Skip to main content

Why D&D has Hit Points per Level for the Acquisition of Gold

The hard-working Hit Point remains one of those aspects of D&D that its players continually debate. It's not just what hit points represent. It's also how they increase per level. This post is about the latter factor.

In D&D, the older a character is, and (in most cases) the more injured the character has been through trials, traps, and monster attacks, the more resistant to further attacks the character will become, because surviving bloody injuries implies success and, indirectly, leads to leveling up.

That's how D&D has always worked: characters who have been beaten within an inch of death come back not with permanent injuries, but even harder to kill because they level up afterwards, which is normally what happens when you survive in D&D. (Yes, I know of many house rules that introduce permanent injuries, but hit points still go up per level.)

It's as if serious injuries just make you healthier. Anybody who has spent time in a hospital knows otherwise.

In early D&D, it was weirder, and it still is for those who like old ways: the more you get gold coins, the harder you are to kill.

How did ballooning hit points come about?

A prolonged side-remark: defenses of inflating hit points and early alternatives

The standard reply to the objections to the lack of realism in hit points began soon after D&D's publication. Gary Gygax explained that hit points are partially metaphorical (AD&D Players Handbook, p. 34, 1978). They represent stamina but also, "at higher levels," other factors that go into avoiding injury. They represent a mixture of "skill, luck, and/or magical factors." A character's level gains give more hit points because higher-level characters are better at avoiding injury.

The same thing goes for gold giving more hit points: there are explanations on offer. Some people like those explanations, and others don't. The main argument in favor is that it's fun to motivate players to seek treasure (and I have no argument against fun). But there was another, earlier reason that acquiring gold converted, in effect, to hit points. That's what I'm discussing below.

Before getting into that, I want to say again that there are a lot of situations in D&D where the explanation for hit points as avoidance capacity doesn't work even within the fantasy verisimilitude, as I wrote about once before. If two characters are hit by surprise arrow attacks--hit mind, you, not avoiding the hits--why should the young and hardy 1st-level adventurer die immediately while the elderly, seventh-level wizard, who was hit by an equivalent attack, have no chance of dying? Neither one saw it coming. Stuff like that disrupts the game rationale. For Gygax, it was because hit points may represent "luck."

Most D&Ders are content to hand-wave those problems. Almost fifty years on, the hand-waving still works.

It has never satisfied everybody. It was one factor that led many early players to develop alternatives to D&D, starting in 1975. Hit points per level was a feature not replicated in most other early RPGs: Tunnels & Trolls, Boot Hill, Bunnies & Burrows, The Fantasy Trip, Traveller, RuneQuest... The list continues and it is long. These games all have numerical scores representing how much damage a character can take, and they include ways to increase that number somewhat, but you don't get hit points per level that increase so directly as in D&D.

Other rules were created to circumvent the effect that hit points had. Critical hits, for example, meant that a high-level character was not apparently immune to the first three sword blows.

The games that began as close adaptations of D&D (like Empire of the Petal Throne and the Palladium Fantasy game, the latter of which evolved out of Siembieda's house rules for D&D) retained leveling and increased hit points along with that. Such games are very much a minority of the systems out there.

As far as I've seen, there are very few role-playing games that lack rules for quantifying injuries numerically. Even early divergent games like Ars Magica (1987), with its descriptive levels of woundedness, still treat injuries as countable (numerical) steps that have numerical ramifications on the dice rolls which give numerical results. As long as dice are used to resolve uncertainty in RPGs, it's hard to imagine that a purely descriptive, unquantified system is possible. (This excludes freeform and diceless games, by definition.)

The rejection of the rule that says that treasure gives experience and power-ups was equally widespread. Even the current edition of D&D emphasizes only giving XP for defeating monsters, not for finding gold. The insistence that GP = XP, once standard, has become reactionary and "old-school." But the "old-school" gamers have not had to fight to retain hit dice because the latest versions of D&D still depend on it. I have argued that hit dice are the main feature that makes D&D what it is. So let's look at the origin of hit dice.

