A Lesson in Humility

This is an exciting time for R&R development. Some pretty key system obstacles have been overcome and while there is a lot of work left to do we are entering into the stage of development where the game is moving from “functional but clunky” towards “functional and elegant” (as opposed to the stage where it was moving from “broken to slightly less broken”)

I was struck recently by an example of design convergence where 3 separate and seemingly unrelated problems all led to the same solution…and how I was guided there by taking my own advice about design goals and asking myself the hard question of “what exactly am I hoping to accomplish by having this rule at all”.


1) The first problem was with combat. Combat uses the same system as everything else so there isn’t really a separate combat system, its just an application of the normal rules for everything. The normal rules for everything for those not familiar are (in brief) an opposing dice pools / count successes system where the net successes can be used to purchase bonuses to future rolls (or penalties to future rolls of an opponent).

This works great for combat because all of the fancy exciting swashbuckling heroics are simply handled by making a roll (like “swing from the chandelier”) and then rolling those successes into the next roll (like “kick the bad guy in the head”). The successes from that roll then go to buying a penalty for the bad guy for future rolls (like “dazed due to boot to the head”). The tricky part is determining when the bad guy is actually put down. Its all well and good to buy a penalty described as “in pain from being run through with a rapier” but when does the guy just drop down “dead”? Further complicating things is my design goal to feature two different kinds of damage. Faux Damage: the robot is simply programmed to pretend to be hurt/dead, and Malfunctions: some important part actually broke. This goal is important to support the larger goal of highlighting the difference between a programmed robot acting for the benefit of an audience and a Spark with the emerging freedom to ignore its own programming. A Spark can simply decide not to pretend to be hurt.

Things had been working ok with this. I used a simple threshold where scoring enough successes in a single roll was enough to put the opponent down (or cause a Malfunction) but otherwise you just assigned additional penalties. Now here comes a great example of the ripple effect of changing rules elsewhere. In the last version of the rules the die pools were typically 8-12 dice with bonus dice taking it even higher. Handling time was the pits on opposing pools of that size. So, current pools are 2-6 dice with bonus dice on top of that. Much quicker. But…while 6 dice opposed by 3 has the same odds of pass/fail as 12 dice opposed by 6…it doesn’t generate the same number of net successes. All of the sudden it was virtually impossible to generate enough successes to meet the threshold and actually put anyone down. Simple right? Just lower the threshold. Except that I found if you lower it enough to be able to put someone down in a reasonable manner, it was so low that there wasn’t a reason to do anything else. The same number of successes it would take to assign a penalty for a boot to the head could now just take the bad guy out all together…so who’d do all the crazy cool stuff.


2) The second problem was related to Power (as in battery charge). I wanted power to be featured in the game because one of the design goals was to feature the impact a few Sparked (free thinking) robots have on a programmed (conformist) society. Power was an easy way to justify why these persecuted individuals had no choice but to act within (or against) that society rather than simply wander off into the wilderness and set up their own free society. Pilgrims may have been able to journey to The New World and set up a new community, but robots need access to the sort of power that only the sole remaining fusion power plant can provide.

The problem was in coming up with a way to incorporate the purpose of Power into the game in a way that didn’t bog down into tedious bean counting. To date…every version of R&R had involved tedious bean counting. Bean counting can be ok…but I already have several other sorts of beans to count that are far more fun and important (like Inspiration which tracks your progress towards true sentience, and Influence which you can spend to have macro level changes on society as a whole (build your own faction / clan / guild type stuff). Plus, tracking all of that power for NPCs would be a nightmare for the GM. In the game Power lets you take extra actions (so you can pull off the “swing on the chandalier” “boot to the head” move in a single turn…like a combo. And also to oppose other robots (like the bad guy ducking under your boot).

