Law & Economics, or Players Respond to Incentives

I once had the pleasure of lunch with John Zuur Platten, the business partner of Flint Dille, Gary Gygax's old friend and collaborator. Through Flint, John had had the chance to learn much about Gary Gygax and the origins of D&D. John explained to me that "to understand D&D, you have to understand that Gary thought like an insurance actuary. D&D is fantasy fiction through actuarial science."

Re-reading the 1e Dungeon Master's Guide, the truth of this claim becomes obvious. Previously inexplicable rules - like the disease tables, or potion mixing - clicked. Gary wrote those rules because he wanted to account for the actuarial risk of living in a fantasy world. Gygaxian Naturalism exists in a like manner.

Unlike Gary, we've had the benefit of enormously powerful spreadsheets, computer modeling, and random generators at our desktop. Seeing how useful those tools have been has left me rather in awe of the men who could develop the engine for D&D without them. Reflecting on the actuarial science behind early D&D lead me to think about  what fields lie behind ACKS.

The field of Law & Economics is a chief inspiration. As taught in law schools, Law & Economics is the application of economic principles to the analysis of legal rules. Whereas legal formalism will analyze law based on moral and cultural premises that serve as its axiomatic groundwork; and critical legal studies will analyze law based on its relationship to postmodern power structures; Law & Economics analyzes law to look at what it does to human economic behavior. L&E assumes that people are utility maximizing and respond rationally to incentives.

Law & Economics as applied to ACKS has been inordinately helpful in evaluating rules for their effect over the course of a presumed ACKS campaign. If players are treated as utility-maximing, rational, and incentive-reactive, many game rules which seem like good ideas are revealed as bad. A great example is 'Shields Must be Splintered', a current darling of the OSR.

Shields Must be Splintered allows a character to avoid a killing blow by sacrificing a shield. The idea is to add increased utility to shields, and to give low level characters an opportunity to survive despite their low hit points. It feels historically plausible, as shields did break in battle, and it has a "story game" like feel that gives the player some dramatic choice in play. Yet this rule has deliberately been excluded from ACKS.

Why? Consider a 7th level fighter with a bag of holding. A bag of holding can carry approximately 1,000 shields. Such a fighter becomes invincible if Shields Must be Splintered. He need only reach into his bag, and voila-he produces a one-charge item that prevents death. If a bag of holding isn't available, some 0th-level shieldbearers can carry quite enough shields.  Shields Must be Splintered only works so long as the players don't rationally react to the fact that a cheap, widely available, 10-gp item is essentially a one-charge magic item that prevents death. As soon as players realize this, and have the resources that shields and shieldbearers are negligible expenses, it breaks the game.

Law & Economics also explains why so many post-D&D RPGs defaulted to "GM-is-storyteller" railroads, regardless of any other mechanics in play. Such RPGs eschewed the "silliness" of D&D's xp-for-gold model, and instead relied on subjective XP awards from the GM. Everything from the d6 Star Wars RPG to Shadowrun to 2nd Edition D&D to Vampire: The Masquerade went down this path. But if the XP award is a subjective award from the GM, that means that the rational utility-maximing player will seek to maximize the GM's subjective impression on his play. What does that lead to? It leads to the soft fascism of "What does the GM want us to do?" It leads to players who compete for the GM's attention and favor.

Ironically the generation of games after the post-D&D 90s kept the subjective XP, but then bent itself backwards to limit the GM's power in other ways - not understanding that the power to reward is the power to control. This is why ACKS has carefully chosen and strictly objective XP awards. ACKS is a game about rising from a nobody to a somebody, and the XP awards align with those goals - no GM intervention required.

A third example of the application of L&E thinking is in our domain rules. In ACKS, adventurers who establish a domain can make profits from the domain, and in doing so, they can earn XP to advance their character. Contrast this to the Mentzer Companion rules, which sharply limited the XP gain from domains, and, worse, made domains inherently unprofitable - this was purposeful, in order to force the characters to keep adventuring to support their domain. But players don't need incentives to keep adventuring - adventuring yields gold, magic items, and XP, which are the three main incentives of the game.  What players need (if domains are desired by the game designer) are incentives to build and run domains. The lack of reward structure for domain management has been one of the main reasons that domains have been the province of retirement, not action, in past fantasy RPGs.

ACKs may not be to everyone's taste, but I am confident that ACKS is at least among the best-designed fantasy RPGs from the point of view of connecting the rules of the game (its laws) with the incentives of its players (its economics).

Comments

Damn that's an interesting post! I've long commented the AD&D books show signs of heavy influence from Gary's training as an insurance salesman, the strictive details of race and class in particular. It's quite something to see someone in Gary's orbit say essentially the same.

In the social sciences, we make strict distinctions between humans as rational actors and humans as maximizers. The latter position was heavily assualted during the post-processualist, anti-positivist movements of the '90's and is considered generally untenaable today, although a few Human Behavioral Ecologists remain bound to optimization models. Far to big a subject for a brief post, but a simple characterization would be to say that while humans will seek desired outcomes, these may be far from the "best" outcomes in a maximizing sense.

However I don't think that issues with optimization theory impacts the ACKS rules much, so I'm just sort of musing here.

I'm right with you on the objectivity or fixed XP awards. I have to say though that I have reservations about tying XP to domain or institutional generated wealth. I don't see a reason why the king becomes a 14th level fighter, for example, because he collects taxes, or the mafioso boss becomes the grand master thief because his underlings are successfull. I'm exagerating the mechanic here (I think), because it's not clear to me why level need be tied to social rank this way. I also think, players will find that there are many rewards to runing domains and that XP can be gained through plunder and war too. I'm just suggesting that perhaps there are alternatives that achieve the same goals.

