A look at Fog of War in General Staff

Today's post is about how Fog of War is implemented in General Staff using couriers. Please take a look: Fog of War post

This kind of thing is always good to see in computer games with limited information arriving late. Will your game show the map with where the enemy was seen or will it give a map area name or some other "foggy" identifier of location ("west of the big hill near town X")?

One thing I did note was that despite the two units being what appears to be a mile or two apart on the map, the enemy unit was identified by name. I think this gives the player too much information. At more than a mile one might only say with certainty that "unidentified troops sighted", between a mile and a half-mile an observer could report if they were cavalry, infantry and artillery and probably their nationality, while under a half a mile more useful items like strengths could be given. You'd never learn a units exact identity until too late (prisoners taken after the action perhaps) unless it was something distinctive like Polish winged hussars, Napoleonic cuirassiers or old guard in bearskins.

The thicker the fog, the better would be my vote.

Well if that really is a guard foot battery, the 7th Wisconsin is going to have a real bad day.

I agree with Martin, a distances of around a mile, very little could be known. Think of cars on a highway at that distance. It would only be at close distances that quantitative assessments could be made. Even then, there would be some probability for error. I think adding some random chance for error should be a component in your FOW routine.

In our SOW games, we deliberately give no information about the enemy forces or even friendly forces not under the player's direct command. In fact, division commanders and above have to query their subordinates if they want an accurate assessment of that unit's condition. Otherwise, the player has to rely on what he can actually see. Since we play at saddle level, that isn't much.

Interestingly, after posting this piece on Fog of War I went and ran some errands and came to the same conclusion! I think we'll just post the unit icon (artillery, infantry, cavalry, HQ, etc.) and leave off the particulars of the unit name (and certainly size).

It's one of those curious situations with computers where we have too much information at our fingertips and as the programmer you just want to share it all!

Thanks for the feedback, making changes now!

I've just made the change you suggested:


Now the question is - and I would very much appreciate your comments - were 19th century telescopes capable of seeing an artillery batter at 8.5 kilometers? I have absolutely no idea. I certainly have the ability to restrict Line of Sight to some specific distance.

I would say not, unless the observer was on a good elevation with exceptional clarity in the air. Cool and dry conditions. I would restrict all sightings of specific units to about 2km to 3km or so even in quite good weather. Beyond that, if it was dry weather you could see dust clouds and because these people of times of yore lived in the open and were used to seeing dust clouds raised by troops and so on, they could have a fair idea of how many troops might be there (single unit, brigade, division, etc).

If the weather is adverse of any sort such as mist, fog, rain, whatever then visibility could be about 1km at most, maybe less.

Rather than an accurate decimal for a distance could you round the value up or down? Seeing range accuracy given to three decimal places dents the immersion a little.

I have read a very good book about the twin battles of Gravelotte-St Privat in 1870 between the French and Prussians whose author and title I now sadly find myself unsure of but I recall that the French on the St Privat heights had an exceptional view of the advance of the Prussian army upon them from the south-west and could determine over distances up to 6 or 7 km the progress of the enemy's advance by means of dust, smoke of battle and the often glimpsed ant-like motions of actual units, but they were seeing an army of 70,000 men or more in motion and knew where to look. As to determining what you are seeing if you glimpse only a single enemy unit at an unknown distance in an unexpected direction, it is not surprising that many intelligence reports received at various HQs over the centuries have been dismissed or were inaccurate.

So, what do you think should be the limit of line of sight observation? 6.5 kilometers on a clear day?

LOS would almost certainly preclude you from seeing anything at those distances.  Even with perfect conditions you might be able to say something was on the road, but not what.  Certainly at those distances it would not be possible to distinguish uniforms and therefore the country.  8 km was about the distance Boney first spotted the Prussians.  I'm confident he had a pretty good telescope, but even so, he had to send light cavalry towards them to find out who they were.

Good evidence for the absolute maximum distance for a 19th century telescope being 8 km. But, as a programmer, I have to put in something as a definite upper bound. 7 km?

Perhaps, I need to research 19th century optics!

If you are including LOS in your calculations, then you could use a number like 7 or 8 km. If LOS is not used, then a much smaller number like 2.5 km would be better.

I'm definitely using a 3D Line of Sight algorithm.

Here's a screen shot of the 3D LOS for an HQ unit on a hill:



Areas directly observe (using a 3D LOS algorithm) are clear; areas not observable are grayed out. Note: woods are calculated at 12 meters.

Can the game's LOS change due to weather conditions?

Sure; it's just the addition of a single variable (distance).
I hadn't originally planned on having different weather for scenarios, but it's easy to implement. The speed for each unit is set for each scenario so it would be very easy to reduce unit speeds as well.