of the time hp is the correct solution.
- hp (or any other game mechanic) should not be inserted blindly into a game; think about what you are trying to achieve.
- alternative implementations of hp (or emergent properties equivalent to hp) can lead to interesting games.

Nick said...

I also like hitpoints. I like my abstractions good and abstract. If I want reality I have - well, reality.

Mikolaj said...

I think this is also related to styles of play . E.g. HP naturally enables the perfectionist style of play and its variants. Resets with systems where you lose limbs are more problematic, may give less satisfaction to a perfectionist that likes to design perfect tactics and see them 100% successfull or die. Going ahead scared for life is neither...

Craig Perko said...

I was specifically saying that hit points are uncinematic...

Again, my post wasn't thinking about Roguelikes: my post was thinking about all those games that use quicktime events, such as God of War. They're slowly stepping away from HP as it is, I'm pointing out that they could just go further in that direction, if only they could make the gameplay of them interesting.

HP is very effective in terms of gameplay, but it's pretty dull and unconvincing in terms of drama and immersion. In a roguelike, where numbers are king, it wouldn't make too much sense to put numbers aside in favor of something that doesn't fit into the genre at all.

Some addition of deeper complexities is possible, but without a lot of very clear feedback, I'm not sure it would add much. It certainly wouldn't add the sort of directed drama I'm talking about. What I'm talking about really isn't suited for a roguelike.

I'm not sure whether you're using "negative feedback loop" wrong, as a negative feedback loop tends to return towards the middle: a positive feedback loop means that whoever's winning wins more. This dramatic system is not innately weighted in either direction any more than an HP system is.

:P

Rodneylives said...

It's worth going back an examining how hit points became so common to begin with.

They were a core component of Dungeons & Dragons right from the start, of course. In the original version, all normal men had d6 hit points. Gaining a level gave you an additional d6. That was about it.

One thing about that game that is often forgotten now is that combat was extremely abstract. A combat round was officially a minute long! One didn't "swing at" an opponent, instead a character chose to attack an opponent, and the rolled-for to-hit and damage were understood (or were supposed to be at least) to be result of many factors. A successful attack was probably composed of many hits, reduced damage was the result of the enemy dodging and using armor, and so forth. And hit points themselves factored in being tired from taking blows, tactical advantage and everything else in battle in additional to physical wounds. Thus, a good referee wouldn't roll to see if a monster hit a critical spot, instead, hitting a critical spot would be the thing that happened when 0 HP was reached. The narrative would be adjusted to match the mechanics in this way.

For that, hit points worked pretty well. The problem came about when most people ignored that and took hit points as representing -actual- damage instead of an abstraction. For that, hit points are a fairly poor representation, but they've survived for so long because they're simple, both to implement and in play.

In the original article there are a couple of points I want to address:
- The impulse to make things "cinematic" is way too prevalent in gaming. It's rare that a game chooses to copy a movie worth copying. The same goes for immersion. I'd much more have an interesting game mechanic than a camera swishing around desperately trying to make something look cool.

- Hit points are not a resource, they're a measure. The difference is that the player -doesn't- control them, he doesn't expend hit points to do things, instead they are taken from him. No one wants to lose ANY hit points, but to do other things requires putting them at risk. They might be lost, but they might not. Rightfully, the resources are the things that allow the player to replenish hit points.

Robert said...

It would be interesting to play a roguelike where there are negative feedback loops to damage, and healing is by potion only, balanced by the fact that individually, the foes are all pushovers. Traditionally, if you win the encounter, in not so long you'll be as good as new. But if the focus of the game shifts to, don't let the rats wear you down, what properties emerge?

Soyweiser said...

But you do use hitpoints as a resource. Certain skills (for example the sacrifice attack of the diablo 2 paladin) have a cost in hit points. And there are other ways that you can use hitpoints as a resource. (walking around a monster instead of fighting it (which gives them a free swing at you) could also be seen as the usage of your hitpoints resource. They are taken from you by your the choices you make. The fact that they might be or not be lost does not change them in a measure. To elaborate on this, consider an item that can be used and has a 50% chance of breaking after use. Is this item not an resource?

