Pixel-bashing and OSR-style play

How do you avoid pixel bashing (aka pixel bitching, aka ten-foot-poling every passage) in OSR-style play? Should you?

To me, OSR-style play means (among other things):

  1. Trying to avoid rolls: if you have to make a skill check, or saving throw, or (gods help you) you get in an even fight, then you’ve probably made a mistake. Creatively using/interpreting the environment and the details given to you by the GM rather than relying on rolls… that’s where “skillful play” lies.

  2. Exploring a concrete environment. Even if the environment is randomly generated in play, there’s an expectation that the environment, once established, is there and it’s concrete and the GM’s job, in part, is to make the environment “make sense.”

  3. The environment being explored is dangerous, and you are squishy. Traps, monsters, cursed items, etc. Very few HP. Save or die effects. Be careful!

  4. The environment having hidden things to find and exploit: secret doors, false bottoms, gems hidden behind the statue, back doors, etc.

So… how does an OSR-style game do anything other than devolve into pixel-bashing?

I’m thinking of a recent game of WoDu with @JimLikesGames and @Ignotus, playing Through Ultan’s Door. Almost the first room we enter has a grisly, mysterious scene (hanged masked corpses around a strange fountain) with bear traps hidden under the straw and a lurking beastman. A half-dozen rooms spiraling off of this central spot, some with hidden treasure.

The environment is rich, mysterious, well-defined, kind of scary. Are the bodies just bodies, or undead waiting to get us, or filled with maggot swarms, or what? Is the fountain just a fountain, or something special? Does tossing a coin into the fountain do anything? Are there gems in the fountains eyes? Should we sweep the straw for traps or just blunder about? (Shit, traps.)

We (mostly me) spent a ton of game time poking and prodding, moving carefully, suspicious of every detail. 90% of those details didn’t matter. And some of the details that I ignored (e.g. the straw under the hanged men, recently disturbed) revealed a trap that could have killed (or at lest seriously hobbled) most characters. Yup, I got to make a roll to dodge, but it felt like a chance to recover from a mistake, not an avoidance of the mistake itself.

The whole thing seems designed to provoke careful, methodical, tap-tap-tapping with a 10-ft pole. Con games of Swords & Wizardry that I’ve played have been similar, though the presence of 4+ other players has always led to someone getting bored/antsy and therefore doing something stupid and getting us all in trouble. (Again, reinforcing the value of being careful and methodical.)

So, like…is this right? Is this slow, methodical pixel-bashing (punctuated by terrifying, sometimes unavoidable peril or violence) the intended approach to play? If not, why not? What does “smart, careful play” look like without pixel bashing?

If you want to prevent your players from taking 30 minutes of play time to explore any given room… how do you do that?


I think as a DM you have to set the players’ expectations in the right way so they won’t think this is your style of setting up a dungeon. I’ve played in old-school tables where the DM clearly signed that was how the game was going to be (and there were deadly traps in every other room, and not always clear signs of them). I usually give people some extra details about a place that is trapped (and obviously, also give details or put odd stuff that doesn’t do anything in some other places) because I don’t want my players to be checking for traps in every single tile or rock in the dungeon.

You can also set the expectations clearly to players in the beginning of a session, especially if you haven’t played with them before. Just because you’re playing an OSR-style game I don’t think that gets in the way of using some more open communication with players :slight_smile:


First thing I will point you to is this piece by Chris McDowell (of Into The Odd fame). He makes a ton of good points there. I would also recommend hitting all the previous posts he refers to in that post.

If we are focusing on the traps issue I agree it becomes super slow going and boring AF if for every single room the players (or just the Thief) have to intricately describe how they search for traps. I’m not a big fan of “gotcha” trap situations. (ie any kind of thing where the GM hasn’t noted anything out of the ordinary, you walk into the room, and you spring a trap that just hoses you.)

The GM needs to (as Chris notes) provide the player with information that allows them to make an agency-affirming decision. If the GM says nothing about the room of note, then either there should not be any traps or whatever trap that is there needs to give them a decision to make to mitigate the consequences. I think if you are playing a game where chance (ie a roll of some sort) comes into any sort of traps checking then I think it would be fair to have whatever “decision” the player has to make be chance-affected as well. If, though, checking for traps is entirely descriptive in your games then an acceptable post-trap “decision” needs to have a descriptive option. (ie doesn’t require a roll - though if the player decides to do something that requires a roll then that is fine)

Again it really comes down to information. If you say there is a sarcophagus in the middle of the room I think it would be fair to have it trapped and require the players to describe how they check it for traps (or do a roll if that’s your system of choice) - it is a thing you have noted in your description of the room - that means (or at least should mean) it has import.