3) The third problem was also Power related and had to do with those NPCs. Power as implemented basically wound up with same feel as D&D hit points. You spend them freely not caring how many you have left until you get down to that danger level where the next blow might kill you. That meant that for players there was little tactical consideration to spending them as long as the robot had a lot of them, and then when the robot was running out players would go into hording mode. I had mitigated this a good bit with the idea of Power Drain (harkening back to various D&D house rules) where penalties would accrue as points go down (in this case reducing the Target Number for dice rolls). But ultimately this was death spiraly and I still had the problem of players needing to worry about making sure they didn’t run out of power while NPCs didn’t. I could blow through as much NPC power as I wanted because…hey…they’re NPCs they recharge off screen. That suddenly meant players would be cautious and timid about using Power for cool tricks while NPCs could go hog wild…not the desired game effect at all.


So that’s the situation that had me stymied for the last several months. Many different ideas and variations were conceived and discarded and I was starting to feel pretty discouraged. Surely there must be a way out.

That’s when the pieces started to fall into place.

I gave myself the proverbial hard look and said “if Power is such a pain in the ass…why even have it?” I’ve mentioned the overall game design goal behind featuring power above; but that (I now admitted) could be handled as simply as saying “robots need power…if they don’t get it they shut down…they can go X-days between recharging”. That accomplishes the goal of explaining why they can’t simply roam out into the wilderness…no need for actual mechanics.

“But” sez I “System Matters! Concepts that central to the game should be mechanically represented in the game…that’s a core component of my personal design philosophy!”

“Right, jerk off” sez the other me “System Matters. So what exactly are you trying to accomplish with these power rules?”

See, once I had decided that power was a central feature of the game and accepted that it should be mechanically represented, I had fallen back into bad old “design for cause” habits. I’d structured my Power rules to be a reasonable representation of how being run on batteries would actually work from the robot’s perspective. You had a certain capacity of power (points to track), as you did things you drained that power (spent points), and periodically you needed to recharge (get points back).

“Yeah…that’s how it would work…” sez the other me “you could track the power expenditure of your cell phone the same way…how much fun is that?”

So I sat down and enumerated what I wanted the Power system to do…my goals. My purpose was not to represent how a the robot’s batteries drain down in any kind of realistic simulative manner. The robots in the game are a metaphor (and a fun bit of color), but the game isn’t SimMachine or RobotTycoon. So what were my goals really?

I needed Power (if I was going to keep it as a mechanic and not just meta text it away) to be a limiting resource. Something that required some tactical management and could be “gamed” in the sense of understanding how it works. But its real purpose is to provide a mechanism to force the player to decide what’s important. “Is doing that REALLY worth spending power on right now?” If I have a limited supply to spend before I need to be recharged, and I don’t know when I’ll be able to recharge, then I have to choose between hording and cool tricks. Making that choice is one way in which I as a player make a statement about what’s important to my character as a free thinking Spark. See my character is programmed to save the princess. The GM puts in a big bad that will require me burning a lot of Power to defeat it…do I do it? Or do I simply say screw it and walk away and save the Power. What if the princess is someone my robot actually cares about for real and not just with programming? Do I do it then?

So basically the same sort of function as Humanity in Sorcerer…the opportunity to make a statement by deciding what I’m willing to risk losing Humanity on…only applied to fun swashbuckly stuff rather than deep soul searching stuff.

Did my power rules accomplish this? Nope. Players had so many points to spend that…like D&D hit points…there was no choice. If you have enough of them you don’t care if you spend them, and if you’re running low you can’t afford to spend them. No tactical joy or gamist resource management involved. And I couldn’t simply go with fewer points to eliminate the initial free spending because that would just put you into hording mode that much sooner. I needed a whole new system.

What are the key elements to a tactically enjoyable limited resource that doesn’t bog down into bean counting? Well, the number of points you have to spend must be small enough so that each one represents a tradeoff between current and future effectiveness. And they have to be replenished quickly enough so that “future” means short or intermediate future (because if it means distant future, hording sets in). We see the same effect in Hero Quest with Hero Points. Players have few enough of them so they matter, but a good GM refreshes them quickly enough to avoid hording.

“Ok” sez I “My current power rules are for shit. Got any suggestions Mr. Smarty-Know-it-all?”

“As a matter of fact” sez the other me “I do…” (in case you’re wondering…yes, I do have these conversations with myself…they just take a long time because I argue as much with Me as I do with anyone. During the development of Universalis, Mike Holmes got to play the role of “Other Me”, but for R&R I’m largely going solo.)


