I don't have a good word I think for how I approach games but simulationism is definitely discarded too easily. Stealing ideas a bit from the rest of the thread I think the closest wording to what I do would be "genre emulation" using modeling. The goal is not to model atoms colliding, but model everything that's important to the genre which might impact the players. That's going to feel like simulation but it relies more heavily on abstracting complex systems down to a few variables which are relevant in the context of the game.

IMO the ideal rules operate under a few ordered principles (minimally):

• The rules produce plausible results.
• The players have maximal agency within the world (within the limits of their character's capabilities).
• If players chase the incentives available to them, neither the game nor the world will fall apart.
• If the rules are taken seriously and played out at scale (even imagined at scale), the resulting dynamics cause the setting to naturally emerge.
• The rules model the fictional "reality" in some way, even if highly abstracted. ie: All mechanics are associative rather than disassociative wherever possible.
• All of this is balanced with the consideration of how to present interesting choices within the context of the rules which yield an engaging game that you can actually manage to play and enjoy at the table.

So the game comes first, but efforts must be made to bring the fictional world and the rules into harmony. My biggest pet peeve in a ruleset is when taking the rules to their logical conclusions results in absurdist conclusions about the world. You end up needing to take the rules less seriously to take the world seriously. In the worst cases, the players decide the opposite - to the the rules seriously but not the world. Players also start metagaming more - even if it's well intentioned. They start intentionally limiting themselves and their plans in the world because they're worried they'll "break the game". The best game gives me all the tools I need to model the important bits, let the players trash the sandbox and have fun, and I don't have to start inventing the rules myself to make it all "work" at the end of the day.

the more popular rules-light, narrativist games IME tend to become "mother may I" games. In a game like Delta Green you kind of know how the modern world works so you can have reasonable expectations about how things work and how people will react if your character takes a certain action. In a game where you're in a fantasy world with entirely different metaphysics though, it's so much harder to answer even basic questions without asking the GM. IMO a decent ruleset provides sufficient guidelines for all the common cases to let players and GMs develop a shared understanding of how the fictional world works.

I don't mind if things like armor is abstracted behind armor class or hit points model something like abstract harm. The important pieces are modeled and the game works. What I dislike is when the rules start introducing free-floating meta-currency or things like plentiful cure disease or resurrection spells/potions. If you follow rules like resurrection to their natural conclusion, you need to start inventing weird workaround solutions to narrative problems. If the king is assassinated, why not resurrect him? If wizards can become invisible, fly around, and bombard a castle from the sky, why do we have castles with open courtyards? I prefer systems that actually consider the consequences of the tools they place in the hands of players and sets expectations for the world appropriately.

A perfect example of great rules: If you roll 3d6 in order down the line and keep the attribute requirements for classes, paladins and other unusual characters are not merely rarer because the GM or the book says so - they're actually statistically harder to find among player characters. The demographics actually scale up on the GM side as well so you can get a good estimate for how many paladins would exist in a city (or fiefdom) of a given size. That lets you play more intelligently as the GM and make the world more reactive.

Personally I've changed my mind on modular design after ~20 years of GMing. I used to love it, but now I think modular designs hinder the kind of player where meaning and structure/consequences emerge from the rules. You need at least a few interlocking systems with enough complexity to get interesting results emerging from them. The world is an integrated place, not a disjointed one. Procedures, processes, and subsystems should be thought of more like the "engine" for the world. If I run 1 thousand iterations of your process (and I will over the course of a campaign), does it make sense? Does it make sense at scale as much as it does for an individual player character? These things need more consideration from games.

Quote from: Steven Mitchell on March 11, 2023, 03:20:19 PM

Quote from: jhkim on March 11, 2023, 03:02:30 PM
It seems like you are using "genre emulation" to mean any game trying to achieve anything. Technically, one can say that there are genres of games (i.e. casino card games, Euro board games, etc.) and genres of simulation. But if one goes with that, then "genre emulation" just describes all TTRPGs that try to achieve anything. That makes the term less useful, IMO, because it doesn't describe anything that distinguishes some games from other games.

