Unpredictability of weather

[RonWood]

The feature list on the Matrix product page includes "Sophisticated weather model". Can someone elaborate on this sophistication?

I've used the history mode to look at some of my games on the Edelweiss scenario (a fairly large map). Oftentimes, an area of rain appears rather spontaneously in a seemingly random location (sometimes on the edge, sometimes smack in the middle). And, just as often, the rain completely disappears from the map on the subsequent turn. On occasion, this weather has disrupted some of my otherwise well thought out plans.[;)] While I like some unpredictability, I'd also like some vague idea where this rain might be on subsequent turns. I feel rather helpless after spending an hour or more in a game, only to have a random rain shower at an inopportune location.

To combat this feeling of helplessness, I'd like to see a weather system that moves with a perceptible speed and direction. I'd like it start on the edge of the map, and move across as turns are completed (or, if it appears turn 1, I'd like to see where it was prior to turn 1). Given this, I could better account for the weather when choosing movement, and planning (or avoiding) locations for attack. I'd also like to see some weather systems that are "longer", more linearly shaped - ones that would "sweep" across the map in a broadside direction like weather fronts often do. At the moment, I've only seen rain that appears in a largely-circular pattern.

One way or another, I believe some interesting strategic depth can be added here. How might the AI deal with it? I'd guess that's probably the most challenging part of the implementation.

Great game nonetheless. I've 20+ games under my belt so far, and really enjoying it!

[2xTom]

Hi, glad you like the game [:)]

It's also good that you noticed the weather engine! We felt we could call it "sophisticated" on the product page we actually consulted a for-real meteorologist when creating it. [;)]

The initial design included precisely what you just mentioned: weather "fronts" moving across the map in some realistic fashion. We realized early on however, that this creates problems with replayability and moved on to the current "stateless" design. Let me explain...

The weather engine will indeed create "random rain showers at inopportune locations". Conversely, it might create a "random thaw" in places with pretty much the same effect. This is OK from a historical standpoint I think, as these weather events in fact did happen and did affect operations. Now, if we make the weather engine move these things across the map, the unwanted effect you get is that once you get a mud patch in a tricky location, it gets more likely that the mud patch stays near that location for several turns.

So what happens is either you have good weather and the scenario is a breeze, or.. you run into a cursed patch of rainy weather that completely ruins your day. I admit the latter is not unrealistic, but we opted to keep the scenario weather reasonably similar between different runs. This is why we have weather that is in line with the climate, but otherwise completely random on every turn.

I hope this helps, or at least sheds some light on our design decision there.

[RonWood]

Given the size of the game's current weather systems, I can certainly see your point that one hex may be muddy many turns over. I agree that might be a real nuisance, especially given the short length of some scenarios. I was actually envisioning weather systems that moved faster than their overall size (given the four-day turns). Given my ignorance of the size and speed of the weather systems for this locale, and variations in the size of the scenarios, I can see what I proposed is possibly intractable.

But I'm also thinking ahead and in the abstract, independent of the game's current locale and scenario size (sorry, I'm a software developer). If 2x2 envisions the game engine independent of this specific geographic locale, it'd be a disappointment to see the overall possibility precluded. But I'd get over it. [:)]

Appreciate your insight and looking forward to seeing this game evolve!

[2xTom]

No prob, thanks for turning an eye to the weather model. It's encouraging that people don't notice it for the most part ("it feels right") but strictly for vanity purposes, I enjoy this conversation a lot [;)]

IIRC, when we looked at these options, the "stateful" model you describe didn't look any more complex than the current "stateless", just less appropriate for given map scale and turn length. If these things change, we may in fact add some state to the weather model. We'll see...

[SeaMonkey]

Actually when you think about the weather model being unpredictable, it incorporates, to a degree, a feature similar to what the missing FoW mechanism would simulate. So, in UoC, this weather pattern has even a more significant effect in gameplay. Just think what a FoW scheme would add!

[2xTom]

This is again connected to the specific scale of this game. We felt the game work well even without FoW at this scale. There are other scales (I'm talking about unit and map size here) where I think FoW would be absolutely necessary.

You're right that both the weather and FoW add to the "unpredictability" element. Are they are interchangeable in that respect? I'm not sure. I don't think UoC is lacking unpredictability at any rate. Both testers and now players report plenty of variety in how things play out for them.