The roots of hit dice: military ranks

Most readers probably know that D&D's immediate antecedent was a set of rules for miniatures wargaming on a table to simulate "medieval" warfare called Chainmail (1971) and that a "fantasy supplement" included as an appendix to those rules gave guidelines for using magic and fantasy monsters. This was not the first set of rules to inject fantasy into the reenactment and simulation of historical warfare, but its impact would be the biggest.

In the Chainmail rules, miniature figures normally represented units of several men of the same type. Alternative rules for "Man-to-Man" fights were included by which each figure represented just one man, giving skirmish rules. There were also individual Heroes, with a quadruple fighting ability, and even Superheroes, "double heroes" who therefore had eight times the normal fighting ability of a man.

The rules for combat were modeled on earlier wargaming rules in which missiles (guns and artillery) played a large role. Attacks were taken in turn: alternating shots, as with artillery. You roll to hit and if the piece on the board was killed, it was disabled or removed. Then your opponent rolls to hit with the remaining pieces. D&D combat is a series of alternating shots because its unseen, submerged template is missile fire in miniatures battles. (There are other ways to do it.)

The use of individual miniatures with higher ranks who were harder to kill was modeled on games in which military ranks played a role. This was a factor in tabletop wargaming for many decades prior to the development of Chainmail, as elucidated by Jon Peterson in his magisterial Playing at the World. You don't need to digest that encyclopedic book to get the point, though. Think of the successful pawn that gets "promoted" in chess for crossing the board, or the way pieces in the board game Stratego are numbered by rank, always defeating the lower-rank pieces they meet. (See Peterson, PatW, 341-352).

The Hero miniatures in Chainmail, representing individuals, were harder to take out than ordinary fighting men. This simulated their heroism. A Hero, having the fighting ability of four men, had to get hit or "killed" four times in one turn to be killed actually. Otherwise, he shrugged it off.

The rare Superhero needed to get hit eight times in a turn to be killed.

The idea that certain special individuals needed to be hit more than once in rapid succession to get killed is the root of D&D character levels, hit dice, and hit points all at once.

Granular "hit points" to increase uncertainty

Arneson's Blackmoor campaign used house rules based partly on Chainmail. His campaign was the petri dish for D&D. It was the game which first included dungeon adventures, as far as anybody has discovered. The dungeon setting took those wargame rules to a smaller scale. Already it had been popular among wargamers to use a 1:1 ratio for miniature figures (one miniature = one man). Similarly, in the dungeon each individual fighting man was tracked with one miniature, each with its own stats tracked on a separate card or paper.

Instead of letting each fighting man get taken out with one successful hit, however, the innovation was to scale those individual hits into 1-6 fractional points: hit points. This innovation, modeled on rules for naval battles in which sea vessels could absorb a fixed quantity of damage before sinking (as shown by Jon Peterson, PatW, p. 337), created greater individuation for characters on this smaller, man-to-man scale and also made the outcome of attacks variable. You might not lose a fighting-man character to any single hit, which would have been assumed with miniatures battles. Luck with the dice became a bigger factor. More dice rolls led to the feeling of higher stakes in individual personas. The playful narration of the outcomes of blows that did 1 damage versus blows that did 6 damage became more exciting, something for the referee to relish and to entertain the players.

An ordinary character's capacity for taking hits was scaled 1-6, and attacks did damage on the same scale. Now modifiers could be added, too. And the way would be opened soon for using dice to track damage and hit point scales beyond the six-sided.

A Hero now didn't simply need to get hit four times to die, but had a new figure, the sum of 4 six-sided dice rolls (Hit Dice!) that gave these more granular hit points. Still, on average, four hits would take a Hero out, but these no longer had to be simultaneous strikes. A Hero could take, on average, four hits in a whole expedition.

