Tuesday, 24 September 2013

Looking for Ogre

So both Dan and Arthur are talking about Quantum Ogres, which inevitable prompted me to waffle about it myself. This is a very OSR-centric debate, which is not particularly my area, but that never really stopped me opining before.

Quantum Ogres seem to originate in Beedo’s 2011 post here. Essentially, given a choice of three Woods, the Quantum Ogre always turns out to have been in the first one you explored. This idea prompted a number of posts, many of them critical.

Dan’s article is prompted by this one by Cat at Hack & Slash, which is less focused as it's mostly a response article. Cat asserts that:

  • Quantum Ogreing is bad for players because "the players never have the experience of choosing correctly and skipping right to the end (which is fun for them)"
  • Fun comes from exploring objective worlds and one-in-a-million rolls rather than from lovingly-crafted encounters, so Quantum Ogreing does not increase fun
  • Players will grow to resent Quantum Ogre encounters and will become aware of them
  • Quantum Ogreing as a time-saving measure is a fallacy, because you can effortlessly do all the prep you need by cribbing and using random content

Kevin at KORPG Games discusses whether or not Quantum Ogres can be considered a false choice. Essential quote: "Remove the foreknowledge of the possible outcomes and you’ve removed all knowledge of the palette shift… and with removal of that knowledge, the palette shift itself is actually removed because the knowledge is a requirement of the shift by definition. What robs the game of fun is when the palette shift is known and is in direct opposition to the meaningful choice expectation of the player."

Dan mostly talks about the nature of choices in RPGs and how this interacts with game content and with player fun. He also has some interesting points about the validity of random encounters versus human-genned encounters.

Arthur discusses the nature of interactivity and the different assumptions people have about it, and the way this relates to their games.

I will instead present a rambling set of loosely-connected ideas with no firm conclusions. As usual.

What is a Quantum Ogre anyway?

People have a couple of different takes on this. Beedo, the proposer, says: "Regardless of which woods the players choose first, he'd like the party to have the opportunity to encounter the ogre. The MacGuffin will be somewhere else."

Commenting articles seem to have read this as something like: "Regardless of which woods the players choose first, ... the party ... encounter the ogre. The MacGuffin will be somewhere else." This is subtly different. Moreover, quite a number of people read this in an even more heavy-handed way, bringing in assumptions that the party will a) confront the ogre; b) in the manner prescribed by the DM; c) regardless of their wishes.

Cat does this very specifically in a comment on Beedo's next article.

The problem is not even when the DM plans to put the ogre encounter in front of the players! There is nothing wrong with this. In a game where the players can go anywhere, improv is necessary and not necessarily abhorrent.

The problem comes when the players attempt to gather information, and then the DM becomes evasive, blocking their attempts because he is vested in his encounter. This leads to the known bad case of the McGuffin always in the last wood.

Leaving aside the interesting question of whether "the MacGuffin always in the last wood" is necessarily bad, I think it's worth noting that Beedo does not seem to hold any of these assumptions. I should also note that Beedo later decided "I ultimately came around to the position that any kind of illusionism is a bad practice" in a series of edits, though the reasoning wasn't clear to me and seems to be part of some OSR school of philosophy I'm not familiar with.

I am happy to agree that actively thwarting player choices is a bad thing (though there may be some circs where I'd accept it), and the related idea that forcing players to handle a situation in a prescribed manner is a bad thing. Both of those tend to reduce players to dice-rolling machines making purely cosmetic decisions about their reactions to your plot, and controlled by a mixture of DM fiat and random chance. I can do that more efficiently by playing a CRPG.

In contrast, to me Beedo's original Quantum Ogre seems to be saying that the party will always a) become aware of the available possibilities; and b) have the least final option presented first, such that it can make IC sense to tackle it rather than bypass it. Going back to Dan's point about IC and OOC goals, this seems a pretty good way to satisfy both the IC goal of finding the artefact, and the OOC goal of fighting the ogre, without having to compromise IC motivation. In contrast, an artefact-hunting PC given the option of Ogre Wood or Artefact Wood has little motivation to visit the ogre, nor does a PC who finds the artefact first through a blind choice or through a random table result.