No, that seesaws too far the other way, though I can see how what I said would create that impression.  I'm basically talking sub genre, or maybe sub, sub genre.  There's "heroic fantasy".  Then there's "heroic fantasy with a dash of sword & sorcery in a dungeon crawl" sub genre.  Or even mashups.  Mashups aren't technically a genre, but they are certainly genre emulation, in that when you start blending the sources, there is some fidelity to the various sources.

When I say that if I ran Star Wars, it would be using a customized version of the Toon engine as part parody, that's not a genre as such.  Yet, we could talk about what it would mean to tweak Toon in order to keep the parts of the space opera/space fantasy genre mixed in with wacky cartoons.  I'd be seeking emulation in a way that can be communicated.

Fair enough. Reading this, I'd distinguish between:

1) genre(s) to be emulated - including subgenre like Spaghetti Western, subsubgenre, and mashups of multiple genres
and
2) how the RPG tries to emulate that genre - like specific rules, GM techniques, etc.

I've encountered before claims that D&D is the perfect genre emulation... of the D&D genre. Which is obviously silly. D&D is fine, and it doesn't have to be called a perfect genre emulation to be a fun game to play.

For genre emulation to be meaningful, then there needs to be a genuinely open question about how well the game accomplishes that emulation.

To take your example of Star Wars using the Toon rules...  On the one hand, one could just take it as an experiment and see how the game turns out, maybe adjusting things based on general feel. On the other hand, if the point is defined to be genre emulation -- then there needs to be a target that could be either hit or missed. Describe the mashup genre between Looney Tunes and Star Wars without reference to anything about the RPG, and then the RPG could be judged on how well it gives the feel of that.

I am only on page 3 in reading this topic but the way I see it is every trpg is a simulation from the perspective that it is the very act and nature of playing a roleplaying game that by doing this you are engaging in a simulation. That is the fundamental nature of ttrpgs.

You are using a set of rules to help players engage with the fictional setting via their self made character avatars that they can then get into the persona of to then act and roleplay in that setting and the rules facilitate this process in the hopes of achieving some state of in game and setting verisimilitude where the events, actions, and consequences of the player avatar's actions become and are plausible.

Quote from: Old Aegidius on March 11, 2023, 07:42:02 PM

• The players have maximal agency within the world (within the limits of their character's capabilities).
• If players chase the incentives available to them, neither the game nor the world will fall apart.

I think there needs to be a clear distinction between the player's agency and goals, and the character's agency and goals.

Story game will often include mechanics where the player decides to make the character fail because it will give the player some sort of advantage later on. For me, this is the biggest distinction between a narrative game and a simulation: the disconnect with what the player wants and what the characters wants.


As for maximal agency, again this applies to the character moreso than the player. Take the earliest example in this thread:

Quote from: amacris on March 09, 2023, 06:31:35 PMI want to simulate the free choices of a panzer commander, including the choice of switching to Team Russia, abandoning my tank to become a bandit, etc., whatever.

If the players have all agreed to play in an RPG where they are tank commanders on the Eastern Front, then one player can't suddenly decide to switch to the Soviet side without destroying the game. Note that the game world isn't affected, but for all practical purposes the RPG campaign will end. Either the game needs to be split into two games or some of the players have to quit.

So I don't think a simulationist game needs to account for all possible player choices, just those choices that are within the genre that all the players agreed upon before play commenced. This applies to the selfish Superman example as well.

Quote from: hedgehobbit on March 12, 2023, 12:26:01 PM

Quote from: Old Aegidius on March 11, 2023, 07:42:02 PM

• The players have maximal agency within the world (within the limits of their character's capabilities).
• If players chase the incentives available to them, neither the game nor the world will fall apart.

I think there needs to be a clear distinction between the player's agency and goals, and the character's agency and goals.

Ideally there need not be much of a distinction made because the player's goals and their character goals are in alignment, inasmuch as you can guarantee such a thing via ruleset alone. Your example of failing a test to gain a resource is a good example of what causes dissonance in the player's thought process. The rules should help the player "think" like their character to aid immersion or at least not get in the way.

