Category Archives: Amber

Amber in Chains

So, based on a convergence of guests and available free time, I ran a game last night. I had done very little prep to begin with, but I ended up improvising madly as the number of expected players jumped from 3 to 6. With 3 players I had been considering just trying some 2 Guys with Swords, modified for 3, but for 6 that wasn’t practical, but I also didn’t want to run straight Leverage – I had gotten a request for some supers elements, and what’s more I’m not entirely sure how it shakes out with 6 players. So, looking at the potential spread of players, their overlapping familiarities and my needs, I very hastily threw together a Leverage hack, bolting on many pieces from Smallville, to run an Amber game. Sort of.

The premise of this particular Amber filed the serial numbers off, both in and out of game. At a high level, one of the princes had won, and ruled Amber as the Sun King with an army of elementals at his beck and call. More problematically, he had also wiped the names and much knowledge of his siblings from the universe with his destruction of the pattern and assertion of the new order. Players represented the underground, those with enough knowledge to know there’s been a usurpation, looking to find and restore the nameless. One of them was said to be held by one of the lords of Amber, and the players took advantage of the wedding being arranged in the manor to try to find it.

I won’t delve too much into what happened, but it went spectacularly, with one player taking advantage of a recent widow to get an invite (all the while being urged by her husband’s ghost to kill her) , and with complications that spiraled out of control, including one dramatic romantic proposal accompanied by peacocks resulting in roast peacock getting added to the menu. Things culminated with the discovery that the whole manor rested on the shoulder of the Titan, his release and the subsequent destruction of, well, the whole building. I got at least one request to return to the setting at some point, and with a bit of polishing to the system, I think I’m inclined to do so.

System wise, I stole the four stats from Road to Amber (Force, Wits, Grace and Resolve) and a new set of roles: Soldier, Scholar, Tinker, Priest (which should have been Courtier), Ruffian and Hunter. Straightforward enough, but Priest ended up doing more heavy-lifting for rolls than anything else, perhaps a bit too much. Something I need to keep an eye on.

I let the characters take five distinctions. In retrospect it was too many. I think that many distinctions works with no stats, but in conjunction with the stats they just made for a little bit too much, especially for freshly created characters. It wasn’t a huge problem, but it’s something I’m going to note.

I also tapped into Smallville, loosely, and had each character pick a bloodline (effectively a heritage) and a gift (effectively a superpower). The two could be related, but did not need to be. These were pretty much entirely created on the spot, and they were structured after powers: three special things you could do with a plot point. As such we had:

  • A scion of the house of the moon, who could turn into a shadow
  • A Scion of the Titan, who was a badly trained mage
  • A Scion of Mandrake with an entourage
  • A Scion of Karm who could read thoughts
  • A Scion of The Hanged Man who had Major Arcana related tricks
  • A Scion of Feldane who mastered ghosts

Mechanically (and time-wise) this was the hardest part because I pulled these out of the air. For some of the poweres I just riffed off Smallville, but for the bloodlines, i was totally making things up. Worse, because I was working fast, I forgot to include any abilities based on exploiting opportunities (what you can do when the GM rolls a 1). If I take the time to go back and retune this system, most of the effort is going to go into making the bloodlines and gifts cleaner and easier to use.

Lastly, I hybridized Smallville conflict resolution by adding stress pools for Hurt, Tired, Confused and Upset. They worked like Smallville (opponents can roll them against you, the value gets too high, you’re taken out) but I skipped Smallville’s ‘damage’ roll and used this rule: Look at the third highest die you rolled – you inflict stress equal to that die size or increase the stress pool by one step, whichever is higher.

With that out of the way, I’ll say this – All the good things about running Leverage very easily transitioned over to this. Chargen was a bit rougher than Leverage, taking most of an hour, but that was a function of the new system and the necessity of me making stuff up. Play was full of subplots and complications, but was still all wrapped up in about 2.5 hours of actual play. With 6 players and the amount that got done, I was pretty darn pleased. What’s more, it was very low stress for me as a GM. I went in with a few ideas, but by and large I just let the complications do the work for me (which they did, with extreme prejudice).