In general, the Quantum Ogre discussion seems to have rapidly wandered off into what I might instead call the Massive Roman Achilles Ogre discussion: all roads lead to the Ogre, the Ogre's gravitational field is too large to escape from, and the Ogre can only be overcome in one way. This is, quite frankly, a completely different issue.

The Quantumness of Ogres

Dan makes some points about how random encounter tables are not necessarily qualitatively different from DM-decision outcomes. The dice thing is kind of hard to get your head around, but... okay, here’s some examples vaguely related to it.

  1. A DM plans three encounters: Nothing, Ogre and Artefact. These are allocated to Woods A, B and C respectively. (planned known geographical)
  2. A DM plans three encounters: Nothing, Ogre and Artefact. These are allocated to First Wood Explored, Second Wood Explored and Third Wood Explored respectively. (planned known chronological)
  3. A DM rolls on a random encounter table and gets Nothing, Ogre and Artefact. These are allocated to Woods A, B and C respectively. (random known geographical)
  4. A DM rolls on a random encounter table and gets Nothing, Ogre and Artefact. These are allocated to First Wood Explored, Second Wood Explored and Third Wood Explored respectively. (random known chronological)
  5. A DM plans three encounters: Nothing, Ogre and Artefact. These are put on cards, shuffled and allocated to Woods A-C, so their order is unknown to anyone. They are revealed during Wood exploration. (planned unknown geographical)
  6. A DM plans three encounters: Nothing, Ogre and Artefact. These are put on cards, shuffled and drawn in order during Wood exploration, so their order is unknown to anyone. (planned unknown chronological)
  7. A DM uses a blind random selection method to obtain three level-appropriate encounters from an arbitrarily large set. Unbeknownst to the DM, they are Nothing, Ogre and Artefact. The cards are allocated to a particular Wood A-C and revealed during Wood exploration. (random unknown geographical)
  8. A DM uses a blind random selection method to obtain three level-appropriate encounters. Unbeknownst to the DM, they are Nothing, Ogre and Artefact. The cards are drawn in order during Wood exploration. (random unknown chronological)
  9. There are three Woods A-C. When players express an interest in exploring a Wood, the DM invents an encounter out of whole cloth. (planned spontaneous chronological)
  10. There are three Woods A-C. When players express an interest in exploring a Wood, a random encounter is rolled from an arbitrarily large set. (random spontaneous chronological)

Which of these undermine player agency? Which are okay? Which enable meaningful decision-making?

I'd argue that mostly, we don't have enough information to judge. How far in advance will the DM check "unknown" results so as to provide clues and details? How easy does the DM make it to shortcut or circumvent results that the players don't want? What do the different 'encounters' even mean? There is an enormous difference between flipping over a card as you approach the wood, and then providing all the relevant details, clues and even NPCs necessary for the players to make meaningful decisions about whether and how they wish to interact with it, and flipping over a card to determine whether they find an artefact in a clearing or are attacked at night by an ogre.

I'd also make two broad points. Firstly, the more random a result is, the less likely it is that players (or the PCs) can predict it based on existing knowledge. This may impact their ability to make informed choices. Secondly, the more unknown a result is to the DM, the more difficult it is to prepare the ground, and thus the more difficult it is for players to anticipate it and make an informed decision. This affects pre-existing encounters where the DM is unsure of their location; random encounter results that a DM knows may come up or not; and spontaneous encounters.

In contrast, a DM who knows precisely what is coming up (either chronologically or geographically) can enable a firm decision from the players. Moreover, in a group with a strong DM-player relationship, players can often judge from metagame clues and OOC chat the direction things are going, and make their decisions based on those factors. As such, I am actually inclined to say that randomness undermines player agency, while prepared aspects enable it.

There is another point to make here, which is that a DM assigning content on the fly (whether premade or spontaneous) has the option to twist that content to suit the state of the party and the group's mood at that specific point. While this can always be done with small details, enabling a frustrated group to skip filler content or a badly injured party to not wander into the trap-ridden ettercap settlement just yet seems to me a valuable option and one that, as a player, I appreciate. I don't want to play a game where even if frustrated and bored, I have to spend an hour of game-time combing a forest to discover that it contains nothing of interest; and if I've just fought my way through three random encounters, I might well prefer not to blunder into yet another fight with an Ogre just because a DM had decided there'd be one here.