Hit points (rather than a capacity to take a fixed number of unqualified hits) introduced a new layer of uncertainty into combat, and uncertainty is one of the prime sources of fun in games. (See Peterson, PatW, p. 338, and Costikyan's Uncertainty in Games.) A lucky hero, whose player rolled high for hit points or whose foes rolled low for damage, might suffer well more than four strikes before dying.

The timescale of Chainmail was a big battle, but for dungeon adventures, it was a prolonged expedition, which might take more than one session. The use of hit points and longer sessions came with the idea that damage became cumulative, unlike the Chainmail rules, in which a hero must be hit four times in one turn to be killed, and healing after the battle was assumed.

Once the Hero (taking 4 hits to be killed) and Superhero (taking 8 hits to be killed) were scaled like this with hit points, and allowance was made for cumulative damage, it was an easy development to create the intermediary steps on the scale: a two-hit-die fighting man ("Warrior"), a three-hit-die fighting man ("Swordsman"), a four-hit-die fighting man (a Hero), a five-hit-die fighting man ("Swashbuckler"), etc. In D&D, the levels came with titles signifying not just potency but also prestige. Tracking hit points meant keeping records, but for small parties of individuals already outfitted with personalized statistics written down on cards, that was easier to do.

And so there were character levels, an outgrowth of military ranks in wargames. With D&D, individual characters could gain levels, and continuity of characters from session to session became vital to building up your forces for the next battle with experience. This was one of the essential innovations of miniatures wargames, whereby characters who survived one battle would increase in ability in the ongoing campaign (the word still used, increasingly idiosyncratically, for an ongoing series of adventure sessions in role-playing games).

Monsters and other foes had hit dice, too, depending on how hard they were supposed to be to defeat in combat. And then the clever idea was born that those monster levels corresponded to the level downward in the subterranean dungeon. Players could choose the level of risk they undertook.

D&D became a story of early-career individuals increasing in rank, instead of the wargame in which the story was about armies led by high-ranking heroes.

Why gold for experience points and level gain?

There are lots of intelligent "old-school" essays that attempt to justify giving power-ups (level increases) in return for treasure recovered, valued in gold pieces. The justifications are usually clever, but they are mostly retroactive, concocted long after the rule they justify was coined.

Blackmoor's dungeon expeditions were the first of their kind, as far as I know. But why were parties of individual characters going into the many levels of dungeons beneath Castle Blackmoor?

The reason was particular to Arneson's campaign. This was an ongoing wargame campaign with many serial sessions simulating a medieval-style fantasy battles. Arneson's rules for this campaign (some of which was eventually published in 1980, in sloppy presentation, under the title First Fantasy Campaign) make it clear that this wargame was all about financing the construction of forts and fleets of ships, raising armies and paying for them, and developing one's "area of interest" in a long-term war against an evil opponent, the leader of the "Baddies" to the North, named the Egg of Coot. (This conflict is also part of the roots of D&D alignment, which I discussed previously.)

Heroic characters involved in the great war needed cash to support their war effort. Arneson mentions, though, that he encountered a problem with motivation. He says he "solved" this (FFC p. 3) by giving out experience points not for gaining gold, but for spending it on the development of the character's established area of interest. This might be the capacity to make war, but it might include "wine, women, and song" (suggesting a real degree of player immersion). The more you spent, the higher in rank your character went. Your character could store gold, but if it was ever stolen from its hidden cache, you might lose levels!

One of Arneson's most important innovations, therefore, was to create a scaled-down special game session in the dungeon not focused on battlefield maneuvers, but in which individual heroes of varying levels looked for treasures for their war effort, effectively cash bonuses beyond what their lands produced. The dungeon was initially an experimental side-venture. The ones who got treasure (valued in gold pieces) could finance their own war efforts better. The campaign--a military campaign--required cash.

This also explained how heroes got to be leaders, the original link between collecting gold and becoming a significant figure in the wargame, becoming an individual of note. Heroic leaders got that way by spending cash to finance their wars and to develop their territory. The story of how they gained cash was also the story of how they came to be heroic leaders.

