AI in solitaire/co-op boardgames: How should it work?

I’m curious to hear how other people here would handle this rules interpretation situation I’ve run into with the ​solitaire game “Sherman Leader”…

In several places the rules mention that if a German (AI-bot side) unit has more than one move option or targeting option that meet the criteria for the AI-bot rule, that “you choose” what the German unit does.

I’ve been playing a campaign and I’ve noticed that these sorts of “you choose” situations come up quite often. There is one movement rule in particular where you are supposed to move German units to “adjacent cover,” and there are often 2 or 3 options to pick from. Likewise, often I’ll have 2 or 3 units to pick from that a German unit can shoot at.

How you move and fire German units in those “you choose” situations impacts the difficulty of the game quite a bit.

My question is this: How would you interpret the “you choose” situations if you were playing the game? Outside of randomizing the action, there are two ways to go with it. One is that you interpret “you choose” to mean “you choose to your advantage.” Another way is “you choose what you think would be best for the German side.”

I guess it depends on your take on solitaire games. Me, I’d probably decide up front to make the decision that seemed the most tactically sound for the German side; if more than one had a good tactical argument, I’d roll dice I suppose.

Thanks!

Do you think there might be a sort of unwritten rule of wargaming that would lead most players to interpret it the way you would?

Don’t know about unwritten rules, but I always play the “other side” the best I can for their situation. Of course playing the Germans could be an exception where I think “What would Hitler have done?” and then I do the opposite. ;)

Ha! Thanks. :)

I reached out to the designer and got a reply today. They basically said that they playtested it both ways and either way is fine. They seemed to think in the grand scheme of things it didn’t make much of a difference.

It’s a solitaire game. You are playing for your own enjoyment. You do you! Then again, in general, I don’t give a rat’s behind what the designer wants me to do, only what the system/game lets me do.

This is almost always a failure of solitaire game design. You should never have/get to choose what the AI does. That’s the job of the game systems.

I trust Dan Verssen Games’ claims to playtesting about as far as I can throw them. They’ve fallen a long long long way since Hornet Leader. :(

-Tom

Yes. This is the first purely solitaire game I’ve played, so I don’t have experience with anything else, but that seems like it should be the goal of an AI-bot. At first I didn’t think that the “you choose” situations would come up that often, but in the last battle I played they came up quite often.

It would be pretty easy to expand the AI rules to cover these situations. “If you have multiple targets with equal to-hit percentages and equal defense, fire at the one with the highest OP cost.” That would clear up almost all the targeting ambiguity and get the Germans (for the most part) to act in their best interest. The ambiguous movement situations could be cleared up in a similar way, as “adjacent cover” movement AI could be mandated toward the best Allied target when the German battalion is at full strength and away from Allied units if the battalion is at half strength.

Yes, agree with this, and at the same time I was curious what the designer’s intention was. I wondered how they played it.

I think I’ll tweak the rules a bit to cover the “you choose” situations and get the Germans to act in their own interests.

Ambiguity and heavy luck elements aside, I’m enjoying the narrative the game creates.

I disagree. Many designs (like Spirit Island) always let you choose to your benefit when there are different options and are balanced that way.

Making a player to decide “what the AI would do” against that player’s benefit is indeed more of a failure.

But anyway, it should be clear, I agree on that.

Agreed. Having the player decide whichever option they think will benefit themselves the most is a totally valid design approach (if sometimes thematically muddy). The game is still a string of decision points, and at each one you’re making the choice that you think will be most advantageous to yourself.

But having the player make a judgement call based on anything else immediately breaks that chain and forces them to second guess the spirit of the game and handicap themselves, which undermines the rest of the game’s tension.

I think it depends on the the theme. A game based on historical or realistic events, like many wargames, intrinsically presents you with a distinct context that exists outside of the dynamic of the game itself. So if you are playing a game like Sherman Leader, you have an existing, known framework called “German tactics (or doctrine)” that, reasonably, should govern German actions. The game rules should reflect this, and if a situation emerges where you have to choose, as the player, choosing something that is historically accurate is just as logical or justified as choosing something that makes sense in the spirit of the game (that is, something that assists you as the player).