Cat's article

Cat's article includes a couple of arguments I wanted to discuss, while hopefully not descending too far into just picking at someone else's post.

Ogres should be foreseen and not heard

If you always pre-ordain 'your precious encounter' then the players never have the experience of choosing correctly and skipping right to the end (which is fun for them).

This contains a fairly major assumption: that "skipping right to the end" is a) fun, and b) probably more fun than encountering the ogre.

Let me be unreasonable for a minute, while also calling on Dan’s example of the treasure-hunting PC. A really canny bunch of characters given enough information can work out where treasure might be, and go straight there evading all ogres. Many adventurers would like nothing more than to stroll from hoard to hoard, stuff their pockets with diamonds, and then retire to a life of indolence.

From my perspective as a player, though, that would be incredibly boring – it’s the equivalent of a mystery game where you immediately identify and arrest the culprit and then drive off into the sunset. I have yet to sign up to a campaign on the premise of “plucky heroes collect artefacts from the places they would logically be and retire wealthy after a minimum of hardship”. In fact, I would almost never choose to avoid the ogre unless I had reason to think that encountering the ogre would be less interesting that other things I could do with the time. Honestly, “skipping to the end” doesn’t seem especially fun to me, apart from the brief satisfaction of having correctly interpreted some clues. Quite often, I don’t want any particular outcome to a situation, but simply want it to be fun. To get into some nasty grammar: if it is going to have been more fun for me to have run into an Ogre before I found the treasure, I want the DM to have interposed an Ogre. Usually! And that, of course is the other thing. We don’t necessarily know what we want. We don’t necessarily want the same things consistently.

The other point is the flipside of end-skipping: if you present three woods, where one contains an ogre and one contains the treasure, then there is at least a one in three chance that players will pick the empty wood first, and if you are handling these as though they are equally important, a considerable amount of time could be spend on it. That is okay if the group are interested in playing some people exploring a forest, but if they are keen to get on with quest-relevant activities, then it could be very frustrating. On the other hand, if you dispense with the forest in a few sentences, then was it really worth including it at all? Why does this forest even “exist”, from an adventuring perspective? That’s not to say I couldn’t enjoy myself just exploring a forest, but these things vary between players, genres, and even individual play sessions. I know at least one player who generally just wants to move from dramatic event to dramatic event, without faffing about.

I suppose another thing is, if the DM (always a friend of mine) has spent time and effort devising an Ogre for me, I am generally pretty keen to encounter that Ogre. My DMs can usually be depended on to devise pretty damn good Ogres, and I’m happy both to get the chance to appreciate them, and to give my DMs the satisfaction of seeing everyone get a kick out of their Ogre.

Just occasionally, spotting a way to evade all kinds of challenges and achieve an objective can be immensely satisfying. But I want those occasions to be a matter of player and PC cunning, and the feeling that I have in fact achieved something. I don't want to have bypassed the bit of the adventure where I would do stuff because I picked the right card. And I will only want it occasionally, because mostly I'll be playing the game in the hope of encountering ogres.

Quantum Ogre and the Jerk DM

The flaw of the Quantum Ogre is that, if you have a party who plays smart, he won't be quantum long before you enter the woods, and then you've wasted time by not assigning him to a location already or you become the jerk DM where ESP doesn't work, the ground doesn't hold tracks, and if you try and teleport - suddenly anti-magic fields everywhere.

This is a reductio ad auteuram (to coin a phrase). A DM who will prepare encounters and then assign them on an ad-hoc basis so that PCs interact with each one to some degree, is by no means the same as a DM who will block tactics adopted by the players. There are very different things going on in each case. For a start, the Quantum Ogre situation does not necessarily involve good information on the choice being made. Frequently players are left making choices with only quite broad information (go to the town or the forest?), and while they can draw some conclusions about what they're likely to encounter and the opportunities they'll have, there are many things they can't reasonably predict without performing detailed research. Secondly, arranging for an ogre to be in the forest or in the town does not in itself involve contradicting common assumptions about the setting, the physical laws of the gameworld, or the game mechanics; this is a sharp contrast to banning teleportation or blocking the use of skills in order to make players handle an encounter in a particular way.

I would also say that, in small doses, preventing the use of specific abilities in a specific situation can offer opportunities for creativity and make things much more interesting, providing it's done with some kind of in-game logic and isn't aimed at forcing a particular approach, but at preventing one.