And here at last is where the solution starts to take shape.

**I’ll give every robot a small number of points that automatically refresh each round.**

Since it’s a small number, each one is a trade off: Spend it for an extra action now…or save it to defend myself later (as an example). Since it refreshes each round the decision is only a temporary one and there is no hording effect. In fact, there’s a Spend it or Lose it effect which could bog the game down in pointless Power expenses unless I keep the number of points small enough so you never have enough to do everything you’d like to (a good lesson taken from Euro board game design). If there aren’t enough meaningful things to spend the power one you probably shouldn’t be playing that scene out anyway. Say “yes” and move on.

I kept the idea of Power Drain, but rather than being a death spiral effect as you spend down Power I made it into a voluntary intermediate term tactical choice. By taking a level of Power Drain you can refresh your pool of Power in the middle of a round…for those occasions when you REALLY want to succeed. In that way I preserve the “what am I willing to spend power on” effect. If I have 3 points each round, but need to spend 5…do I accept the Power Drain, or just say screw it.

“Hey, that’s pretty exciting” sez I “It eliminates the bean counting, and it eliminate the NPC problem because now there isn’t a difference in the question of whether to horde or not to horde between PCs and NPCs, and it eliminates the D&D hit points problem. You’re pretty smart.”

“Of course I am” sez the other me “you need to shut up and listen to me more and quit pissing around for months at a time with your head up your ass.” (I won’t bore you with the details of the ensuing cat fight THAT remark caused).


And here’s where the convergence starts coming in.

I needed to figure out a way to determine the size of the pool of Power that would get refreshed every round knowing that I would need it to be in the 2-4 points range. I could just set it as a fixed number for everyone…but where’s the fun in that.

So I started to try to back into an explanation that would make sense as part of the game world. I like to have in-game justifications for things. Its kind of like the designer version of Author Stance. I’m doing it for meta reasons, but I’d like to have something that sounds good too.

It occurred to me that perhaps it had to do with the robot’s relative importance in the game world. Low level shop keepers and soldier types would be programmed to simply execute their basic programming without a lot of flash. High flying heroic swashbuckler types would be expected to do a lot of feats of derring do and would need to have the power available to do that. Ergo…low importance robots have smaller Power Pools, higher importance robots have larger.

Importance in R&R is also tied into how human the robot looks for background reasons I won’t get into here. Less important robots look more roboty, more important robots look more human. This is summed up in a stat called Anthropoid Class which directly represents how human looking the robot is, and indirectly serves as a proxy for the relative importance / status of the character.

Ok, so I can tie the Power Limit into Anthropoid Class…that works because AC is rated on 0-5 with standard robots being 2-3. Perfectly in scale. I also got to thinking, however, that maybe this limit was a programming issue and not a hardware limit. Lower robots didn’t have smaller batteries (that would be a capacity issue not a points per round issue) they were just programmed to consume less power (a cost savings plan for the park…like a lap top’s power-save mode). Given that, Spark robots should be allowed to break that programming. Which works great because that way low importance Sparks can overcome that disadvantage. The degree of a robot’s Sparkiness is measured by the Self Awareness Stat. SA is on a scale from 0-10. Low and behold…half of this is 0-5 so I can simply set a robot’s Power Limit to be the greater of AC or ½ SA. Simple, effective and uses game stats I already have.

And you know…I already been thinking about something similar for combat. If you remember from above there was a threshold for when a robot would pretend to be put down by an attack. I had been thinking that should be based on importance also. Low importance nobodies go down easy, high importance types you have to work for (R&R's organic version of a Mook rule). I could use the exact same rule as for Power Limits…set the threshold to the greater of AC or ½ SA. Cool.

