"Punishing" players

Can we stop talking about "punishing" our players?

It seems like you don't have to go very far online to find someone running a game of D&D whose players have done something that they think requires "punishing". Or, alternatively, to stumble across a Reddit post or four asking whether they should "punish" their players for some action or decision, or whether it's too mean, unsporting or whatever.

Here's why this bothers me. The concept of punishment, and its counterpart, the concept of reward, are normative concepts. That is, they're concepts we apply to actions we evaluate as either good or bad. More specifically, punishment and reward are applied in the context of training specific behaviours; we punish behaviours we don't want to see, and we reward behaviours we want to see more of.

Now, it may just be me, but I don't think it's the GM's job to train players in this way. Doubtless the GM is responsible for training players in some sense - teaching them the flow or rules of the game, getting them to be good sports etc. But this isn't the sort of training that "punishment", as usually meant in ttrpg discussions, is aiming to inculcate. We're not talking about someone being a bad sport or forgetting the rules, and the GM throwing a high level dragon at them as "punishment". The discussions I'm talking about usually take the following form: The players did x; y is a plausible or logical consequence of x, but y is undesirable to the player characters, hence y is a punishment. E.g. (from Reddit): One of my players gave some blood to a demon, how can I punish him?

The troubling thing here is attaching a normative status to the consequences of player actions. What did the player do wrong here, such that it needs punishing? Sure, they took an inadvisable action which they could easily foresee having undesirable consequences. But what makes those consequences a punishment, as opposed to just things that, you know, happen? As a GM, it is your job to rule on what happens when your players do inadvisable things, but it isn't your job to try and discourage them from doing those things, much less using in-game events.

This has a reverse case as well; GMs whose players do inadvisable things are often reluctant to follow through on the consequences of those actions, because the natural consequences are seen as too punitive. The situation here is usually that the players have decided to take on a fight that has been telegraphed as too strong for their current level, or similar. Here the problem is more obvious: The players' getting party-wiped is the logical consequence of their actions, but the GM doesn't want to rule that it happens, because they feel it's too harsh, or whatever, so the GM feels like they're in some sort of a bind.

But this mentality creates all sorts of problems. For one thing, the GM is now running every ruling on the players' actions through an extra filter, asking something like "Do they deserve this?" The first big issue here is consistency: As I've been saying since way back in the mists of time, players need a baseline level of consistency in outcomes for their actions, otherwise they stop making plans and strategising, because there's no point. And the moment where the players internalise the idea that whether a particular action goes through depends on whether the GM thinks they deserve the "reward"/"punishment" is the moment players start to tune out and become passive.

I don't want to make anyone paranoid, but I feel like a thing I keep running into in these GMing theory sessions is the fact that players learn or internalise more than they may be explicitly aware of. For instance, players may not be thinking to themselves "Wow, that was a really railroadey session" - indeed, it could well be that if you asked them, they would tell you honestly that they thought they had lots of agency over the game's outcomes. But those same players may have developed a feel for when the story is dragging them in a certain direction - they learn to sit back and let plot events happen, they learn to go and do the thing the important NPC flags for them, even if their characters wouldn't be interested, etc etc. These two things - thinking intellectually that you have agency, but internalising on a practical level that you don't - can coexist.

I picked railroading as an example here just because it shows clearly how players can internalise things about how the game runs without being explicitly aware of them. But it's not unrelated to the reward/punishment mindset. Think again of the "disruptive" character, like the Thief/Rogue who lies and steals from NPCs - a prime candidate for GM punishment, on this approach - and ask what exactly they are disrupting, if not the intended game progression?

The fallacy of the "punishment"/"reward" conception is, in effect, that the GM thinks themself into a headspace where they aren't just ruling on what happens as a neutral arbiter of the world, but are instead either approving or censuring what the players do. But this carrot/stick mentality makes what happens in game contingent on what the GM wants to communicate to the players. If you stumble into the dragon's lair by accident, the GM might not want to "punish" you, since you didn't really do anything wrong, so the dragon turns out to be a pushover. And that compromises the consistency of the game world in a way that makes players' predictions of the outcomes of their actions impossible, and therefore makes actively strategising and planning novel approaches impossible too.

🕸

You don't see this "punishment"/"reward" discourse nearly so much in the old school ttrpg scene. This is despite the greater prominence in old school style play of things that could readily be considered both rewards (treasure, magic items) and punishments (traps, curses, unbalanced monster encounters etc.). There are, I think, two main reasons for this, and each of them tells us something interesting about the game.