I did use the “Complications start at d8” rule, allowing myself a free budget of d6s, and I think that worked out very well for one specific reason: When players wanted to use distinctions that seemed a little dodgy, I asked them to justify it, and often used those justifications to add some assets or descriptors to the table.

I also used a slightly different method of creating assets, where I let players create “permanent” assets – ones that last the whole game – for one plot point, but only if they went on the table, rather than remaining under the player’s control. That let the players use these assets, but also let me turn them around to use later to complicate things (which I, of course) did. as a technique, I really liked this as a very organic way to handle players stretching their distinctions.

The bottom line of this for me is that I need to work a bit on my Leverage hacks to help make sure I have more of a toolkit on hand to draw from rather than need to totally make stuff up, but with that done, man, Leverage remains a super-potent go-to for a fast, engaging game.

The one tip that I’ll add is this: While the system can support open-ended play, you’re only going to get the speed benefits by having a clear goal in play. In leverage, that’s baked in. For this, I needed to make sure it was inserted (recover the nameless). In the absence of that, we could probably have played all night, and that would have been fun, but it wouldn’t have had anything like this kind of finish. The goal drives play, but it also gives you a stopping point – that’s incredibly potent and important.

Bonus Lesson: Safe Harbor and Consequences

Ok, so something came up in discussion of the Lessons of RTA that probably merits some discussion of its own.

One of the biggest problems with any system of resolution on a MUSH is that it’s a low-trust environment, by necessity of being broadly multiplayer. Small groups may form to play in isolation, and within those groups the play may be very high-trust, but that breaks down for any sort of public play. Public play[1] is essential to a MUSH (because otherwise, why not just play remote tabletop?) but at the same time it introduces most of the logistical complications for the system.[2]

Getting trust in a public context is rough. A referee can handle it, but that’s labor intensive and subject to calls of bias and other nonsense. System can handle it as well, but doing so requires clarity in what the system can and can’t do. If the system allows for undesirable outcomes, players will avoid it. To provide an example, if the system allows for an outcome where you might win by enough to cut my head off, I am likely to assume that is the outcome you will go for. This is basic risk math – if a potential outcome is overwhelmingly bad, I need to prepare for it, even if the likelihood is low, because risk is based of likelihood AND cost. Because death is the ultimate bad outcome, it’s price is effectively infinite, so even if there’s only a fractional chance of it happening, it still looms largest in my personal math.[3]

What players really need is a degree of “Safe Harbor”, a certain baseline of protection against bad actors. This could be a simple set of rules, like “Characters can’t be killed, crippled or removed from play without their consent” but as soon as you lay down such rules, people start to weasel around it. They try to find ways around those limitations, or apply social pressures to force people to give their consent. In response, people refuse to consent to a broader and broader circle of things by virtue of indirect impacts (“You can’t kill that NPC because it would socially cripple my character!” and such) and we find ourselves back at square one – little trust and little fun as a result.

Now, at the risk of sounding tangential, I’m very pleased with the RTA stat system. The stats (Force, Grace, Wits and Resolve) are all about how you do something more than the specific details of what you do. Force could just as easily be strength as it could be kung fu, or just luck in dropping heavy objects. Wits could be intellect, cunning, situational awareness or god knows what. The bottom line is that there is a mechanical core (the value of the stat) that offers guidelines for a very broad range of color (a term for description of actions and things) surrounding it. I mention this because I think this same logic could help the trust issue.

Mouseguard has a wonderful system of consequences as a result of conflict. There are a small set of statuses a character might acquire as a result of a conflict, things like “Hurt”, “Sick”, “Angry”, “Tired” and so on. There aren’t a lot of them, and while they have very specific mechanical effects, the color is entirely flexible. “Hurt” could mean a limp, arm in a sling, a bandaged head or anything else you can think of within that general definition of hurt. This is not as fine a grain as a lot of injury systems, but there’s a good case that a finer grain is not necessary.

