Cropped & edited extract from The Triumph Of Justice by Gabriel Metsu. Refer to the link for terms of reuse.

This article is the result of a confluence of many different vectors, from reading a review in the current issue of KODT of the original “Paranoia” RPG to reading an article at E Pluribus Unum about Twitter mistakenly suspending an account as a purveyor of spam. Setting aside the questions of anti-spam techniques and technologies and their application in this specific case, it raised an interesting thought in general.

Twitter, and the other social networks, are not public services; they are entitled to define whatever rules for usage that they like, and you can either follow those rules or stop using the service.

But these incidents raise a larger question. People write rules all the time, but when it comes to technology, they then encode these rules into Rules and make the detection of infringements the province of automated systems.

That’s fine, too – there’s far too much activity for human agencies to sift through it all, looking for red flags. But no algorithm is perfect, when it’s monitoring human activity. It can’t be; we don’t normally reduce to neat lines; there is always some fuzziness. And that becomes a problem when the same automated systems also initiate action.

The Need For Speed

And yet, there is a need for a speedy response to allegations of spam or other misuse of technology. Even an additional ten minutes can permit hundreds, thousands, or tens of thousands more to be victimized by undesirable behavior. The obvious solution is for enforcement actions to be subject to human review – but not only would this be mind-numbingly dull, since most of the decisions would be correct, but it would be expensive. So the system in place relies on the occasional incorrectly-targeted individual requesting a review of the decision. Those who knowingly broke the rules won’t bother appealing because they know they will lose. You could even make the argument that appealing such an automated decision is prima facae evidence of innocence.

The Dimension-Regency Extrapolation

When I was creating the background for the setting of the current Zenith-3 campaign, Dimension-Regency, I extrapolated from systems that were being put in place in Sydney that not only automatically detected traffic infringements – speeding and running red lights – and automatically issued infringement notices to a future in which judicial decisions were made by Expert Systems in 2050. The discovery process produced an agreed-upon submission of facts, and these facts – and relevant arguments – were input into the system which produced a verdict. Appealing such a decision also needed to pass through this software – if the machine determined that there were valid grounds for appeal, or an unresolved point of law was identified, permission was granted to appeal to a human judge and jury. If no such reason was identified, the software refused leave to appeal and proceeded to a sentencing phase, which assessed the various relevant factors and produced a sentence that was informed by established social mores and lay somewhere between the minimum and maximum mandated sentences. Once again, permission to appeal the sentence might be appealed, using the same process already outlined.

The point was that 98% of lawsuits and criminal cases were resolved without requiring lengthy court procedures, cutting the court backlog massively.

What I didn’t realize until I read the article referred to in my opening paragraph is that this perfectly and precisely forecasts the rule enforcement procedures and policies employed by Twitter, and no doubt by other social networks, today. (At the time, there was no such thing as social media; I based a lot of my material on IRC, which was the social connectivity of its day.

Rules Without Oversight

Computers without adequate human oversight are a feature of movies such as Dr Strangelove and the Terminator franchise, to name just a few. It’s unnecessarily alarmist to view the shortcomings of the Twitter implementation of their rules and policies as the thin edge of the wedge, but nevertheless, the eerie similarities to those dystopian views remains.

So how does all this matter to RPGs and to GMs?

The RPG Equation

Aside from the obvious utility to sci-fi RPGs, there is, once again, the broader issue. One of the most significant changes in the last decade or two has been a move away from the philosophy that “the GM is always right”. This was stated explicitly in AD&D, was watered down slightly in 2nd Edition, and watered down still further in third edition to only being able to arbitrate when there was a conflict within the rules.

There is a very real debate that has raged throughout gaming (even on these pages) between the old-school philosophy and the new-school. I’ve made my position in this debate very clear on a number of occasions, and have also held extended discussions with my players on the subject.

Certainty

Players like certainty. They like especially like certainty about the rules, the knowledge that they can have their characters attempt to do something and that it will be arbitrated according to standards that they both know and understand. The old school demands absolute trust in the GM by the players, and that is not always forthcoming, with good reason; every GM and player has suffered at the hands of apparent abuses of this authority. This produces a trend toward demanding that the rules, as written, must be enforced when clear, and let the chips fall where they may.

… vs. Control

GMs like certainty, too, but they also demand the authority to decide – based on the circumstances and campaign – exactly what that certainty contains. Nevertheless, the truth is that a GM who tries to force the players to his will risks having no players, just as a player who refuses to obey the GM risks being asked not to play again. Everyone is giving up a portion of their free time in the expectation that they will have fun, or at least be entertained. If that doesn’t happen, the game is in trouble.

And uncertainty makes for unhappy players; which means that every diversion from the written rules must be justified, and (even then) is on perpetually-shaky ground. So there is constant pressure on the GM to follow the rules as written, like some mindless automaton, and an increasing expectation on the part of players that the rules will be as written in the sourcebooks.

I don’t advocate players being at the mercy of GM whim, but I am convinced that the current trend toward the role of the GM-as-an-umpire has gone too far. When players can state, with peer approval, that they would never play in a game that has house rules, or never cast a spell that has been modified by the GM – both of which have happened to me in the past – there is a problem. Eventually, inevitably, like the situation reportedly facing Twitter, removing human oversight over the rules and their interpretation leads to trouble.