As for agency, I agree that the rules need not cover every potential outcome. It's more important to explain a dynamic than to enumerate every possibility. However, once the ruleset has introduced an incentive then IMO it needs to follow this incentive to its logical conclusion and plug any holes that would, to use your example, encourage the players to defect. It's not quite the same thing but D&D's level loss for alignment change is an example of a rule that tries to plug a hole around alignment requirements which are baked into the rules (among perhaps other considerations). If players can change alignment on a whim without any consequences, alignment requirements cease to have meaning and the new incentive will be for players to act erratically.

Maximizing player agency IMO requires providing tools for the GM so they aren't totally left to fend for themselves if the players do something plausible within the setting. If characters are playing tank commanders, I'd still like some basic guidelines, processes, or procedures to help adjudicate outcomes outside of the main tank vs. tank scenarios. What happens in a combined arms scenario? What happens when commanders get out of their tank and are speaking with locals in nearby towns, get caught outside of their tanks with nothing but their sidearms, etc?

Many games lack proper tools. IMO a lot of modern "rules-light" games look more like "rules-incomplete" to me.

Quote from: jhkim on March 11, 2023, 03:02:30 PM
Right. The Ron Edwards approach of classifying games into one of three GNS categories is dumb, since the standard is hybrid -- which was originally recognized in the Threefold. But having the language to talk about "this muffin is too sugary for my tastes" is useful -- when one person might want a heartier bran muffin and one person wants a sweeter blueberry muffin.

I don't like sweets. So I totally get wanting less sugar in my muffin. The problem I'm having is in seeing this as a legit analogy to the sort of thing we're talking about here. Whatever we want to call it, simulationism, emulationism, realism, logical consistency, verisimilitude, I don't get is how doing less of that improves the drama/narrative/story/etc.

I accept we're doing hybrids. I accept there exists a theory that assumes hybrids. I get how these categories could be used to describe player preferences in broad terms. I even get how they could be used as areas in which a GM has strengths and weaknesses. I just don't see these big three or big four models say much beyond that. I don't see them translating to recipes.

I get how less sugar in my muffin gives it the less sweet flavor I prefer. I don't get how get how NOT tracking the number of bullets in Dirty Harry's gun would make the Do You Feel Lucky Punk encounter more dramatic/better for the story/more credible/etc.

So if I call out a GM or an RPG that handwaves tracking ammo, coming back with "Well, we don't care much for verisimilitude. For us it's all about the story," that's bullshit. Those things do not oppose each other at all in my view. They support one another. If you want to say, "Look, dude, we're just trying to have fun, and tracking ammo is just more work than we want to do," I'd say fine. That's an honest response. Bandwidth limits absolutely call for trade-offs. The benefits to drama, realism, and game (resource management) just aren't worth the extra work of tracking ammo. Fair enough.

Quote from: Lunamancer on March 13, 2023, 12:33:11 AM

So if I call out a GM or an RPG that handwaves tracking ammo, coming back with "Well, we don't care much for verisimilitude. For us it's all about the story," that's bullshit. Those things do not oppose each other at all in my view. They support one another. If you want to say, "Look, dude, we're just trying to have fun, and tracking ammo is just more work than we want to do," I'd say fine. That's an honest response. Bandwidth limits absolutely call for trade-offs. The benefits to drama, realism, and game (resource management) just aren't worth the extra work of tracking ammo. Fair enough.

Yes.  If you dig deep into all the bits, you can end up with some decisions that look strange at first glance.  This is where breaking it down into the components is useful, but only if you put it back together again--and then test that hybrid in play to see how it works.

How my system is handling ammo would probably annoy every purist on the planet, not to mention some people that don't mind hybrids.  Yet, it works as designed.  We are semi-tracking ammo.  You fire an arrow.  If it hits, don't mark it off.  If it misses, mark it off.  If it is a critical hit, mark it off.  If you fumble, either the arrow broke (miss) or you stuck it in a friend (hit), mark it accordingly.  This leads to cynical comments about the unfortunate friend helping you with ammo retrieval.