This is also why the goal of early D&D was to become powerful enough to build your own castle (temple, wizard's tower, domain, etc.). Originally, in the transition from wargame to dungeon exploration game, it was not that the dungeon led to domain-level play. The game started with what we today call domain-level play; the dungeon explained how they got there and facilitated bonus cash for this purpose.

D&D evolved as a way to explain how the heroes in the tabletop wargame got to be heroes that were hard to kill in war (had more hit dice) and that had special properties, and got more attention, as miniature pieces.

In a way, D&D evolved as a way to create background stories for heroes, those beloved individual miniatures in wargames.

It soon outgrew its origin, however. Instead of being an opportunity to build resources for a wargame, it became a new kind of game in its own right: the hidden-map dungeon exploration game. It became a story about lowly characters risking their lives for treasure and power, not for the defense of their domain. And what caused them to become heroic? They became harder to kill. They acquired more hit dice, more hit points, just as in the wargame.

These rules were developed to explain and elaborate wargame mechanics. They were not invented from scratch to represent dungeon adventures as a new idea. Many oddities and expectations of D&D's rules are holdovers from this stage of development.

Arneson's alternative to hit point inflation

 Arneson developed his rules over time. In First Fantasy Campaign, Arneson mentions that characters who advanced did not receive more hit points, but became harder to hit. When struck, characters made saving throws to avoid the damage. Fighters' saving throws against strikes improved faster than those of Clerics and Magicians (sic).

In other words, Arneson's campaign (eventually or originally, it's not clear to me) had defense rolls against attacks. This is rather like other well known rules from other games: the contest of dice pools in Tunnels & Trolls and Fighting Fantasy, the "soak rolls" in Ars Magica, defense rolls in GURPS, etc.

D&D got stuck with hit points per level, one short step from being a Chainmail mechanism for miniatures warfare.

Another "hit point" alternative to consider: Flip the hit point/level ratio.

In real life, young and healthy people are more physically fit and more likely to endure and heal from physical hazards. Injuries leave lasting effects on the body, diminishing physical capabilities for those who suffer them. Middle-aged athletes and soldiers get hurt. They become less effective. They retire.

In D&D, characters start off comically fragile. It's even a trope of the game. They get more powerful and seldom retire.

What if you flipped this? What if first-level characters were given a large number of hit points (or Endurance, or whatever), and serious injuries shaved those off permanently, little by little, until it was too risky to continue adventuring? This solves the problem of super-fragile starting characters, for those who don't like that aspect of D&D, and has more verisimilitude.

Well, a lot of games do it like that, in effect, already. But if we insist on D&D, what if we give high hit point values to level-one characters, and stipulate the conditions under which hit point loss is permanent?

It could be fun. Give each starting character twenty hit points (or however many you think is fair). Every time the character's HP fall below a certain level (say 5), take two hit points away permanently. Certain kinds of attacks (like "level drain" or acid attacks or whatever) give permanent HP losses. Only rare magic, blessings from the gods, and the like, can increase hit point maximums permanently.

I'm just throwing this out there. Adjust to taste.

Advanced characters become better at fighting, magic, etc., but they do not gain hit points per level. Advanced characters who have been beaten up one too many times decide when to bow out, as in life. Picking a fight is still really risky and may reduce your PC's long-term playability. Instead of having lots of PCs die off at first level, you'll have players who want to ditch the PCs who got badly hurt too soon, because they know that they have a more limited duration of play.

What is a D&D hit point?

Hit points are a way of simulating significance in a story. Significant characters don't die easily. We want to see them continue. The more time we invest in a character, the more endearing and enduring the character becomes. The death of a higher-level character means more.

In this way, every edition of D&D, including the "old-school" varieties, is about story mechanics. It's a "storygame" in which the drama is expressed through hit points. One score, hit points, indicates more than any other, without any verisimilitude, how much risk a character can take and how important a character is by virtue of the character's likelihood to continue to appear in the game and the story it generates.