I think this could probably be stolen for a MUSH (or LARP) pretty effectively as a way of creating implicit safe harbor for conflicts. If there are a list of explicit consequences (with mechanical effects) then players know exactly how bad it can get, and with the flexibility of the color, players can protect their concept. That is to say, if Derek and Anne get in a fight and Anne wins, the rules might dictate that Derek gains the “hurt” status[4], which will have certain mechanical effects, and last a certain amount of time. Anne controls the mechanical element (whether or not Derek gains the status) and may use that as a point of negotiation. Derek has authority over the color (though he will hopefully take the events of the scene into account) so he ultimately can decide whether one of the cuts he took was meaningful or if he wrenched his shoulder dodging the last attack or whatever else he can think of.

Now, this works great for swordfights, but this also opens up social conflict in interesting ways. One big problem with any social conflict system on a MUSH is that players are intensely attached to player autonomy, partly because of the nature of the games and partly because – historically – a lot of games in the past have had strong streaks of mind-control sex and other potentially squicky behaviors that leave people super wary. In the risk/reward calculation, surrendering autonomy is possibly even worse than death, so the mistrust is even more intense. But if losing a social conflict means a character is “angry” or “flustered”, but the way to play that is still clearly in the hands of the player, I suspect you might find people a little more willing to mix it up.

So, I lay this out in the manner of an exorcism. I do not need to be spending brain bandwith designing MUSH systems, so I leave my thinking on the table, and go get a cup of coffee.


1 – Sometimes cynically referred to as “Bar Play.” It’s an old joke, but most public scenes on MUSHes tend to happen in bars because they are one of the few places that people can legitimately just hang out until something interesting happens. This is one of the consequences of a lack of direction – if you want to play but have nothing you want to accomplish (or more importantly, no way to actually proceed with it) then going someplace public and hoping to bump into people tends to be the best option. This is probably not ideal, but it’s a reality of the structures of play, and it has some upsides – for many players the socialization element is the reason they play in the first place.

2- At its most basic, it makes it very difficult to delineate when a scene begins and ends – people come and go, and events rise and fall. That’s pretty hard to hang any sort of scene-based mechanic off, and the ability to begin and end scenes cleanly is essential for a lot of the more clever rules out there that might otherwise be very helpful in a MUSH context.

3 – There’s a whole lot of fascinating psychology to this, but the bottom line is that we are far more sensitized to dramatic, extreme risks than we are to ones that seem low key. Consider the fear of flying: you’re much more likely to die in your car, but we’re used to our cars, and a car crash is less dramatic than a plane crash, so we discount it.

4 – There might even be more variety. Maybe Derek can fight defensively and get the “Tired” status instead (or put “Tired” on Anne). Maybe Anne has a gift that lets her poison her blades, so she can choose for Derek to be “Sick” rather than “Hurt”. The mechanical options are broad and fascinating (especially in how it relates to types of fights and potential tactics and abilities) but would start turning into full-bore game design from here.

Road to Amber, Lessons 6 Through 10

Following up on Tuesday’s post with the second half of the lessons from Road to Amber.

Lesson #6. People Lie Without Even Knowing It
If there is one single thing that makes staffing a game absolutely maddening, this is it. I’m not even talking about the people who intentionally lie about things (though there are plenty of those) but rather people who are sure they want one thing, but actually want something entirely different. This comes up most often in the context of risk and conflict, but it can really come up in almost any context.

The simple truth is that the least reliable yardstick of what people are looking for in play is what they say they want. There are exceptions, of course, but you will drive yourself mad trying to find them. Ultimately, you need to apply a skeptical ear to requests until you’ve had time to see people in play (and specifically, how well they can handle being in a bad position).

Lesson #7. Ownership is Powerful, but Stagnates Over Time
One of the most effective ways for staff to delegate responsibility is to give the control of setting elements to players. These prop controllers (“PropCos”) have the autonomy necessary to make the decisions regarding matters that might come up in play without needing to consult staff. This tends to be awesome for a while, but eventually the situation will come up where the prop may be endangered, and the player is put in the unfortunate position where he might end up diminishing the prop if he allows that. Since that would mean diminishing the potential for fun in the future, he makes the rational decision to preserve the prop. Thus, with every good intention, we start a decision-making chain that ends in the prop never changing, ever.