The Ironic Uncertainty

The irony of the whole situation is that a mindless and dogmatic application of rules – the only thing that a machine understands unless it is programmed very cleverly – has produced a situation in which uncertainty is dominant. When using a service in exactly the manner advocated produces an outcome deemed contrary to the rules, and hence a penalty, no-one is certain what they can or can’t do.

There are a lot of deep issues involved in this situation, Oversimplification of a situation to the point where black and white rules can be followed dogmatically produces inflexibility and cases “falling through the cracks”.

The Hierarchy Of Rules

I believe that there is, or should be, a hierarchy of GM rule authority. I’ve discussed this before – notably in Blat! Zot! Pow! The Rules Of Genre In RPGs. In it, I advocate:

• House Rules Trump Official Rules
• Simulation (i.e Realism) Trumps Rules
• Genre Trumps Simulation
• Plot Trumps Genre
• Campaign Trumps Plot
• Gameplay (i.e. Practicality of Play) Trumps All

It could be argued that Fun should ride atop this hierarchy as the ultimate trump card, but I reject that; these guidelines (and a lot of other GM advice) is all about making the game fun, so I don’t consider this necessary. It might also be argued that Justice or Fairness should be explicitly stated at some level, probably between Campaign and Gameplay, but – once again – I reject this; the assumption must be made that at each of these levels, the GMs rulings are fair, and indeed the whole purpose of this hierarchy is to be certain that fairness and justice has an opportunity to play a role, by explicitly considering situations in which a lower order of rulings might produce an unfair result.

Rejecting The Rules

Examining this hierarchy in terms of enforcement, it should be clear that there are three areas where GMs can and should override the existing and agreed-upon rules.

“Realism”
The first is when the rules dictate an unrealistic outcome (in terms of what is ‘realistic’ for the GAME world, which is not necessarily the same as what is realistic in the ‘real’ world); when this happens, GMs should forget the rules and dictate whatever outcome is ‘realistic’.

Genre
The second is when Genre trumps realism, even within the game world. This safeguards against flaws in the game world’s conceptual mechanics, and explicitly mandates occasions when the GM should override not only what is reasonable and realistic within the game world – and never mind what the Rules say. This is also an area where the players can have their say; while they aren’t always in possession of all the facts, and so cannot state unequivocally that a ruling is contrary to Simulation, the definition of what is and what is not canonical within the genre and sub-genre are far more open.

Practical Gameplay
The final point of appeal is to the question of gameplay. It doesn’t matter what the rules, or earlier rulings say – if its not a practical approach, the GM needs to short-circuit the process and get on with the game. It should take less than 2 seconds to identify whether or not there is an issue to consider at each of these levels, and no more than 30 seconds to make a decision – I would prefer to say less than 10, but sometimes it takes longer than that to find, never mind read, read the relevant sections of rules.

If it takes longer than 60 seconds, my rule-of-thumb is to decide what happens without recourse to the rules, running that decision through the above gamut (assuming that my ruling is a house rule). If you can manage to make a decision in ten seconds or less, then the entire process of arbitrating even the complicated situation should take a minute or so – not an unreasonable time frame.

An alternative hierarchy

Some people have suggested to me that the hierarchy itself is incorrect, and should read:

• House Rules Trump Official Rules
• Plot Trumps Rules
• Campaign Trumps Plot
• Simulation (i.e Realism) Trumps Campaign
• Genre Trumps Simulation
• Gameplay (i.e. Practicality of Play) Trumps All

This is a very interesting point, and one that merits further discussion. In a nutshell, should subjective in-game reality overrule plot/campaign, or can plot/campaign overrule subjective in-game reality?

I chose the first hierarchy over the second because a plotline can focus on the (temporary) overriding of established in-game reality, but it has been since pointed out to me that this simply enlarges the scope of what is permitted within the in-game reality by displaying a new phenomenon. It’s the difference between ‘subjective in-game reality’ and ‘established in-game reality’.

A Decision

After much thought, I have decided that the original hierarchy should stand, because it places the welfare of the campaign above that of the ‘realism’ of the campaign. Objective reality might say that a properly-maintained and prepared weapon will fire reliably – but if it is better (read “more fun”) for adventure or the campaign for the gun to misfire, then hang it all, it should misfire!

Liberation!

And that brings us back to the point of this article. A slavish adherence – a ‘foolish consistency’ as Emerson described it – to the rules of anything does more harm than good. Certainly, consistency is important, and should not be ignored by whim or caprice – but flexibility and an admission that circumstances alter cases, rendering some of them fuzzy, must be built into every decision-making system – whether we’re talking about Twitter rules, RPG rules, traffic tickets, mandatory minimums, or Skynet.

Sidebar: Recommended Reading
I couldn’t let this article pass by without plugging one of the best books I’ve ever read on the rise of machine intelligence and the inflexible application of rules: James P. Hogan’s The Two Faces Of Tomorrow (3.82 rating at Goodreads) – review. amazon page. And, most astonishingly, it was written in 1979, when computers were in their infancy. The science-fiction – and the science in back of it – still rings true today, 35 years later. And that’s very, very, rare. I give it a personal 4.95 out of 5.