I think in your example, unless straw on the floor was found in preceding rooms, that is a signal that it is something to interact with. One piece of this though that I will note is that a lot of this is only learned through previous encounters and really through previous sessions. New players aren’t going to know this. (Though I would highly recommend anyone new to OSR stuff (GM or player) read through the Principia Apocrypha.)

I think @seaandsailor has a good point about setting expectations. Especially if you are running a published module as opposed to something you put together yourself. For example if I was running Grimtooth’s Museum of Death or Tomb of Horrors I would say up front - this is a deathtrap dungeon - act accordingly.

All that being said - I still struggle (on the GM’s side) with this stuff. Especially when it comes to things like wandering monsters or rooms with some big fuck-off monster that can’t be bypassed. Sometimes it is hard for me to see what the set of options are for the decision - which can make me feel I shouldn’t make use of the thing the module says I should make use of.


It seems like, with all the retro clones that now exist, at least one of them would give GM advice along the lines of: “when a group enters a room or other situation, the GM should say ‘you do some cursory prodding of the room and A, B, C, D and E catch your attention’ as a shorthand message to the players meaning ‘there is a lot of crap that might be going on here, but I’m just telling you up front that only A-E are worth using table time to mess with’.

Or: the party is given a bowl of tokens and the understanding that pixel-bashing is not necessary. If they blunder into something that could have been discovered by pixel-bashing, they just spend a token from the bowl and retcon that earlier (off screen) prodding found that out, so what do you do about it now? Maybe each token represents time that passed (presumably to represent said poking), and maybe some clock advances as tokens accumulate. Maybe clever solving puts tokens back in the bowl. (Call it, the “Ten Foot Bowl™”).


In my games I shortcut a lot of the exploration by asking the players what they are doing more generally. If they are running down a corridor then they will spring any traps there, Save to avoid. If they walk down the corridor I might give them some clue that something is amiss. If they say they are stepping very carefully and checking every stone, crack and bloodstain then I will probably have them safely find the trap and ask them how they want to avoid or disarm it.

Other factors might come into account too. Exhaustion, magic effects, tricky enemies, whatever makes sense in each scene.


I have not technically used this framework in play yet, but I’m considering it for future sessions and games.

When you move forward or inspect an object in a dangerous area, decide if you are proceeding carefully or quickly…

  • If Quickly, a Turn does not pass, but…
    • Any creatures present get a bonus to Surprise you
    • Unless it is explicitly triggered by an action a PC takes, there is a chance (X-in-6, considering placement, age, complexity, etc) that any trap is triggered, per individual.
      • If it is triggered, those affected receive a saving throw to avoid or reduce its effect.
  • If Carefully, a Turn passes, and…
    • The Dungeon/Hazard/Event/Encounter Die is rolled
    • Any trap that you would trigger by moving into and exploring the area is automatically spotted, though its trigger, purpose, and function may not be obvious.

I have seen OSR players establish a standard operating procedure for certain things, such as checking a door for traps. They let the DM know how they do it (look without touching first, check for air currents, tap different colored surfaces, etc) that is established as how they are doing it each time without going through the whole process each time. Them the DM can say, Ok, as you’re tapping on the different jewels one of them sounds hollow. The player knows that the previous steps didn’t present obstacles or give information and they are jumped right to the important part.

Edit: Can’t edit the first line OSR typo on my phone. Established marching order is another player tool.


First of all, I got really excited about the title. I thought we are going to talk about pixel art in OSR or something along those lines.
Is that a term widely used in OSR? I have never heard it before, I guess it is a spin on “Pixel Hunting” from point & click adventure games…
I kinda hate using that term for OSR :stuck_out_tongue:


If this is what your group is into? If so, sure - go for it! It would be too methodical for my playstyle, but I never really got into the “combing through the dungeon” part of OSR (or RPG in general), but I can see some groups enjoying it and treating it kinda like a puzzle.

So, I guess my answer is - it depends on your group. Talk to them, see if this is something everyone enjoys. Established rules and methods are important and it is good to know them, but it is more important to know why and when to break them.

…and to connect it to my first paragraph rant - here’s a tiny pixel skeleton I made a long time ago:


For what it’s worth, here’s what I would say to the OP of this older thread. (I’m sorry I just stumbled on it.) I’m the author of Through Ultan’s Door. The Ruins of the Inquisitor’s Guild, the dungeon you explored is from my home game (and issue 1 of my zine). Here are some things that are currently going on in that very game, three years in.