Tabletop wargames simulated battles between armies featuring high-ranking individuals that were better fighters and harder to kill. These high-ranking (high-level) figures were represented by special miniatures. The heroes, who may have castles and led armies, were there from the start, leading low-ranking fighters and managing fortifications. The heroes were valuable assets, a focus of special attention.

On this front, the innovation of Arneson's Blackmoor, and of D&D, was to tell the story of how low-ranking fighters got to be heroes. The goal of the game now, in effect, was to become the kind of figure that had castles and led armies. To finance fortifications and armies, gold was required. Gold granted leadership status. The dungeon was initially a diversion, a chance to find extra caches of gold, beyond the finances provided by the produce of one's domain. Gold in the dungeon was initially a means to an end: to fund a war effort and the construction of a domain involved in a long-term war. This was the initial rationale for granting experience for gaining gold. Gold made low-level characters into high-level leaders because gold paid for the greater campaign efforts.

Soon enough, however, the focus of the game shifted from the great war to the lively tales of quirky individual characters who preferred to spend their time risking their lives in weird and wondrous dungeons. Arneson began to grant experience points when PCs spent gold on their specific interests. This was a secondary development.

In the ways described here, the most basic components of a D&D character were established in wargame antecedents. D&D told the back-story for potential heroes and wizards on the battlefield.

This matched the fantasy fiction that players were consuming then. Whether it was the humble hobbit Bilbo who risked adventure and made it there and back again with a pack full of treasure and a ring of invisibility, or the barbarian orphan Conan who eventually trod the jeweled thrones of the Earth under his sandalled feet, or the villager Ged of Gont who became a mighty wizard of Earthsea, tales of ordinary individuals who became mighty heroes were the fandom model for the shift from the battlefield to the dungeon excursion.

Yet unlike those fantasy heroes who were mostly loners, the ordinary individuals of D&D now adventured together in bands, like a small miniatures battle squad, making for a more lively and social game in which each player character's function was ideally complementary with those of the others. Even here, there were fantasy fiction models, such as the Fellowship of the Ring, in which one of each kind of character was recruited for the mission.

Gold granted levels because heroes used it to finance wars as leaders. In effect, gold made a character significant, a hero, and so harder to kill as represented by more hit dice.

gold > heroic rank > hit dice > victory

Then add the initial steps that made D&D distinct:

dungeon > survival > bonus gold > heroic rank > hit dice > victory


  1. Nice analysis.

    IMG a hit point is the extra effort required to avoid being hit by an attack that would otherwise hit and injure you. If you have run out of hit points then your dodge, parry, or block was too slow and ineffective and you were hit and injured and rendered hors de combat.

    The more experienced you are, the more effective your defence is, and the less effort (proportionally) you need to make to defend yourself.
    Even 1 point of fresh injury was of concern, and rendered you incapable of continuing to fight effectively (possibly because it had killed you). One of the distinctions between Chainmail and D&D was that a "kill" could be an soft kill for an individual, not just a hard kill - it was a figure removed from the battlefield, due to injury, death, capture, or cowardice.

    Injury IMG represented actual physical damage (a wound) and was a separate score. Each point of Injury represents a -1 modifier for each point to every dice roll, including every individual hit die when an attempt at first aid is made after the battle (and hit points are rerolled). Rolling a positive amount of hit points meant you were walking wounded (again with that negative penalty to any attempted dice roll). Rolling 0 hit points (there were no negative hit points) meant that you were seriously wounded and had to be carried out of there. Assuming you were alive of course.

    Normal Injury requires you to make a saving throw against Death. Again, with that negative Injury modifier. However armour added a bonus to your saving throw (as armour is designed to make you harder to kill).