A problem of with a system without hitpoints and a DF like combat system is that it is very easy do die horribly. A critical hit from a kobold who found a enchanted knife could end your adventuring career. While this can be exciting, it can also be nasty if your favourite character is killed by a lowly creature. It requires you to be on your toes a lot. And it could create situations in which you did not know death was imminent. (kobold finds enchanted crossbow of poking you in the eyeball with additional range enchantment, and you didn't see him yet). Normally in roguelikes death is quick and around every corner, but it is mostly fair and fun. Not being able to defend against instant kill attacks, that you cannot expect is not really fun. That is why I think we also have hit points.

But I do like Craig Perko's idea of the tension mechanics. Using that would remove my complaints about a DF like system.

It seems to me that a lot of people who write about games design have their pet peeves of things that they don't like (and think that if these concepts disappear games will be a lot better). Craig Perko has Hitpoints as a concept. Squidi has experience points and levels. I dislike systems based on the four elements (not that I write about games design).

Andrew Doull said...

Thanks everyone for the feedback - just goes to reinforce my theory that it's better to write half an article to get more people involved ;)

Craig: You're right. I'm misusing negative feedback horribly - I am in fact talking about a positive feedback loop with negative consequences.

Glass2099 said...
This comment has been removed by the author.
Glass2099 said...
This comment has been removed by the author.
Glass2099 said...

One idea that came across my mind while reading this forum is the adoption of "rules" in place of hit points, such as those used in chess and poker. From chess one can extract ways in which players may attack and defend, which would be more conducive to party play. Poker deals with a randomized set of situations from a neutral pool of finite situations, but how one chooses to use these situations can allow a trump effect over an opponent's decisions (metaphorically speaking). Furthermore, I always felt a bit cheap when playing the main FF series on the Gameboy Advance because of the attack command. I mean, come on, in real life and especially in Japanese anime/ manga there are so many ways to attack that can be imagined which are distinctly separate from special abilities and magic (though, I hate the calling out of techniques associated with these Japanese media). The point is that as my rule instead of HP bound characters would level up (by gaining actual experiences like being trained in the way of the sword, gaining new/ better equipment and items and abilities and magic, taking battle damage (subtracting from one's pool), healing (restoring options to one's pool), performing character enhancements in battle (temporary pool increase) or out of battle (permanent pool increase), and/or how many times a certain enemy or like enemy is fought to help determine the details and outcomes of future battles vs. gaining experience points, and the levels would be more like classes of power that clearly define a character as having a significantly minimal chance of being defeated by an enemy on the level below him/ her) they would gain more trump options in their random pools vs. the sizable amount of subservient options that characters begin with and more and different/ better ways to use these options while still leaving the decision up to the player.
Another mechanic I would want to omit is the rigid turn-base rules that most are familiar with in FF I through X-2. Turn-base is a very solid mechanic for RPGs, so I prefer it to fast-paced real-time like in FF12 or Tales of Symphonia. Also, the active time bar (with a pause button) was a fantastic innovation that I see as an abstraction of the god-like speed portrayed in anime battles with the bar representing a composite of one's reflexes and recovery time, except portrayed at a pace a player can engage in comfortably. Here are some specifics I would include: No action lists that include showing at what point an enemy or anyone outside of the party is going to attack because one has no control over non-party members (the exception being that a character uses a mind control ability, effectively making the target a temporary party member); button combos (ex: Sabin of FF6, ring system from Lost Odyssey) and unique strategies for defeating enemies (AWSOME ex: most boss strategies in the Legend of Zelda series); the ability for characters to be patient and strategize among one another (ie. the first three FFs made one use a top to bottom way of ordering characters, and FF 6 made one wait until the time bar was filled before even being able to issue an order), thus creating a battle where one choose what order characters attack in, and once a character has attacked having the decision of using them to begin a new party strategy or act independently (the defend option would be available as well as a skip-all option to act independently and a skip option so that one does not have to follow the same order of character use over and over again when setting up a party strategy); the defend option should make one use a shield if available (a possible block to a lesser attack if not) instead of being an unrelated command to the mostly random appearance of a shield to block an attack; spells of the like of FF's Haste should not only increase the fill speed of characters' time bars as it does in the later FFs but also increase the number of attacks one can perform as it originally did (though only according to the rule that normal speed(= 1 attack) x a positive whole number (by how much the speed increases, n)= Z (the n number of attacks)); and, finally, logical placement of items, especially on enemies, who should not normally (in the case of most non sentient monsters) have money on them, nor should shops always be allowed to sell military-grade weapons and spells of mass-destruction. Thank you for letting me explore these ideas that I have had for a while and helping me clarify them with your own knowledge.