In the last session, the players were trying to foment a slave revolt in a community to which they have serious connections in the inverted white jungle of the dreamlands. Another player character, was exploring an ancient treaty between the spirits of the air and the predecessors to the Zyanese, because she read an in setting conspiracy theory book that led her to believe this ancient treaty might allow them to turn the spirits of the air against the Hidden King of Zyan, with whom they’re engaged in a cat-and-mouse struggle. I have another character who learned an art of dancing in the dreamlands and has brought it back to the waking world, and has started a dancing troupe. In other words we have dense, faction-based play, with rich player developed goals, including artistic and religious goals. How has the game evolved to a point where we’re doing anything but “pixel bashing”? This would require an INCREDIBLY long answer–basic teaching a whole style of play that this question presupposes (in good faith) you are not familiar with. The basic answer is that they emerge over time from open, player-driven interaction with an already existing environment.

That potential is there in seed with the very first dungeon–although not if it’s played as a one shot. It is not intended to be a one shot, but rather a very small scale introduction to a whole world. Lethality is absolutely crucial in OSR games, since the whole suspense of the thing involves playing smart and taking risks with a fail condition in place: death.

It’s natural for players to be cautious when they can’t judge how lethal an environment is, and are interacting with it for the first time. I had a couple sessions like that when I first ran that dungeon, where not much happened because folks were poking around. And then it picked up. It’s not the most dynamic dungeon, it’s more a slow burn mystery introduction to an initially inscrutable world. (It could use a little more immediate pizazz I think in retrospect, I don’t think it needs to start with quite such a slow burn.)

But I do want to say this. There are almost no traps in the dungeon. There are a couple of crude bear traps, hidden rather obviously under straw beneath what is plausibly interpreted on first sight as bait. Poking around with a 10’ pole would literally do NOTHING for you in that dungeon.

And those traps are there for a reason. One of the factions in the dungeon, the Guildless, hate the Zyanese who wear masks (they are pariah exiles). They killed the people hanging, and then used them as traps to draw out the white swine to the south. In my game, the party allied itself pretty quickly with the guildless. They’re mutes, but they understand speech and some of them can read, so communication was possible. This alliance opened many possibilities, and was also an initial source of halting information about the strange world around them.

About handling lethal environments: the crucial thing is that the players always know when they’re risking something. No player should ever die from something they can’t avoid. When that happens, it’s always a failure on the part of the DM. My players right now are very skilled. As a result, they handle dangerous environments with relative ease. They’re as careful as they need to be, but aren’t slowed down by it really ever. So it’s also a player skill thing, and finding a rhythm with a DM who you trust.

Again, this is only the beginning of what would be a very long answer if spelled out properly. I’d be happy to answer further questions.


It seems to me that the main missing piece of the puzzle here is a standard aspect of play in the “old days” that has since more or less disappeared:


Gygax wrote, “… you cannot have a meaningful campaign if strict time records are not kept.”

If you look at any really old ruleset, you’ll find very specific rules for exploration turns, light exhaustion (how long do those torches last?), and wandering monsters/random encounters.

All of those rules served an important purpose: to put pressure on the PCs to decide how much time they would spend exploring, poking, prodding, etc, and (sometimes) how much light and/or noise they were willing to make.

When you know that fiddling around with the statue for 15 minutes has a chance of bringing down that marauding gang of Orcs on your ass, it makes “pixel-bashing” a very serious and important concern.

If you’re holding your last torch, and by the time it goes out, you’ll be trapped in a dungeon with no easy way to find an exit…

Remove those pressures, and pixel-bashing becomes a relatively safe bet for the players. Those rules were there for a reason.


I agree that it becomes safe, but it doesn’t mean that it is entirely bad (it depends what you are going for). I can still see some rewarding play in “puzzle solving” this way.

Nevetheless, I would totally incorporate that looming danger. This reminds me of…

A really cool example of making passage of time feel tense and uncertain is in an unfinished game by Emanuele Galletto called FAR BELOW. Sadly, the quickstart seems to have vanished with the advent of google+ and it doesn’t have a license in the pdf, so I am not comfortable sharing it here openly.