After the fight, if you hold the field, you can recover most or all of your ammo.  It's a GM call, based on what the opponents were using, the environment, and often a die roll. In an underground cave, plenty of time, and your opponents had arrows left in their quivers, you just restock.  If you don't hold the field, then the GM makes a call to knock a few more out of your quiver, based on the length of the fight.  More often than not, it's somewhere in between, because some of your "hits" broke the arrow and some of your "misses" are recoverable.  You might also be able to salvage some broken arrows which can be repaired or cannibalized to repair a lesser number.

Sounds counter-intuitive, and it certainly doesn't map exactly to the simulation as a step-by-step process.  That's because it is working on at least two different abstraction layers:

- We need to balance simulation and handling time. This is a compromise.  We found that if you tracked every arrow, bolt, and sling bullet, when you held the field it hardly ever mattered except in really long fights or fights with cliffs, lots of vegetation, etc.  If we don't track them at all, or use something like an "ammo" die, it's too far from the simulation. When you don't hold the field, the GM has a pretty good guess as to how much you are down.  Sometimes it rounds off for or against you, but not enough to justify the extra handling time of doing it the long way.

- At this higher layer of abstraction, we are simulating that when you fire an arrow, you really don't know what's going to happen. It's a whole lot less work than rolling to check for every single arrow breaking or not (which is way too fine a grain for this system), yet accounts for the same dynamic in a rough and ready way.

- Critical hits are different because the system says that ammo from a critical hit is stuck in the target and cannot be easily removed.  This is simulation, game, and drama, in that it affects decisions by someone that has been hit, hurts to get it out, and takes proper skills to retrieve. There are player decisions to make, with consequences for possibly extra damage to the target.

- As a purely game thing, marking misses and not hits speeds play. The person who hit is rolling damage.  When they miss, instead they mark the arrow.  Averages the time between hits and misses.  The rationale is that misses break on armor, smash into a wall, disappear in the trees, etc.  The critical hit things goes against this, but there's already a bit of excitement and slow down around that event, and the player is excited to mark off the ammo as part of that.

- As a drama thing, nothing kills the drama of a fight like the conversation about ammo usage.  GM: "Did you remember to mark off the arrow?"  Player: "Don't remember." For some strange reason, the disappointment in the miss reminds the player to do that more than hitting does, and it compounds the disappointment at the same time. 

- The whole set of rules for ammo tracking are just a tool to set a baseline.  The GM always adjudicates.  OTOH, when you start getting really low on ammo, each player always has the option to track it explicitly.  When you are down to your last 5 arrows, you tend to be much more careful about the shots.  So not a big deal to track in those circumstances.  Each player can make that decision on their own terms.   I'm always happy to let the player decide where the threshold is.

- The system as described is not always transferable to other game systems.  Part of the reason that it works is because of the expected number of hits and misses tend to be fairly close to 50/50 for most characters, most fights.  The penalties to hit for greater than short range tend to be strong enough that players are hesitant to make that shot--which was also a design goal.  In a system where you wanted eagle-eyed elves or Robin Hood wannabees making incredible shots, the system would be counter-productive.

It may sound like one of those wacky set of rules from AD&D to that doesn't really map to any particular style, but that's because it started with a design to support other parts of the game, and then evolved with experience.

Great article!

A fellow Simulationist here also. To me, as some have said before, the feeling of playing/gamemastering in a "real and breathing"
game world brings the greatest of immersion. I guess I´m just a braindead autist.

Quote from: Steven Mitchell on March 13, 2023, 11:32:00 AM
- We need to balance simulation and handling time. This is a compromise.  We found that if you tracked every arrow, bolt, and sling bullet, when you held the field it hardly ever mattered except in really long fights or fights with cliffs, lots of vegetation, etc.  If we don't track them at all, or use something like an "ammo" die, it's too far from the simulation. When you don't hold the field, the GM has a pretty good guess as to how much you are down.  Sometimes it rounds off for or against you, but not enough to justify the extra handling time of doing it the long way.

Allow me to confess my sin. Not only do I not track encumbrance religiously. I also handle it inconsistently.

In principle, I'm actually an even bigger stickler than the rules for deciding what you can carry. It's not enough you keep enc within your weight allowance. You have to be able to explain exactly how you can carry everything.

