Let's correct a certain misconception. Free Kriegspiels never really went away, nor did their evolution through Matrix games and the like ever stop. In fact, we have a relatively new expression of that, framed and couched within its own set of rules and directions, called Open Strategy Games.
openstrategygames.github.io/…
It's particularly interesting because it is very clear that the referee is specifically that, a referee, an intended adjutant of the actions taken by the players. The referee does not bring the fiction during play. That's not his job. His job is to simply judge what the outcome of the actions chosen by the players are.
While there are mechanics, none of those mechanics apply to judging the outcome of those actions. The mechanics are about turn sequencing and structuring what you can do/accomplish in a given turn, limiting the number of turns, etc. The mechanics are entirely outside the space normally considered for being "the important part" of having mechanics.
There's an interesting bit early on in the text that I think is totally worth putting in this discussion:
The goal of the game is to achieve your objectives.
The point of the game is to create a credible narrative.These two sentences encode the complete design philosophy. You should be trying to win: pursuing objectives, building leverage, outmaneuvering rivals. But you are simultaneously playing to find out what happens. A player who sacrifices their own objectives to make the narrative more interesting is playing correctly. A player who exploits a rules gap to achieve their objectives through means that collapse the fiction is playing
incorrectly.Play to win. Play to find out. Both at once, all game.
People with a background in modern story games and narrative first games will recognize a lot of this directive. In fact, you could have pulled it right out of Apocalypse World nearly 20 years ago, and definitely right out of Blades in the Dark a decade ago. You play the game to create a credible narrative while trying to achieve your objectives. A character/faction without objectives has no play that they can make and is unsuitable for play. A referee or player that comes to the table to win at all costs without being interested in a credible narrative is also unsuitable for play. You play to win and play to find out.
Frankly, I would love to get into an Open Strategy Game at some point, but I have absolutely no idea who's playing this, when and where and how. I'm certain there's an entire community devoted to this kind of gameplay, but they are not contiguous with either my experiences in RPGs or my experiences in tabletop war games, even though this is firmly and explicitly in the overlap.
OSG is all the best things that Braunstein wishes that they were without all the crap things that they actually are, which is fascinating.
But let's take this back to free kriegsspiel with intent, that is, with structures that the referee is bringing to the table intentionally in terms of expected outcomes. Here, in an OSG, that would be part of the brief, which everyone receives, if not in full. A brief can certainly direct you to lean toward things going wrong, though I would argue that this is not a reversed gamist wargame where unfun is the goal because failure is not inherently unfun. (The obvious, though painful, counter to the suggestion is that there is a large fan base who love TPK dungeons. You can't tell me that failure is unfun when those people exist.)
Part of the problem I have with your interpretation here is that free kriegsspiel, used outside of simulation, does not undermine its own basis. It's simultaneously storytelling and trying to win. The Prussian prince may very well enjoy a free kriegsspiel experience run by his nanny, not because he trusts his nanny to be providing a perfectly accurate simulation of battlefield reality, but because their shared experience of telling a story is gratifying.
You've sort of cheated here by setting up a situation where deceit is part of the process. That's really kind of unfair. Most people don't enter into game playing with the expectation or experience of being deceived about what is about to happen. They go into play with the expectation that it is some level of abstraction with which one will grapple.
Likewise, you've inserted the element of deceit in your second example, where the test of adversity requires the players not to know that things are tilted toward degenerate situations. But does it actually change things for those players if they were told up front that this is going to be a test of your ability to deal with things that go wrong, and you should expect things to go wrong?
Is this not actually the right attitude to carry to war if you were wanting to teach them something as an effect? Is the deceit actually an important component of the desired result? I don't think so.
We have military training now where we instruct that those being put through the ringer are told that things are about to get really terrible to a level that they should pray they never experience on the real battlefield. Yet they learn from it.
So with these two pillars of your argument addressed, I don't think your third pillar holds up either. For free Kriegspiel to work long term, it does not have to be grounded on judges who are actually interested in simulating what would really happen, at least not in the sense of having a deep amount of subject matter expertise. They have to be devoted to the idea of "play to find out." That part's true. They have to be willing to let the players fail, but that's not inherently unfun. That's just a side effect of playing hard and not playing soft.
I think it's only a very narrow slice of the tabletop hobby which are being instructed to fudge when things are "not fun," rather than when "everyone at the table agrees the result is unreasonable." It's kind of funny that I don't see these debates over in the story gaming side of the world. Not because it's assumed that you're going to fudge the dice when things are unfun, but quite the opposite, because the players are empowered at almost every turn to make good decisions under bad circumstances. Thus, if the circumstances break against their interests, that's part of the game. Failure is not unfun. Overcoming failure, overcoming things being against you through no fault of your own, overcoming someone else's failure being dumped in your lap—that is the fun. Play to win, but enjoy losing from time to time.
Large language models have nothing to do with this here. Even if you trained one specifically on all of the military doctrine ever written, that would just make it a set of rules. It's just a set of matrix math, figuring out what token comes next. That figuring out may be quite good, but it's still an abstraction, just like any other abstraction. A representation, just like any other representation. It's rules, just more of them in a language you can't read.
At that point, it's not about whether or not the rules are accurate. It's about how much abstraction you have to deal with to communicate both intent and outcome. Rolling a single D6 with modifiers as your core resolution mechanic is just as valid as a 300 gig large language model trained on all of the doctrine of the world, as long as your expectation is set properly. It's just as good as looking at the outcome and the influences, the leverages, and the obstacles, and having someone who isn't a subject matter expert consider, make a judgment, and render it to people.
We have a problem sometimes in that the simulationists want to claim that simulation is better than narrative, while ignoring that simulations generate narratives. The only reason they think the simulation is good is because it generates the narrative they expect.
This is a fine position to take, but not blindly or ignorantly.
Free kriegspiel, Matrix games, or its lineal inheritor Open Strategy Games don't require grounding in simulationism, philosophically or mechanically. They really only require a spirit of drive, of curiosity, and of appreciation. Everything else is bringing something unnecessary to the table.