    If someone wanted to ensure you didn't survive they could take a round to administer an automatic coup de grace. Provided they weren't being distracted by other combatants, of course. A defeated foe can generally be safely ignored for the remainder of the battle (absent heroic action). [My definition of melee was if an enemy could move to engage you, you were in melee. In Chainmail for example, if you were in 30' of an an enemy you were melee.]

    Injury heals at the stated 1 point per week (OD&D). Magical healing will heal 1 point of injury per die of effect. The Injury will still be a negative modifier to the number of hit points regained though. [Because IMG all damage done by magic is reversible, this meant moving the points of magically healed injury to a different column where they healed normally (effectively doubling the healing rate) but had no other effect. Unless dispelled by the spell sanguinary condemnation which moved these points back to the active injury column. But this was just to bring healing into alignment with my campaign's definition of magic.]

    It has seemed to work quite well for the last four decades or so.

    As for xp for gold we never awarded xp for gold, feeling it was double-dipping. Gold was itself valuable and we found it was incentive enough for a character, especially when applied to strongholds and the like. [A nice little fortified manor (and the estates to support it) was about a Level 4 purchase in my game.] However i was a heretic amongst my associates and allowed my players to buy xp for gp. Which not only helped eliminate it from teh game, but helped provide backstory, social contacts, and could be used, for example, to determine the size of a magic user's library or laboratory depending on how much they had spent on it.

    1. Thanks for sharing your house rules. The one you have devised, whereby hit points refer to avoiding damage, and some other numerical factor accounts for real injuries, is one of the more common remedies that I've heard about.

      I wonder how you would treat a missile strike that hits a target unaware of its coming. Would that damage go straight to "injury," because HP refers to the ability to avoid damage, and one cannot avoid an attack without knowing about it?

    2. That would be classed as a successful sneak attack. Which does immediate Injury, but since it occurs before "combat" actually initialises (sneak attacks are only possible outside of combat) it causes an immediate reroll of hit points (at the injury penalty from the attack) and then "combat" triggers.

      Only the missile-specialist class characters (ie Rangers with Stealth IMG) could make such an attack from ambush as a special attack. Other characters, even if they achieve total surprise, still allow characters to dodge or block the incoming attack at the last moment (just not very well, which is why the enemy get the bonus to hit).

      Sneak attacks are also possible as a special attack with melee weapons (but generally only by Assassins (a Fighter sub-class in my game) using Stealth to conceal the weapon from the target (and bodyguards that may give warning). There is a penalty to the stealth roll based on the size of the weapon.

      Remember that whilst hit points can be synonymous with fatigue (which is why they get the Con bonus), they actually represent "defensive" experience. You don't get more fatigue as you gain levels, you get more effective at defending yourself with less effort. You are also more aware of the possibility of a potential attack. So an experienced character is more likely to pull off the last minute block/dodge. Or not, if they don't have enough hp to absorb the damage from the attack.

      [Any fatigue effects (which are pretty rare, such as wearing heavy armour in the desert or having executed a forced march) generally manifest on the character's hit points are a single 1d6 die roll reduction (for each cause) in initial hit points to a minimum of 1 hp.]

      On the other hand, because of the wuxia basis of my game, guns always do Injury if you fail your saving throw against them. In tribute to Once Upon a Time in China 2 IIRC, where the expert martial artist falls apart and is seriously harmed by the small .32 revolver used by the British villain. Note that it is impossible to be skilled at using a gun. Also a prejudice from the wuxia basis of the game.

      [Speaking of films, the ideal effect of hit points is in the ambush by the clan members in After the Rain where Misawa easily reduces his attackers to 0 hp, but most of them actually survive the fight (with at least one notable exception).]

  2. Great treatment.
    I really enjoyed the linear presentation, cogently explained.

  3. This is simply a great essay on both RPG history and game design! I found it perfect! The nice finding is the Costikyan's quotation. That was a gem to my eyes, for a particular reason: it is a confirmation of my assumption... I have just written a post about this here:
    Thanks a lot!


Post a Comment