[smacktoward]

Good article. Modeling FoW is one of the classic hard problems in wargame design, especially for boardgames, where the options for hiding information from one player without hiding it from all are limited.

The problems with "Good Guy" FoW go a little deeper than the article indicates, though. It correctly identifies human error (a platoon leader misreading a map, for example) as one reason why a commander's intelligence about the status of his own units may be incorrect, but there's other factors that can cause such problems as well, such as:

- Communications latency: the time between an order being issued and its being received can be significant, and lead to units following orders that were based on a situation which no longer exists, or disregarding orders because by the time the orders arrive the situation looks different to them than it did to HQ

- Subordinate initiative: lower-ranking units have their own leaders too, and those leaders have free will just like the Big Brass at HQ does, which means that sometimes they take action on their own authority, either in the absence of or in contradiction to existing orders

A textbook example of the latter problem is Union general Dan Sickles' advance on the second day of the Battle of Gettysburg (see: https://en.wikipedia.org/wiki/Battle_of_Gettysburg,_Second_D...). Sickles' corps was part of the long "fish-hook" defensive line that the overall Union commander, Gen. George Meade, had formed his army into after the Confederate assaults the day before. But, seeing slightly higher ground in front of his position, Sickles on his own authority marched his corps forward to occupy that ground. The result was disastrous; Sickles' troops were too far forward for the rest of the Union line to support them, so they were smashed when the Confederates hit later that day.

This type of thing is extremely hard to model in a wargame, because ultimately players want to feel that their winning or losing is down to the decisions they made, not due to random factors or the game itself pushing against them. So, while balky subordinates like Sickles are not uncommon in real life, how do you model that in a way that doesn't make the player feel that the game is essentially playing itself? It's a sufficiently hard question to answer that most designers don't even bother to try.

[ctdonath]

Maybe a key lesson is that the game is essentially playing itself, that in battle so many factors work independent of your intent that one must accept that there is little one can do to change the outcome, and how important it is to insightfully work that little one can do.

[smacktoward]

Yeah, it's possible to imagine a game whose objective is essentially pedagogical -- to teach the player that generals in war aren't really in control of much at all, that they're just trying the best they can to hang on to a bucking bronco.

Would anybody play that game, though? Is there a way to make a game fun (or at least tolerable), when its core mechanics are all designed to frustrate your intentions? I'm not sure.

[msandford]

That game is called Fluxx. A part of the game is that you can play cards that change the rules. So people draw cards and work on strategies that inevitably change before they can be finished. Eventually someone looks at the board and realizes that they've won. Usually completely by accident.

Personally I hate it, but there are plenty of people who like it too.