If you have a game where the only context is you vs. the game system, I agree, it makes more sense to choose a beneficial result. You only have that one context. In a historical game though I think it is equally valid to err on the side of realism. Depends I suppose on why you, the player, are playing the game.

Me, I play wargames to recreate or simulate real or possible events. When I solitaire a game, I always choose the most beneficial results for the side I am playing. It’s a bit different than playing a specifically solitaire game, sure, but the principle is the same. Would the Japanese or Russians or Zulus have done A, or B? Of course, usually I’m playing both sides, so it is not as important as in a true solitaire game.

I do agree 100% that any choices the player has to make should arise from designer intent, not from failure to write good AI rules.

2 posts were merged into an existing topic: Grognard Wargamer Thread!

But then, instead of relying on the player to have internalized that “German tactics” framework and be disciplined enough to apply it regardless of the situation, wouldn’t it be better to make it an explicit part of the solitaire rules? Have an order of priorities or a flowchart or something to determine what the opposing force does?

Of course, playing both sides to simulate or explore history is totally valid too, but to my mind the point of having a dedicated solitaire game/mode is to give an experience closer to a computer game, in that there’s a particular situation or obstacle that I have to overcome, using all resources available to me to the best of my ability.

I want to be playing against the system, not myself. It’s unsatisfying to get to the end of a close game and think that I won, but that maybe I wouldn’t have if I had made more optimal choices for the opponent.

This was my mindset as I was playing Sherman Leader: If the system was letting me choose, I assumed I should be choosing something that aids me, as I’m playing against the system. I’ve been making some videos of the series, though, and one viewer felt the victory was tainted because I wasn’t choosing optimally for the Germans.

That got me thinking that perhaps this is another wargamer unwritten rule I wasn’t aware of, like never having drinks on the gaming table, or how you’re supposed to stuff random counters from old wargames in friends’ couches when you visit.

What? I mean, seriously, what? I don’t even…

I’m astonished that you’re saying this, because it is demonstrably false. Spirit Island never makes you decide what the AI will do. Ever. The AI adheres to a specific set of steps about what it will do, where it will do it, when it will do it, and how it will do it. It is the exact opposite of what @Godzilla_Blitz is talking about. You never have to make decisions on the AI’s behalf.

I defy you to show me when that happens.

Are we even talking about the same thing? You’re saying making decisions on behalf of the AI is “a totally valid design approach”? Can you give me an example of a game that you’re talking about? Partly because I never want to play it, but also because I’m wondering if you and Juan are talking about something completely different.

-Tom

Dude, there are so many better options that whatever latest boondoggle Verssen Games has foisted off on you. :(

-Tom

I’m enjoying Sherman Leader, tbh, but I’m really impressed by how single-player wargame options have expanded. I have a few stacked up in the queue that I’m looking forward to:

Fields of Fire 2
Silent Victory
Enemy Action Ardennes
D-Day at Omaha Beach
Target for Today
Storm above the Reich
RAF: Battle of Britain
Empire of the Sun (w Erasmus 2.0 bot) (arriving Thursday this week)

In particular, I’m looking forward to D-Day at Omaha, RAF: Battle of Britain, and Empire of the Sun

Good luck to you on Fields of Fire. I just gave up on that one after many hours of confusion. I understand there are new rules, which might help.

Empire of the Sun is fantastic, of course, and I had a good time with D-Day at Omaha Beach. RAF is supposed to be good, but I haven’t had the chance to play that one.

I’m not sure how great/valid the design is, but Dungeon Alliance maybe?

As I’m sure you know, it has pretty basic AI selection criteria like ‘ready the highest XP monster’ or ‘target hero with lowest health’. There are often ties, and these are always resolved by the player however they see fit. *I think it also gives basic minimum directions, like needs to attack or have line of sight or something.

Since the game is basically a puzzle, I think that factors into your decisions as you try to ‘game’ the situation.

However, in the multiplayer game you are also moving monsters to benefit yourself and hinder others, so it seems to fit that the AI works a bit like this too.

A post was merged into an existing topic: Grognard Wargamer Thread!