Weather forecast
Moderators: Slitherine Core, Panzer Corps Moderators, Panzer Corps Design
-
kanniethexen
- Corporal - Strongpoint

- Posts: 71
- Joined: Sun Jan 11, 2015 7:55 pm
Weather forecast
It is stated that the weather forecast has an accuracy of 80%. But suppose that the weather forecast is raining. Does that mean an 80% change for rain, a 10% chance of clear weather and 10% chance of cloudy? OR is the chance for cloudy weather somewhat more than clear weather? (for instance 80% - 13% - 7%) as cloudy weather is closer to rain than clear weather.
Re: Weather forecast
Unfortunately I do not know the logic behind the predictions, but the chances for particular weather types are set in the editor for each scenario, and they can be changed form turn to turn or left at their starting values. How (and if) the prediction deals with manual weather changes halfway the scenario I do not know.kanniethexen wrote:It is stated that the weather forecast has an accuracy of 80%. But suppose that the weather forecast is raining. Does that mean an 80% change for rain, a 10% chance of clear weather and 10% chance of cloudy? OR is the chance for cloudy weather somewhat more than clear weather? (for instance 80% - 13% - 7%) as cloudy weather is closer to rain than clear weather.
What I do know is that there are several weather variables that can only be set manually by the scenario designer. The chances (in %) for Clouds/Rain/Snow/Sandstorm can be set manually, the chance for 'Clear' weather is simply 100% minus all the chances for different weather.
Also the starting 'moisture level' of the ground can be set, this determines if the ground is Dry/Muddy/Frozen (snow covered), and this is changed by the weather conditions (explained here: viewtopic.php?f=121&t=40599&p=380399 )
So it is possible to allow only rainy or clear turns without other weather types, so I guess the weather prediction can't give an indication how likely a third weather type would be. Maybe it does some calculations based on the % set and the weather during the past turns? It might be fun to check to see how accurate it really is, I never bothered to test it much.
-
BiteNibbleChomp
- Lieutenant-General - Do 217E

- Posts: 3231
- Joined: Mon Jul 01, 2013 6:35 am
Re: Weather forecast
I think I have worked out how it works:
Say we put into a scenario 60% clear and 40% rain chance.
On T1, there will be a 60% chance that clear will be chosen, and 40% that rain will be chosen. There will be a 60% chance for a predicted clear on T2, 40% for rain.
Let's say it chooses rain for a prediction:
On T2, there will be an 80% chance of rain and 20% chance for clear (If there were two other options instead of one, it would be the percentages of each to be weighted to equal 20%). On T3, a similar setup will apply.
However, the scenario's settings state that on T3, change the weather chance to 100% clear. This will first become apparent on T4 as a prediction, taking full effect on T5.
- BNC
Say we put into a scenario 60% clear and 40% rain chance.
On T1, there will be a 60% chance that clear will be chosen, and 40% that rain will be chosen. There will be a 60% chance for a predicted clear on T2, 40% for rain.
Let's say it chooses rain for a prediction:
On T2, there will be an 80% chance of rain and 20% chance for clear (If there were two other options instead of one, it would be the percentages of each to be weighted to equal 20%). On T3, a similar setup will apply.
However, the scenario's settings state that on T3, change the weather chance to 100% clear. This will first become apparent on T4 as a prediction, taking full effect on T5.
- BNC
Ryan O'Shea - Developer - Strategic Command American Civil War