In any case, ACKS certainly does bind all these aspects of the game together to a degree never achieved before and provides a great platform for customization of a comprehensive campaign.

Thanks for the very kind words on the blog post, and on ACKS!

Regarding rational action v. maximization, I think you're right that it's probably too much to even attempt to address here. Nobel Prizes are gained and lost on such debates.

I think awarding XP for domain management makes sense in the context of our assumed fall-of-Rome-ish (Auran Empire) setting, Warring States Japan, Dark Ages, or Bronze Age clash-of-cities setting, or similar environments where rulership was a highly dangerous activity. It makes a lot less sense in campaigns where you've got the rule of law, fixed heredity, stable societies, and so on. In these cases, adventurers are truly outliers and most rulers would just be normal men or low level characters...

Cheers!

Very cogent post. The part about "GM soft fascism" is probably the very core of why I play the way I play.

AD&D as the D&D Actuary's Edition makes a lot of sense. The idea of the DMG as the (uncompiled) source code for a Gygaxian Naturalistic world once again illustrates the truth that "It's not stupid; it's Advanced".

If players are treated as utility-maximing, rational, and incentive-reactive, many game rules which seem like good ideas are revealed as bad.

Love this as a basis for rule analysis. So true to lived/played experience.

One niggle though:

Such a fighter becomes invincible if Shields Must be Splintered. He need only reach into his bag, and voila-he produces a one-charge item that prevents death.

Non-problem IMO, and definitely not a gamebreaker.

SSBS allows a character to ignore damage from one attack per round, not to ignore death from damage.

* How long does it take to fish another shield out of a Bag of Holding? One round? Mr Bag o'Shields has just been shut down as anything other than a speedbump for so long as he continues to dive into his bag.
* Who says the opposition are attacking one at a time?
* What's the ROF for a bow in ACKS at the moment? 2/round?
* Fat lot of good a shield does against falling damage, hot oil, etc...

(this last section cross-posted from Sean-n-Sorcery blog)

Chris, regarding SSBS, I had a very different experience. We tested it with mid-level characters and found it quite game breaking. Perhaps my group is more, uh, "utility maximizing" than yours? An encounter with a cyclops or a giant becomes much, much easier with SSBS, for instance.

I found it useful to think about it this way: How much would I expect to see a one-use magic item that enabled a character to ignore damage from one attack per round cost? Would it be "more than 10gp"? The answer is yes!

Or consider that a potion of healing enables a character to spend one action to regain 1d8hp. This is mathematically the same as avoiding an upcoming attack worth 1d8 hp. A splinterable-shield enables a character to spend one action to avoid [any one attack] worth of hp. A fighter is therefore considerably better off with an extra shield (10gp) than with an extra potion of healing.

I agree that the Shields Must be Splintered rule makes shields too good. This is certainly true for a game that includes armor as heavy as plate. Generally, shields fell out of favor when plate was in common usage as you needed a two handed weapon to cut thru it. Also, it fails to account for the fact that very few characters in the stories that D&D is based on actually make use of shields on a regular basis.

When is this game coming out already? I need it right now! :)

This is a fascinating critique of the 1E DMG, and it intrigues me now to reread sections of ACKS through the lense of maximizing utility, or whether the right behavior is being incentivized. I've spent the most time with the campaign sections, so I can see how this philosophy comes through pretty clear in campaign choices the players might make.

One wrinkle to SSBS you might have missed; the original version of the rule I saw Michael Curtis using involved having to make a Death Saving throw in order to sacrifice the shield; it's not an automatic escape-death item.

SSBS - the original rule comes from Dragonquest.

I've never had the hunkering to go out and buy a retro-clone...but you guys seem to be doing it right!

What's the ETA on release?

We are discussing that by email right now, but the answer is some variation on not much longer!

I had thought this for the longest time about EGG. Sure when he took the necessary job of an insurance guy for a while to put beans on the table, it may have appeared off course, but it very much honed his skills at rul-ifying life, and by extension "fantasy life". I also worked for a few insurance companies, and through the insurance comes a tradition that's attempting to capture the framework of life, random crap that happens, random event (why are there so many actuarial tables about "random encounters"? Because that's his background, and putting tables around fantasy life helps organize it much better.) . And just because he was a salesman, and not a product manager, the training that they give you and some thinking about the matter would've very easily led him to the ideas on how to construct such a game.

So he applied his science fiction and fantasy reading into a well-designed, even if rules-heavy, construct through which everything can be accounted for. For example, how to put a framework around the risk of a loss occurring on a property? Well, where is it? Is it a crime-ridden neighborhood? Could it be overtaken by a wildfire? The well-worn tradition of insurance very much aids in getting your head around these issues, over many decades of trial and error.

So not only did EGG inherit a good deal of fantasy and SF literature and cultural tropes from which to build the game, he also inherited the skills and traditions of the insurance field.

Isn't this the way it is often is, take a few seemingly unrelated disciplines, and combine these, and all of the sudden you have something surprisingly new and worthwhile.

OTOH, a "Splintering Shield" +2 that you could expend to save your character's life one time would be a pretty cool magic item. :)

casque monster beats pro et aussi Inters experience Geffen community relationsPhysician
reteqrewqr
La conceiving d'insertion verticale,

Add new comment