This one’s pretty hard to shake once it sets in, and most attempts to do so result in bad feelings and other unpleasantness. You can try to establish strong rules for what owners can and cannot do, but those are only worth the paper they’re printed on (so to speak). I sometimes wonder if there might be some way to subject props to regular review, but that only feels like half of a solution. I expect this one to bug me for a while yet.

Lesson #8. Nobody Reads Anything Longer than an Elevator Pitch
An exaggeration, sure, but it speaks to an important point – people want to get data quickly and get on to the next thing. Of all things, twitter makes for a good set of guidelines for this: ask how you would express an idea with explicit limits on the number of words or characters you can use.

Lesson #9. It is Easier to Ignore Direction than to Find It
The greatest problem for players in any large game is answering the question “What do I do now?” Some players excel at answering the question, and can enthusiastically make their own fun, but even the best of them will occasionally find themselves bogged down. In those situations, it helps if there is some default action to take or direction to go, something that can be picked up and run with. There’s a lot of resistance to this idea since it tends to manifest as “metaplot” or “Railroading” but unless the GMs are forcing it down people’s throats, then this is a good thing to have. The people who need it can tap into it, and the people who don’t can ignore it.

The dark side of this is that the most effective kind of direction takes the form of a problem. In and of itself that’s not bad, but it becomes a problem when players come to resent problems and want them to go away, not realizing that in doing so, they are taking away their direction. They play and solve the problems of the day, but unless they’re willing to embrace new problems, they are killing their own fun.[1]

There’s no good fix of this, short of people realizing that problems are what fend of boredom in fiction, and that’s not happening anytime soon.

Lesson #10. Finish 100%, Show 25%
This is sort of the flipside of “Prune, Don’t Seed”. Have everything done and in hand, but don’t show it all. The image that the players will create from incomplete information will be vastly cooler and more compelling than whatever your idea was, and the only thing that will ever come from a full reveal is disappointment. This is total man-behind-the-curtain, illusionist sleight-of-hand but it is incredibly potent and incredibly important.

This is one of those ideas that I need to step back from a little because i find it profoundly self-evident. That’s great for me, but it makes for a piss poor explanation, so let me break down the benefits.

  • Because players are figuring things out, they’re more invested than if they were just told
  • Because more players will be thinking about it, they will generate more ideas than any one writer ever could
  • Because you’re not showing it all, you can change things if the players come up with a cooler idea

Of these three, I cannot overstate the potency of the first. People are, by their nature, very strongly invested in conclusions they come to on their own, right or wrong. If you can tap that, your game stands to benefit a lot.


Extra Bonus Rule 0: Work With Ninjas
The simple reality is that Helix, God of the MUSH, is a coding machine, capable of creating entirely new systems and doing so in an incredibly short timeline. That allows rapid prototyping, which is incredibly useful.

However, Helix is also one of the kindest, smartest, and also most pinpoint-orbital-strike-weapon-like people it has ever been my privilege to call a friend. She was a joy to work with, and that underscores the most important point. The people you work and play with may have all the skills and talent in the world, but if they’re not people you enjoy, none of it is worthwhile. I don’t regret my departure from RTA, it was the right decision, but I do miss the collaboration.


1 – I feel like the Lorax here, if he were a nerd rather than a hippie.

Road to Amber: Lessons 1 through 5

As noted in yesterday’s posts, I’m drilling down into the lessons learned from Road to Amber. The first half get run through today, and hopefully the rest come tomorrow.

Lesson #1. Everyone Wants to Control the Network

I blame Neal Stephenson. Basically, if you introduce the possibility that players can control some manner of in-game resource, there will always be a few who want to control actual things, like goods or armies, but the first thing that most geeks will ask for is the network. They want to control shipping or the bureaucracy or some other thing whose power comes from the ability to direct and control other things. This is not a huge shock: gamers tend to be inherently Jominian, and this is where their interest lies.