In a nutshell time is tracked with a dice pool. Whenever time passes (some rules state that time passes, but generally whenever players take some time to do stuff), GM gets a die (in my game I had a glass bowl and I would drop a die into it, the sound it made was a nice reminder of the looming danger). Then GM would roll the dice and interpret them for dangers. Players can clear the GM’s dice pool by making camp. This made for some tense choices during the game, similar to (although not as tense) how the Dread tower works.

Anyway, sorry about the tangent, but maybe it will give someone some ideas.

That’s really cool! I’ll be thinking about that.

But you misunderstood me a little:

I didn’t mean that messing around with interesting puzzles and traps was “relatively safe”. After all, that could be quite dangerous!

No, I meant that “pixel-bashing” - going around and messing around with every square inch of boring, unimportant scenery - becomes relatively safe, and that can really halt the game.


This is some terrific context, Ben. If I can tease out some more from you on this, I’d love to hear about the play style that leads to this rather than what Jeremy described. As he put it:

If you want to prevent your players from taking 30 minutes of play time to explore any given room… how do you do that?

I know that the designer of Into the Odd, at least, advises referees to simply tell players where there are hidden traps, presupposing that if the PCs are not in a hurry, they’re doing all the poking and prodding and careful peeking—traps only become a problem in that game if you’re under some other pressure and barreling through the dungeon willy-nilly. But I have had experiences with other GMs who expect a lot of perception rolls, or a lot of very detailed description about how you search, waiting for that moment when a player forgets to be careful so they can pounce. As a result, until I get a sense for what kind of GM I’m playing with (or simply get bored), I fall into the hyper-cautious pattern Jeremy describes, and I know this isn’t an uncommon problem. How do you signal to your players that they can leave their 10’ poles at home and see those traps as part of a wide world rich with opportunity?


I’m not Ben, but I have played in his game (only 1 session in the Inquisitor’s Theater), fancied up Ultan’s Door’s maps, played in games with him and I think GM’d for him a few times. I’ve also done a lot of thinking on how and why I enjoy classic style games, especially B/X and 0E D&D. The engine of a classic dungeon crawl - that which builds tension - is a set of exploration mechanics specifically: Combat Lethality/Factions/Unbalanced Encounters, Resource Management/Timekeeping/Random Encounters and Secrets/Puzzles/Moral Play. Without these things, especially that second category Dungeon Crawls don’t work too good. There’s a reason people who play different play-style largely elide past exploration or find it frustrating.

I could explain in detail here - but I’ll leave the specifics for Ben. I do think that my most recent posts on this theory blog about the Risk Economy might help some though.

P.S. If you are using perception rolls to run Ultan’s door you are not using it as intended.


I agree with Gus.L about the “engine of a classic dungeon crawl”. The tension and consequences from choices is part of the meat and drink for the OSR-style play. Interaction is key too…like the first room in the example with the hanging masked corpses, straw, fountain…all those questions you mentioned: “Is the fountain just a fountain or something special? Does tossing a coin into the fountain do anything?”…etc. In my opinion, getting players to start asking themselves those type of questions is a sign of a great adventure because it provides the interaction component. It makes the players want to poke around and experiment with things to answer their own questions. It causes the players to look through their equipment and figure out what might be helpful in situations and if they should waste a resource now or save it for later when its more of a life-threatening situation and when would they know its life-threatening…and blah blah…tension! its great!
The “smart, careful play” statement I’m not sure I agree with as I think that translates more to ‘experienced’. As you play OSR-style, you begin to pick up clues, like the ‘recently disturbed straw’. That’s a clue that will prick up your ears after falling victim to a trap before. No one wants to poke a 10’ pole around everywhere…you are correct that it does eat up time and can get extremely boring. In my opinion, its clear that the designer doesn’t want that either and that’s why he left a clue with the disturbed straw. Traps that only injure without clues have a place–mainly for resource management but should be extremely rare. I’m referring to hallways with a pit trap. Trapped chests don’t usually have clues and thats ok in my book because most people always check chests for traps…at least in OSR-style play.

They would record down to the minute in dungeons. This is also why there are wandering monsters.

Modern rpgs have made time a meaningless currency now. In D&D now, the “optimal” strategy is to move 5 feet, check the floors, walls, and ceiling for traps. Then move another five feet.

While we’re thinking about time as a factor, one other solution–admittedly an extreme one, and appropriate mostly for convention “tournament” style play–is having an out-of-character time limit before something bad happens or an opportunity is lost.

Fortress of the Ur-Mage for D&D5 / Dungeon World makes use of this form of pressure, and it’s an absolute blast if that play style is compatible with your group!