What I generally find in practice is, players do not want to commit to having a potion in some awkward place just to justify how they can carry it. Because the moment they do, they realize they're admitting their character wouldn't be able to get at it quickly in a pinch. And so players tend to streamline their adventuring gear and supplies. And because armor in 1E just imposes lower movement rates regardless of encumbrance, I almost never find players actually in danger of being encumbered at the start.

Loading up on loot during the adventure is a different story. That's when I begin to track encumbrance.

So I don't lose an iota of "game"--encumbrance still serves the function of imposing a limit on how much loot PCs can take back and requiring players to prioritize the treasure.

I'd argue "simulation" is actually improved since I add the requirement of "make this make sense" and am not satisfied just with the numbers adding up.

And it aids in "drama" as well. The neglected step child of the Hero's Journey is the return journey. And for good reason. If you've just conquered the dungeon, why should walking back through it be exciting anymore? Well, if you got to go back through it but burdened, where you can't outrun wandering monsters, where you can't just leap that pit, where you might have to toss the idol to Satipo, that makes it all new again and worthy of continuation of the story.

And it saves half the work of tracking encumbrance.

Quote from: Lunamancer on March 13, 2023, 01:02:44 PM

Allow me to confess my sin. Not only do I not track encumbrance religiously. I also handle it inconsistently.

In principle, I'm actually an even bigger stickler than the rules for deciding what you can carry. It's not enough you keep enc within your weight allowance. You have to be able to explain exactly how you can carry everything.

What I generally find in practice is, players do not want to commit to having a potion in some awkward place just to justify how they can carry it. Because the moment they do, they realize they're admitting their character wouldn't be able to get at it quickly in a pinch. And so players tend to streamline their adventuring gear and supplies. And because armor in 1E just imposes lower movement rates regardless of encumbrance, I almost never find players actually in danger of being encumbered at the start.

(snip)

I do similar things for similar reasons, with similar results.  I don't have armor directly affecting movement rates, but I do have armor hitting encumbrance, and then thresholds for light/medium/heavy encumbrance are fairly forgiving in numbers, but brutal in the results for extended movements.  So yeah, except when hauling treasure out (or unconscious friends), everyone moves heaven and earth to stay under medium.  And if fleeing for their lives, they start ditching things to get down to light.

The exception for me is because I have fairly nasty, semi-realistic restrictions on basic supplies (food, water, camping gear, medical supplies, etc), characters often are pushing medium when they go in.  They hope that they use enough of it so that by the time they come out, they've got treasure to take its place, but sometimes that doesn't work out so cleanly.

It probably makes a difference too that I'm using a silver-based economy, not AD&D "gold rush", with the characters flirting with destitute for the first few adventures.  Leaving some of your food in favor of some treasure is not the easy decision it would be in the typical D&D session, especially since some of that early treasure is of questionable value.  Of course, players getting too aggressive with trading food for treasure runs the risk of running out on the way back, which is equally true for you and me.  It's only the risk/reward calculus that changes a little.  On the other hand, some of those near destitute characters can't afford enough equipment to come anywhere near the encumbrance lines.  So those first few adventures going in, I can more or less ignore it.  Very handy for introducing new players. 

Quote from: Lunamancer on March 13, 2023, 12:33:11 AM

Quote from: jhkim on March 11, 2023, 03:02:30 PM
Right. The Ron Edwards approach of classifying games into one of three GNS categories is dumb, since the standard is hybrid -- which was originally recognized in the Threefold. But having the language to talk about "this muffin is too sugary for my tastes" is useful -- when one person might want a heartier bran muffin and one person wants a sweeter blueberry muffin.

I don't get how get how NOT tracking the number of bullets in Dirty Harry's gun would make the Do You Feel Lucky Punk encounter more dramatic/better for the story/more credible/etc.

So if I call out a GM or an RPG that handwaves tracking ammo, coming back with "Well, we don't care much for verisimilitude. For us it's all about the story," that's bullshit. Those things do not oppose each other at all in my view.

I agree that story and ammo tracking aren't inherently opposed. But I do think that there are different approaches that work for different people depending on the genre and their goals.