And as I thought about it, I realized that the Power Limit solves (I think) the threshold problem I was having in #1. See I couldn’t generate enough Net Successes to hit a reasonably high threshold…namely because opposed rolls would cancel out most of the successes a successful attack would generate. Opposed Rolls had always cost Power but when Power was free and easy (especially for NPCs) robots could Oppose all day. Now with a low power limit not every attack will be able to be Opposed. Now when your heroic or Sparked character with Power Limit of 3-4 faces off against the nameless guard with Power Limit 1 or 2, the hero is able to run the opponent out of power and then launch an unopposed attack against a lowish Threshold (which, by the way, is how I originally envisioned combat working way way back in Alpha…so hurray for that). Against the Power Limit 3-4 (or even 5) major villain, however, whose threshold is also higher, the hero will have to work a lot harder and get creative in order to persevere. Testing will be needed to ensure “work a lot harder” doesn’t equal “impossibly tedious”.


So what a neat convergence. That simple “Greater of AC or ½ SA Power Limit” rule helped solve (I think) all three of these problems. Since I’ll be referring to that number in a couple of different places I’ll need to come up with a name for it for ease of reference on the character sheet. Maybe I’ll just call it “Importance” like in Uni.

“But wait,” sez I “that rule sounds awfully darn familiar…didn’t I write a rule like that once before…?”

“Hmmmm….could beeee….” Sez the other me snidely.


Sure enough, there it is in my rules for Faction Building. You know, up above where I said one of the resources that the game had was Influence Points which could be spent on cool change-the-world type stuff. You collect Influence Points from your Contacts…robots who owe you favors…by calling in those favors and having them do stuff for you.

How much influence each Contact can provide depends on how important they are. They might be programmed to be an important character in Auvernais and thus have access to lots of resources and people who do what they tell them…or they might be important because they’re a Spark and have manipulated things to their own benefit or have some pull in the

Spark community. Anthropoid Class is a proxy for programmed importance, and Self Awareness is a proxy for Sparky ability.

Care to guess how I defined Influence nearly a year and a half ago now? That’s right.
Influence: the greater of Anthropoid Class or ½ Self Awareness.


“Geeze…you mean it was sitting right there all this time!?” sez I
“Ayup…” sez the other me
“Why didn’t you tell me! Why did you let me waste all this time reinventing the wheel?” sez I.
“To teach you a lesson in humility Fuck Face.” Other Me sez laughing.
“Asshole” sez I

Beautiful :)

as the sort-of person here (that i imagine is quite common) that has only ever dreamed of putting my game ideas (seriously) on paper, this is the kind of experience i like to read about-

1) The problem and its origins
2) The solutions and their origins
1a and 2a) all the reasoning that lead from 1 to 2 without leaving out the boring parts like failed attempts

Good luck on the rest of your game :)

Cheers

That was a fun read. It's a lonely, schizophrenic affair, being a designer.

I kept the idea of Power Drain, but rather than being a death spiral effect as you spend down Power I made it into a voluntary intermediate term tactical choice.

Each level of Power Drain reduces the Target Number on your test rolls by 1. So an NPC robot could spend power out the yin yang, but they'd pay the price for it immediately in the form of the lower TN rather than at some undisclosed time in the future where they run out of power...which would never get played out and so isn't a real hindrance.

The first time you're called upon to make a Test Roll and the TN is modified to 0...your robot shuts down due to lack of power.

Since power refreshes each round you more or less have as much of it as you want...as long as you spread your actions over several rounds. Its when you absolutely want to do a bunch of stuff right now that you can accept a level of Power Drain (or more) to get the power you need.

That's the current theory anyway.


Tieing Power to a die roll was one of the options considered. Currently you make a Test Roll and any die that comes up as a 1 (which would always be a Success since its a low roll system) can be taken instead as a point of Inspiration. So you're choosing between acquiring that important resource at the expense of current effectiveness.

I had toyed with the idea of making dice that rolled a 10 cost power, but while I liked the symmetry of that, it had a number of hazards that caused me to discard it. I'd tried a number of different ratios...every two 10s on a single roll...every 10 but the first...those options became just as fiddly as points.

Each level of Power Drain reduces the Target Number on your test rolls by 1. So an NPC robot could spend power out the yin yang, but they'd pay the price for it immediately in the form of the lower TN rather than at some undisclosed time in the future where they run out of power...which would never get played out and so isn't a real hindrance.

There's two possibilities. The easiest to implement is to have the TN drop occur immediately upon refreshing the extra points. This would still tend to be a net positive to effectiveness.