It also seems strange to me for Cat, who shrugs off the time needed for sandbox preparation (of which more later), to quibble about wasting time by not assigning de-quantising Ogres. If you have forests A-C prepared, and locations 1-3 on a map, just how much time is being wasted by associating one with the other during the game rather than beforehand?

Finally, how quantum is an Ogre? Some “ogres” can’t reasonably be predicted by PCs, even by asking locals or carrying out research. A ruin or notable monster might be learned of, but there are plenty of other Ogres that are either commonplace or impossible to discover without going there. Other places aren’t frequented by anyone, so nobody’s there to warn you about the ghost, artefact or quicksand.

A Choice of Ogres

Looking again at Dan's points about choice, you could also bring up the question of what precisely the choice is the players are trying to make.

Are they trying to choose “do we go and find the MacGuffin, or encounter an Ogre?”, or are they trying to choose one of three locations to nose around in? If it is the latter, then arranging for a Quantum Ogre is not in any way infringing on that choice. Or perhaps they might be choosing “do we go for some undemanding content and mess about in the woods, or push on to another stage of the plot?”, which is different again, since getting to MacGuffin Wood doesn’t necessarily enable them to advance the plot, nor does meeting the Ogre preclude it.

If the players choose “undemanding content”, I personally don’t see any failing of a pregenned Ogre over a random encounter; a random encounter might prove interesting and exciting, by throwing out something the DM wouldn’t have thought of, but could equally well seem nonsensical. Odd results on random tables can seem out of place narratively, either being so petty that it seems pointless to waste time on them during your epic quest, or so substantial that they shouldn’t be an unexpected diversion.

I might also say, if you’re going to argue that PCs should have the chance to learn about ogres in the wood and avoid them, they also deserve a chance to learn about many kinds of random encounter, which in effect means pre-rolling the random encounters and planting some clues. Your prep time just went up.

I think this is my main take-home message from Dan's article: “It is only wrong to force the players to encounter an ogre if they specifically want to avoid encountering an ogre.”

I don’t care one jot about any supposed loss of agency in depriving me of the ability to discover that a wood, chosen blindly out of three, contains nothing whatsoever of interest. This is not a meaningful choice (and I think the use of that term on various posts in that chain is highly misleading). For one thing, much of my real life consists of doing things that are not particularly interesting; I don’t need to do them in my gaming time too. I don’t generally feel the need to roleplay the time spend on long stretches of dull road, waiting for the kettle to boil or asleep. For another, if empty wood is called for, why not just use the 99% of Ogre Wood that does not contain the Ogre? We can do all the same things there, it’s cool, seriously.

Essentially, this choice is only relevant if I have some information that Wood B contains an Ogre, and decide that I would like to explore the other two woods in the hope of finding the artefact without encountering said Ogre. Of course, I might believe that the Ogre is likely to have the Artefact, in which case I might go there first. If there’s plenty of information indicating the location of the Artefact, then perhaps there’s no point having the three woods anyway, and you might as well combine them. At that point, the question stops being a case of Wood-choosing, and becomes a tactical one: there is an Ogre in the Wood where the Artefact is, what is our best strategy for dealing with the situation?

Another point is that people want different things at different times. Sometimes I want to see interesting content of whatever type. Sometimes I'm quite invested in finding the artefact to push the plot forward and don't care about the side content. Sometimes I'm not that interested in the plot and would rather fight an Ogre. Sometimes I'd rather explore the setting my DM has created. And sometimes I really don't care what I do IC, so long as I can spend time with my mates OOC. Arthur talks a lot about types of interaction, which is also quite relevant here.

Agency and Cooking with Ogres

Broadly speaking, in playing a game, the agency I'm looking for is ability to control my interactions with the content of the gameworld. I may choose not to explore an element as fully as the DM expected, or to dive right into it. I may not interact with it in the expected way. If I show no interest whatsoever in interacting with some content, and especially if I (as a player) show signs of avoiding it, I expect my DMs to respect that wish within reason and within the expectations of the group as a whole. I expect freedom in the way my character behaves, the ability to shape events and to affect the plot to at least some degree; though all of these are subject to the genre, the game system and the kind of campaign I signed up for. However, I am not bothered about having 100% total freedom to explore any part of the world as I choose, because this strikes me as being both a bit pointless, and kind of arsey. I am particularly not bothered about having an objective world from which I select elements to explore, because while I very much enjoy exploring the creations of my DMs, I don't think this experience would be enhanced by knowing that if I'd chosen the wood over to the right I would have discovered something different and missed out on the current wood.