I think your example is an interesting one. To me, the whole point of that Dirty Harry scene is that there is tension over whether there is a bullet left in his gun or not. If we're playing an RPG where tracking ammo is part of the game, then the player marked off the bullet, so there's no tension. Everyone at the table knows that the gun is empty.

Quote from: Lunamancer on March 13, 2023, 12:33:11 AM
If you want to say, "Look, dude, we're just trying to have fun, and tracking ammo is just more work than we want to do," I'd say fine. That's an honest response. Bandwidth limits absolutely call for trade-offs. The benefits to drama, realism, and game (resource management) just aren't worth the extra work of tracking ammo. Fair enough.

To me, the choice of ammo tracking isn't a simple fun/not-fun. It will depend on what sort of game I'm running.

For example, if I'm trying to emulate a gritty military action genre, then I'll probably track the exact weight of gear and the count of ammo. Then it serves a purpose of bringing the players into planning and tactics. On the other hand, if I'm emulating an over-the-top wacky action like Bad Boys or Hard Boiled, then I'll be looser about those things and go more for the "Rule of Cool".

Alternately, I might make the choice between the two based on the level of tactics/strategy that I'm trying for. I might go for faster conflict resolution that gets more to the bigger picture, rather than tracking action and ammo shot by shot -- so more like Memoir '44 or Up Front rather than Squad Leader.

Different games and priorities will mean different choices about the rules and practices.

Quote from: jhkim on March 13, 2023, 01:20:19 PM
I think your example is an interesting one. To me, the whole point of that Dirty Harry scene is that there is tension over whether there is a bullet left in his gun or not. If we're playing an RPG where tracking ammo is part of the game, then the player marked off the bullet, so there's no tension. Everyone at the table knows that the gun is empty.

I meant to address that too, and got sidetracked in my long post.  One of the main design goals of my system was to introduce more player uncertainty.  Ammo is one tiny part of that goal.  Specifically, I think it should be a rare character that knows exactly how much ammo they have left when the battle is winding down. I couldn't find a way I liked to give that exact result, but it being kind of vague once the battle was over is the next best thing.  And good enough, since I primarily care about another wave of foes arriving than I do about recreating running out during the fight.

Also, I think there is another nuance to the Dirty Harry scenes (plural because they revisit it in a later movie).  The bad guy doesn't know.  The audience doesn't know.  But it is strongly implied that, despite his nonchalant claims to the contrary, Harry does know. That's part of the point when he pulls the trigger on an empty chamber the first go around, and then kills the guy in the later movie who doesn't give in.

Quote from: Steven Mitchell on March 13, 2023, 01:30:02 PM

Quote from: jhkim on March 13, 2023, 01:20:19 PM
I think your example is an interesting one. To me, the whole point of that Dirty Harry scene is that there is tension over whether there is a bullet left in his gun or not. If we're playing an RPG where tracking ammo is part of the game, then the player marked off the bullet, so there's no tension. Everyone at the table knows that the gun is empty.

I meant to address that too, and got sidetracked in my long post.  One of the main design goals of my system was to introduce more player uncertainty.  Ammo is one tiny part of that goal.  Specifically, I think it should be a rare character that knows exactly how much ammo they have left when the battle is winding down. I couldn't find a way I liked to give that exact result, but it being kind of vague once the battle was over is the next best thing.  And good enough, since I primarily care about another wave of foes arriving than I do about recreating running out during the fight.

Also, I think there is another nuance to the Dirty Harry scenes (plural because they revisit it in a later movie).  The bad guy doesn't know.  The audience doesn't know.  But it is strongly implied that, despite his nonchalant claims to the contrary, Harry does know. That's part of the point when he pulls the trigger on an empty chamber the first go around, and then kills the guy in the later movie who doesn't give in.

While it's more bookkeeping for the GM, I'd say the way to handle that is to not let the player use any tool (not even scratch paper) to track ammo use while in combat (and then give them some other thing to mentally track of (ex. At the end of each of their turns roll two dice and have them multiply them; if they can give you both the rolled numbers and the answer at the start of their next turn they get a small benefit... now they have to hold four numbers, one their Ammo count, and pay attention to the rest of what's going on.