Concider a 4 die roll with a target number of 5 (50%). The expected value would be 2 successes.

If you pay a power to take an extra action (to keep it simple lets also assume 4 vs. TN5), the 2 successes from the first roll can roll into the second making the second 6 dice at TN5 with an expected value of 3.

If that power resulted in a drop of TN down to 4, then you'd have 6 dice at TN 4 which would have an expected value of 2.4 Still better than 2 but not as good as three.

If the TN is lower or the number of dice lower then the effect is less tempting. A 2 die roll vs TN 4 would have an expected value of .8. Accepting the drain to take an extra roll would only make that a .84...so almost a wash.

The second possibility is to not have the TN drop take effect on the roll that caused the Drain but only the next. That makes the choice more tempting which I like but is a little more complicated to explain in the rules.


All robots will have essentially the same range of TN. Its based on attributes (which represent the robot's hardware) so since all robots started out as non Sparks they all have the same original hardware (depending on what they were designed to do). I am planning rules where Sparks with access to the right parts and a competent mechanic can upgrade themself, but that's mostly chrome.

The frequency of Stunts has more to do with what the robot's programmed role was. If they were programmed to be a high flying swashbuckler, then that's what they'll do. If they were programmed to be shop keepers than they won't. In game terms, high flying swashbucklers were the "stars of the show" while the shopkeepers (by and large) were the "extras".

As such the stars were made to look more human...which mechanically translates to a higher Anthropoid Class. which results in a higher Influence, which results in them having a higher Power Limit with which to pull off the stunts they're programmed to do. For shopkeepers, low level soldiers and guards and the like the opposite (less human, lower AC, lower Influence, lower Power Limit).

A player character based on shop keeper who Sparked will thus actually have a lower power limit and not be as effective at stunts as a swashbuckling NPC.

BUT, the way I have Self Awareness set up...its easier to increase (for various background reasons) for lower AC robots than higher. So a low AC Spark player character could increase their Self Awareness.

Since Power Limit is based on Influence, and Influence is based on the greater of AC or 1/2 Self Awareness, by increasing their Self Awareness (becomeing more Sparked) they can overcome that initial programmed limitation and catch up in that respect to player characters who started out as Swashbucklers.


Did I make that clearer or more confusing?

Cool, and clear (mind you I have the short version you were selling last year).

One note. Naming that quantity Influence may have worked before... but I suspect that this function will rapidly swamp the previous one. So I would personally suggest that you work your nomenclature a little. Frankly calling this just plain Power - connotations of electrical and social power - suggests itself right off. The increments you spend, up to (power) limit each round etc., would then need a subsidiary name. The physicist in me winces, but I'd suggest Watts here nonetheless. Spend a Watt to get an extra action, etc etc, up to Power per round; spending more Watts than Power can only be done by a Drain on your power core. Cool.

Another note, just on scaling... I don't remember where you use it other than in deriving stuff, but a 10-point Self-Awareness scale could be converted down to a 5-point scale either by losing some information, or by preserving the "odd point" in a binary toggle. Maybe Self-Awareness 3 and 3+ give the same dice (Power = higher of Anth.Class or SA), the only difference is that a 3 increases to a 3+, a 3+ to a 4. Fractions bad; consistent currency scale good. Not pushing it; just thinking you might want to consider this route.

Does this mean we get a full version for GenCon?

- Eric

Right now I'll have 3 uses for the expression =>AC or 1/2 SA
The power limit,
Influence to generate points with
Damage threshold.

I might wind up calling the meta-stat "Importance" after all and have the three subuses be Power Limit, Influence, and Mortal Wound.


As to scale, the 1-10 scale is important for the d10 checks which are the primary function of Self Awareness. I think trying to match scale by making it 1-5 would be more trouble than its worth.


As to GenCon...I'm hoping...but not expecting at this point. We'll see how quickly I can get this latest version ready to be playtested.

“As a matter of fact” sez the other me “I do…” (in case you’re wondering…yes, I do have these conversations with myself…they just take a long time because I argue as much with Me as I do with anyone. During the development of Universalis, Mike Holmes got to play the role of “Other Me”, but for R&R I’m largely going solo.)