The problem is that they also tend to think they’re being incredibly clever about it, and will often make a big show about how the thing they want is clearly boring and unexciting. This I kind of novel and amusing once, but eventually it just makes me want to bash heads in with rocks. Thankfully, I was given some perspective on this recently. See, my son loves these little dried fruit snacks we give him, but we need to be careful not to let him see the bag. He’s figured out that the bag is where the snacks come from, and he will ignore snacks in favor of pursuing the bag. He’s 8 months old, so I’m pretty proud of his reasoning skills. But I figure anyone older than that needs to realize that this clever idea is one that can be essentially grasped by an infant. Now, when I encounter that phenomena, I just imagine Jamie diving for the bag of snacks, and I smile.

So what’s to be done about it? My instinct is to allow it, but set things up so that they understand why the bureaucrats and shippers tend to get the short end of the stick in any real conflict, but that’s just mean. The reality is that if there’s something that people want that much, it should be enabled. The trick, I think, is to make it more dependent on other players – if you want network play, then nodes really need to be other players.

And that’s sort of the rub. Network play tends to be pretty boring in and of itself because players don’t *need* the network. It’s more fun (and more profitable) for them to horse trade directly among themselves. If I need iron or horses or whatever, why not go directly to the iron or horses or whatever guy? The upshot is that the network guy only ends up working with a very generous player, one who is committed to connecting other people. That’s rewarding, but it’s a lot of work. There are other solutions: network guy is often treated as controlling things no one else cares about, but the myriad ways that can go wrong are just overwhelming.

All of which is to say, I don’t really have a solution for this one yet, except that it’s something you need to keep an eye on. If you’re going to allow PCs to take network roles, then you need to have some reason for people to seek them out as connectors. There may be a temptation to give them more information, but that’s a bad plan, as that tends to turn them into information brokers. That’s cool for them, but it’s a whole other niche.

Ideally, if you have a way for currency or resources to change hands, the network guy should be desirable because he makes such transitions more efficient. Consider a game of merchant prince’s, where one family is dominant in shipping. When two other houses want goods to change hands, only 50% of the goods are transmitted (representing cost of shipping and such). If you can get a member of this third, network-ey house to sign on, then 75% are transmitted and the shipper gets some percentage, either fixed or negotiated[1]. This works because there’s a clear benefit in going to the network, but at the same time if the network is an asshole, he can’t *stop* play, as much as he might want to. As a bonus, our network guy is dependent on the other houses for his resources, so he’s drawn more tightly into play.

Lesson #2. Focus was a Win
So, in addition to experience points, players accumulated another currency called “focus” it was accumulated fairly quickly (about a point a day) but it capped at 10 points, so there was strong incentive to spend it rather than let it go to waste. Generically, Focus represented the effort a character put into things when he wasn’t playing, so it could be spent to study things, build things, influence events and so on. The specific things it could be spent on were cool and all, but the real triumph is that this tapped into the part of our brain that loves gaining and spending points. People enjoyed the very act of earning and using focus, and really seemed to enjoy that they could regularly do things with it.

From this I take the lesson to always include something like this. Whatever form it takes, the main idea is to have something that comes more frequently than advancement, and which can be spent for concrete things.[2]

Lesson #3. Complexity Should Emerge from Play, not Setting
This is the nicest possible way to say that players are not interested in nuance.

When we started things out, I wanted to noble houses to feel like part of the setting, not like one-note caricatures (this is the merchant house, this is the military house, etc.) The same thinking was applied to the creation of the Golden Circle shadows (the alternate worlds near Amber). In both cases the idea is that there might be a strong theme but that it would just be the strongest note in a more complicated piece of music.

This was a knee jerk response to too many years of seeing organizations composed of stereotypes as the norm, and it was just a bad idea.

The reality is that players want the stereotypes. They want the groups to be simple enough to summarize in a word or two. This sounds awful on the face of it, since it seems to suggest that players aren’t that bright, but that’s not the problem. The real issue is that they want clarity, because clarity provides them with clearer impetus for play, and play is the ultimate goal. Creating a simple foundation lets players create exceptions (and oh how they love creating exceptions) and lets the complexity emerge from their play.