First, although the GM arguably has a particularly prominent role in old school games (people are constantly going on about such games running on "GM fiat" - by which they actually mean the GM acting as arbiter), crucially, the GM has no agency. The ideal OSR GM is effectively an entirely passive referee - their job is only to judge the logical consequences of the PCs' actions, applying what they know about the world to do so. Even when the GM is strategising e.g.: about how a band of orcs could best construct an ambush to take down the party, they are doing so because that's what the orcs would be doing - if there are limitations, like the orcs' not being very cunning or brave, then the GM works within those limitations. Their aim, in terms of their role in the game, isn't to take down the party, but to figure out what the orcs would do.

This means the GM doesn't want to train the players to respond in particular ways, because there are no paticular ways the GM wants the players to respond. The old school GM doesn't need players who are responsive to plot beats or cues as to the intended encounters, because there are none. Indeed, figuring out which encounters you can and can't handle (and coming up with plans to circumvent the latter) is a huge part of the game for the players, not a prerequisite skill the GM needs to train into them so the actual game can progress.

This is why the idea of the overbearing, all-powerful GM that sometimes gets applied to OSR play is a misconstrual: The ideal OSR GM looks on paper to have a huge amount of power, because so much is given over to them to rule on, until you realise that they have no goals towards which to use that power. The ideal OSR GM is, at the table at least, essentially a sort of nonentity, in a sense that finds its expression in things like the Blorb.

[I could go into a whole side rant here about how this misconception contributes to the pivot to the approach in D&D 3e, which tries to tranfer some power away from the GM to the system itself, as well as to the advent of storygames, which attempt to transfer that power to the players, but that would be a whole other post. That's not to diminish these approaches or games in any way, of course. I think of them as being a little bit like the spork: An innovative way of solving a poblem that sort of already had a solution, but without whose existence the world would be considerably poorer. There are probably people out there who prefer the spork for what it does that other cutlery doesn't so, and I salute them.]

The second reason you don't see the "punishment"/"reward" discourse nearly so much around old school play is to do with players' approaches. Again, this is something of an ideal, but I think the OSR way of approaching one's character is largely to see what happens to them, and, provided it's decently interesting and adjudicaed fairly, to be happy with whatever fate befalls them. This means both in terms of not pre-plotting a destiny for them, and letting their distinctiveness arise through play (e.g.: in what character-defining magic items they discover), and in terms of not having a backstory where the most interesting thing in their life has already happened to them.

It also means accepting a good, unforeseen character death - heroic, colourful, or even just funny - as a suitable and satisfying end. And, before anyone springs from the woodwork to push back on this being how this works ("That's not how classic ttrpgs really used to play! We used to get really invested in our characters!"), do please note that this isn't at all incompatible with caring deeply for your character, wanting them to succeed, fearing for their life and so on. In fact, you'll find yourself so much more deeply invested in your character if there is a genuine possibility they might not live to see the end of the expedition. Think of it more like having a healthy approach to character death: When a character dies, it isn't that the game has failed you in some way (or worse, that the GM has inflicted a punishment on you) - that would require some sort of expectation going in that your character will survive. That's an outcome you want to try and secure, for sure, but it's not an expectation built into the game.

Indeed, that lack of expectation about what will happen to your characters is what makes the game exciting. With that mindset, it doesn't make sense to look at Faustian pacts with demons coming back to bite you, or getting immolated holding back that dragon as "punishments" - they're colourful pieces of the tapestry of your character's life (sometimes the pieces at the very end), and the whole game is about accruing those pieces. My table has seen player characters mutated into mindless Chaos spawn, have their souls sucked away to the howling void between worlds, age seventy years in a few minutes, and crash-land a winged horse into the ocean. On a certain approach those might be considered "punishments", since they were all the consequences of characters taking actions that were perhaps ill-considered - which is just to say that they were largely self-inflicted. But these are exactly the events my players remember most fondly (even when there's some healthy anguish mixed in there). Rather than being punishments for mistakes they want to avoid repeating, they're exactly the sort of thing that keeps them coming back to the table. Sure, they'll think twice next time they touch the evil obelisk, but that just means they'll approach it in a more interesting way - they're not going to not touch it.

🕸

If there was a short way to sum this up, it would be: The concepts of punishment and reward are about reinforcing particular behaviours. But what are you trying to train your players to do? To not do stupid/dangerous/ill-advised stuff? Then you're trying to train them out of the best part of the game, for them and for you. To follow an intended progression of encounters or set-pieces? Then you're training them to be passive players, and that never ends up with a satisfying game, for anyone - try just talking to them about your vision for the game. Either way, this lab rat stuff is working against you.