As a player, I don't want to feel that I am obliging my DM to devise a vast swathe of original and interesting content so that I can completely miss 95% of it, and that I may well miss the most interesting elements (to both my PC and myself) because I chose to take the left door rather than the right. Cat (I don't mean to pick on Cat, they just happen to have written the article that mentions this) suggests that this isn't a problem: content can be quickly generated through web resources, recycling and random content. However, I'm not sure this would really fly outside the OSR. While I’m really not that experienced as a DM, I've never been in games based around walking from hex to hex, or strict geography. Groups I've been in tend to want to go somewhere specific, and are more interested in what happens there than in the intervening geography and any semi-random content that might have been allocated to it. This means it’s more important to devise narrative locations than geographical ones. The Dark Tower and the Big Lake are locations. N13E04 is a datum.

Being creative does not, in fact, make you less creative. The more you create, the more your output increases! Let's ignore that there's enough free material on the web to stock 1 millarn over 9000 hexes and dungeons with no more effort than hitting print... (Cat again)

Random unconnected stuff does not fit well in my games. I want things to make sense, both as a DM and a player, as do other members of my group. They want to know what people were doing somewhere, and why. They like to investigate the history of things, or ask about their religious significance. They generally aren't very interested in being attacked by six generic bandits, or even investigating a cave several miles away because “it’s there”. As such, it takes not just time but also creativity to come up with things that fit the location and the situation (let alone the setting material as developed so far, the tone of the campaign, and so on). It's difficult to transpose material that's quite specific, particularly into a new campaign that may not share the same kinds of assumptions. And there is a vast amount of free most things on the web, but as I can say professionally, the main problem with information is excluding the bits you don't want. If your campaign can readily accomodate absolutely anything you grab off the internet, I don't think I want to play in it. The vast majority of stuff will need adapting at the very least, perhaps rebuilding from scratch, and that means time. Time is precious.

Random encounters are not everyone's cup of tea - players I know aren't generally keen to just fight something because they saw it on the road somewhere, and predesigned tables may well not fit the tone of your campaign. There are some very large tables, with truly interesting and creative possibilities, but it’s a bit of a devil trying to make these conform to your own campaign. I’d also say that the less your group is interested in Killing Things and Taking Their Stuff as prime game content, the harder it is to come up with good random content that will enthuse people and make sense to them.

Quantum Ogres are an attractive prospect for exactly this reason: a DM can produce something interesting and relevant, and ensure that the players will see it. There is nothing about Quantum Ogres – even as defined by Cat, who dislikes them – that obliges the players to interact with them in any specific way, or indeed at all. All you are doing is removing the possibility of never knowing that the Ogre exists.

1 comment:

  1. I've been mulling this over for a while and I think I can sort of see where -Cat is coming from, at least as regards arranging encounters geographically rather than chronologically.

    Specifically, I tend to favour "chronological" arrangements over "geographical" ones because I *don't* like to keep track of where the players physically are at any given point. But a flipside of this is that I *also* don't ask the players to make choices about where to go on the micro level. Rather than asking "do you want to explore Wood A, Wood B or Wood C" I'd be more inclined to ask "do you want to explore the surrounding woods".

    I *can* see that if the players are invested in playing the sort of game where you choose where your characters go *in detail* that they should expect that choice to matter (and I agree, and I think -Cat also seems to agree) that this should include giving them meaningful information about encounters in advance: Do you want to go to Wood A, which is rumoured to contain an ogre, Wood B, which is rumoured to contain a magic sword, or Wood C, which is rumoured to contain nothing very interesting. If I then go to Wood B and get ambushed by an Ogre, then head to the Ogre's supposed lair in Wood A and find nothing, only to find the magic sword in Wood C I think I'd be justifiably narked.

    I think here the issue isn't really about quantum ogres, it's about coherent play expectations. If the players are going to encounter Ogre/Nothing/Sword no matter where they go, don't ask them where they want to go.