This is a pretty simple lesson to apply since it actually calls for less work on the part of the designers. You don’t need to flesh things out – paint them in simple stark lines, and players will (hopefully) take care of the rest.

Lesson #4. When Rushed, Prune, Don’t Seed
The initial rollout of RTA was slightly hurried, partly out of necessity and partly out of the discovery that people were already playing. That meant in a lot of cases we had to plant a flag in the ground where an idea was going to go, but just leave it at that with the expectation that we’d get back to it. It seemed necessary at the time, but as the game went on it became clear that with the pace that new things needed to be dealt with, the time and resources necessary to go back to those seeds was not always going to be available. This would be incredibly frustrating when a player found one of these flags, but it simply wasn’t ready. Either we’d have to sprint to fill it in, or disappoint the player.

The alternative would have been to roll out with less initially – have fewer options available and fewer things going on, limited to those things that were fully fleshed out. That seemed very unappealing at the time, but in retrospect I think it would have been a much better idea. It would have helped forge a stronger central idea of what the game is about, and it would have allowed for staged release of new elements. The fact that this ends up looking like the average MMO release schedule is probably not a coincidence.

Lesson #5. Watch What People Use
Helix[3], bless her soul, loves statistics and hard numbers, and that meant that the game was set up to do really robust anonymized reporting on various game statistics. This meant we could see what people were spending on, what mechanics were being used or not, and generally speaking how people were interacting with the game. This made it incredibly easy to see what wasn’t working, and it also helped to identify gaps that needed filling.

All of which is to say, put some thought into transparency up front. This is not limited to electronic media – do people take home their character sheets? Do item cards change hands? What questions are people asking GMs? Where is play happening? Where is it _not_ happening? These kinds of questions are the ones you want to stop and ask yourself during a game so that you know what’s going to work better in the *next* game.


1 – Negotiated is really interesting since it tends to drive people towards their “fair play” reactions. The optimized transporter might say “You can transport 51% and I will keep 24%, everyone wins”, and while he might be able to sell that to some, most folks will tell him to go piss up a rope. The marginal advantage is not worth the benefit he gains, even if it costs them nothing. Of course, if other members of the transportation family can offer cometing bids, it gets all the more interesting.

2 – Arguably, gold might fulfill this role in some games, but that depends a lot on the surrounding structure.

3 – Chief Poobah and Grand Code Wizard of RTA. Also the ninja referred to in the still-mysterious rule 0.

10 Lessons Learned from Road to Amber

So, a while back I helped create a MUSH[1] called Road to Amber. It’s still going, and you can read more about it on the wiki, but I’ve since bowed out of my role in it due to the time constraints of changing jobs and giving birth to my first kid. I’m pretty proud of what we accomplished, and I think there were a number of ideas in RTA that are going to see use in other games down the line. However, there were also a lot of lessons I learned from what worked and what didn’t work. I’ve been meaning to write them down for a while, and this seems like the venue.

Top Ten Lessons of Road to Amber
1. Everyone Wants to Control the Network
2. Focus was a Win
3. Complexity Should Emerge from Play, not Setting
4. When Rushed, Prune, Don’t Seed
5. Watch What People Use
6. People Lie Without Even Knowing It
7. Ownership is Powerful, but Stagnates Over Time
8. Nobody Reads Anything Longer than an Elevator Pitch
9. It is Easier to Ignore Direction than to Find It.
10. Finish 100%, Show 25%

Bonus rule 0: Work With a Ninja.

Vague enough? This is a bit of a teaser, I admit, but I’m laying these out here so I can start getting into them over the next couple of days.

Edit: And here they are…

Link to Lessons 1-5
Link to Lessons 6-10+1
Unexpected Extra Lesson


1 – A MUSH is an online game played in a text environment – think of it like the old infocom games, except other people are also walking around in it, and play centers around talking and interacting with those people. This type of game has been around for about two decades now, and RTA is far from the first Amber-based MUSH. Amber gives itself well to the flexibility of the medium.

Structurally, it has a lot in common with LARPing, so some of the lessons can migrate back and forth between the two.