4545

That is a statement I completely disagree with.

For a number of reasons, from playability (number of steps that one needs to process in order to resolve an action) to mere logistical constraints (number of counters for instance) on the gaming materials, board games do have to compromise a lot in fidelity of the processes. That’s self-evident I think.

Another different matter is whether the outcomes of the games, as in specific landmark events and the order they happen in the game, rather than the process of mechanically playing the game, are consistent with historical constraints and/or a reasonable counterfactuals.

That is the magic of board games, you can “bake in” a range of possible outcomes in the rules to generate that illusion of “realism”. To what degree the outcomes need to be “baked” into the rules before “crying uncle” on credibility is still a controversial matter: P. Sabin wrote quite a bit about that, and I don’t think it is resolved.

A magic which also can be readily be included, and in fact it is included, on most computer war games.

The AI would have to be capable of doing the same, which would be a daunting task, but not so much on the coding side as on the data identification, acquisition, and contextualization side, as the developer would need to build in effect an expert system applying a set of concepts adapted from historical practice for the use of the computer.

A “daunting” task… until somebody actually does it. In very specific games and sweet spots, it has been done already. Also, and interestingly, these three “wins” I link to have been associated to funding by defence organisations, as tools for training and/or operations research.

Neither of these things required huge resources and they’re the three examples of which one can find publicly made statements on technical aspects.

4546

Tiller’s stuff was turned down for having inadequate AI for DoD exercise purposes (his AI “method” is to have the human attack all the time), I don’t really see that as Star Trek M5 computer “win.”:)

4547

Wait, I thought you were talking solo digital vs ftf board. You are moving the goalposts!

4548

Have you played the Midway game? Allegedly it has fully functional AI - so I would appreciate a link for that.

4549

Yeah, I played some Campaign series a couple weeks ago and uh, the AI is very, erm, incapable. I mean Panzer General’s AI was the same thing- have the player play against a clock and occasionally send out a tank to bother him.

4550

No, I have been consistant. Online MP and FTF Board. PC or Vassal as the method, I could care less.

I’m the guy who said “save the AI for the tutorial” remember? :)

You know…Mr. Grognard Sunshine?

4551

Well, then I’m discussing something else. We can consider our discussion null and void :)

4552

My Uncle Joe told me. :)

Seriously, It was turned down for Joint Exercises. I was at a few of the meetings. The Army, USMC and DoD had a higher standard than the USAF did with regard to Tillerware. So was Soviet Afghan War.

4553

I’ll never get those moments back, B. I need that Scotch now…

4554

I consider online m/p (in wargames) to be such an infinitesimally small section of the market that I can’t really discuss it. It certainly isn’t driving design. Design drivers are solo digital vs ftf board, in which the markets and experiences are completely different.

I have heard some (certainly anecdotal) assertions that even VASSAL is used more for solo play than for opposed. People like having the maps and counters set up permanently and easily for solitaire.

4555

Ah, pesky Uncle Joe. :D

Well my reading of your remark is that the key issue was the matter of joint exercises. Tiller’s aerial simulations of course works for USAF, because USAF thinks they are the only branch of the US military that does something useful. Of course, I can see Tiller’s stuff not being suitable for the USMC - which is pretty much about joint naval, aerial and land operations - or the DOD.

Note as well the “sweet spot” bit… generalizing and composing existing systems is where it is at at the moment.

4556

For Simulations we used…drum roll…Real People as the “AI”. :) Because AI is bad at Stragetty. And Tacticalnessesses

4557

Yeah, that is that whole “I pushed some cardboard around”. Is that a “play”. I’ve never figured that out. Nobody has any durn numbers. What is a “play”? :)

Anyway. We are purely in “What I like!” land. There ain’t no objective data. And that’s cool.

But I have multiple turns of that “minuscule portion of the market” to attend to. :) And some TOS thread work. So I hereby consider this horse dead. In a sunshiney way. :)

4558

generalizing and composing existing systems is where it is at at the moment.

I mean it for real. No passive aggressiveness at all - it is a statement of fact.

If you mean the kind of “AI” as “programming a fixed set of behaviours by hand” yes, you’re totally right.

4559

I accidentally quoted you on the first comment up there. Inadvertent. Ignore that. I just mean to respond without highlighting text of yours. I think we agree. But now, this stallion is:

4560

Haha, great gif I will borrow it :)

4561

I think we are working at cross-purposes here. When I talk about fidelity, I’m talking about the relationship between a deep historical understanding of the subject being simulated/gamed, and the execution of systems designed to exploit or implement that understanding. I find that in computer games, the need to have two separate skill sets, programming and game design, usually undermines this sort of fidelity. In board game design, there is far less between the idea the and the implementation. This is why, in my opinion, so many computer wargames end up counting rivets. It’s easier to develop models, from a programming stand point, that add up values than it is to implement complex simulations of doctrine and battlefield practice, which are easier I think to do in a board game environment, where you can rely on the interpretive powers of the players to some extent to grasp the intent of the systems.

In other words, I do not equate fidelity with numbers, but ideas, but computer games are always, always, about numbers in the end.

As for how daunting the AI work is, yes, you are correct that from a programming perspective it’s quite doable. The trick is finding a way to have the right design skills and the programming skills in the same spot. One-man teams cannot create games that meet the needs of today’s gamers, in most cases (graphics, interface, size, performance, etc.) and larger teams have a real problem with communicating complex design parameters between designers and programmers in ways that work for everyone. Then there’s money; Tiller himself told me that the reason he doesn’t do much with AI in his games is that it doesn’t pay. No one cares enough to pay what would be necessary to cover the costs for good AI. They will pay for looks, or other things, and that’s where most developers wisely put their money.

So I will agree to disagree, I guess; I find the quality of military history-type deep thought in the best board wargames higher and better realized than that in most computer wargames. The demands of the two mediums, and the commercial climate for both mediums, drives this I think. I do not think this is absolute, or that there are no computer games that are really solid here, or that all board games are great, far from it, just that in general this is what I find. YMMV.

4562

@TheWombat I see the question of “fidelity” very much the same way. When you add units, for instance, you may actually be pushing the design in a more ahistorical direction because of the interaction of those units with the system. Boardgame designers are forced to balance this in a way computer game designers aren’t. The boardgame representation is much more a part of the design. Because units, terrain, and whatnot are much more easily changed in a computer setting, it’s easier to just through more “history” into the game in the form of data without realizing you’re pushing the history off the rails.

I very much agree with your entire post.

4563

Not sure about that Bruce to be honest. I have mentioned Sabin’s writings on the matter - and the somewhat ferocious criticism - which exposes the issues of overfitting rules to outcomes.

In the same way one could fudge things by giving too much agency to players in a computer wargame, which I think it is usually the most common issue as the player has to essentially many command roles at the same time, in board games there is plenty of examples of rules fitting too perfectly historical outcomes as understood at the time of making the game.

As interpretations of history change due to new data being available, those games turn out to obsolete in a very fundamental way.

Certainly, pure data based design isn’t a guarantee as you can find examples of biases in selecting which data is relevant. So I would say that as long as the human element is there, selecting, composing and summarising there is going to be an element of fancy to these games, computer based or cardboard based.

My bottom line: The cardboard format is not a _superior _ format, it is cheaper to work with and better understood when it comes to tell apart what works from what doesn’t.

4564

As for fidelity, clearly you use a peculiar definition @TheWombat, but to each their own.

I am not sure how one benchmarks ideas, to be honest.

I think that the passing remark

is quite obvious and the converse is true too. The best computer wargames surpass most of the carboard ones too.