The idea is to skirt the edge of a player's short term memory capacity even as the GM keeps an actual tally of the limited supply. Maybe they can keep an accurate count, maybe they miss or add a shot or two because the number is close to one of the numbers they have to multiply and remember between turns.

But when you actually reach the point where "Do you feel lucky?" comes up it could be a legit question for everyone but the GM.

An even easier method for a more abstract level of combat would be that each round of combat involves, say, 1d4 shots that the GM rolls secretly (sorta like how the original AD&D arrow weights were for about equal to the number of shots a bowman could get off during the one minute combat round... the attack roll is for whether any of those shots actually land).

Each round the PC knows about 2.5 shots are being used, but a string of 1's or 4's could see you either with more or less shots left than you believed you had.

Quote from: Steven Mitchell on March 13, 2023, 01:30:02 PM
Also, I think there is another nuance to the Dirty Harry scenes (plural because they revisit it in a later movie).  The bad guy doesn't know.  The audience doesn't know.  But it is strongly implied that, despite his nonchalant claims to the contrary, Harry does know. That's part of the point when he pulls the trigger on an empty chamber the first go around, and then kills the guy in the later movie who doesn't give in.

The way I've always played it and have always seen it played syncs up really well. Players are left to track their own ammo. It's not a given that all players will know how much ammo the one has. Maybe if they made it a point to pay attention and count. Same could be said for members of the audience watching Dirty harry. Not only that, as GM, I don't also track it; the player is on the honor system. It's just not something I want on my plate. And so that means in a situation like this, not even I as GM know whether Dirty Harry's got another shot. And so it's pretty neat that how what everyone around the table does and doesn't know corresponds perfectly to what their corresponding roles do and don't know.

It's interesting to see how people track arrows. My problem with ammunition is that it is only impactful if either the players or I put in a lot of effort tracking it. As a player, I eventually always forget. As a GM, I have better things to do and I'm already juggling more interesting considerations. Since tracking it is tedious, it basically gets dropped (similar to tracking poundage).

Even with very meticulous tracking, ammunition is not generally expensive enough to present a scarcity problem. In practice, having 100 arrows or 20 arrows is the same if you are only worried from moment to moment about the next shot you're taking in combats that last a few rounds at most. It's just way too many arrows either way to present a chance of running out except over a very long and grueling crawl. Once you take a step outside the arrow tracking world, you're now adopting an abstraction with a lot of tradeoffs either way.

I decided to give players a choice to explain why they missed a target whenever they have a clean miss. Either they're getting tired, they've failed to maintain their equipment properly, or they're just stubbornly stuck on too hard of a target so they've wasted the remaining ammunition in their quiver. All choices have their tradeoffs. The idea behind offering a choice is for the sake of the game (interesting choices make interesting games). Tying a miss to a big consequence ensures everyone is more worried about ammunition scarcity. Provisioning equipment for an expedition is much more interesting under a system like this because one way or another you're going to need equipment and potentially one or more backups for key equipment. All of this also reflects how people often explain away their failures in real life haha!

I like the modeling - more skilled archers miss less often, which also implicitly means they have better endurance drawing and loosing their arrows, as well as more experience maintaining their equipment in the field. A less skilled archer is more likely to neglect or abuse their equipment, overestimate their capabilities, and/or get stuck on a target and stubbornly waste shaft after shaft losing track of their remaining quiver arrows. I reused this for melee misses (melee of course cannot waste ammunition). People get tired, equipment starts to break down or gets used up, and more equipment is needed to keep everyone in tip-top shape. This equipment usage eventually opens up more space to carry loot. Altogether, I think it's a good pacing mechanism and makes people think a lot more about what they're carrying altogether. Sidearms are actually useful because maybe you've decided you're not doing a good job keeping your blade sharp.

In case anyone wants to use it in their game: Any d20 attack roll which would have hit if the target were unarmored is NOT a clean miss so no penalty. So generally if you'd hit AC 10 (plus whatever non-armor bonuses). By the time you hit Thac0 10 (or +10 attack bonus), you're mostly past the point of worrying about this stuff. Considering you're reaching name-level at that point, I think it's fine. You have people who fight your battles and sharpen your swords for you anyway.