Combat simulator question

Printable View

10.07.13 18:57
Stecker78

Combat simulator question

Think I need to help from a more experienced player to explain me some things about the combat similator.

Following simulation (2200 times):

Bandit camp: 60 guard dogs, 20 roughnecks, 120 rangers
My general: 200 recruits

Min. loss: 139 recruits
Max. loss: 168 recruits

Alternative simulation (of course also 2200 times):
My general: 173 recruits, 22 militia, 5 soldiers

Min. loss: 140 recruits
Max. loss: 171 recruits

How's this possible? I would say that my army in the alternative similation is stronger at any time.
10.07.13 20:14
Sto_Helit

Quote:

Originally Posted by Stecker78

Think I need to help from a more experienced player to explain me some things about the combat similator.

Following simulation (2200 times):

Bandit camp: 60 guard dogs, 20 roughnecks, 120 rangers
My general: 200 recruits

Min. loss: 139 recruits
Max. loss: 168 recruits

Alternative simulation (of course also 2200 times):
My general: 173 recruits, 22 militia, 5 soldiers

Min. loss: 140 recruits
Max. loss: 171 recruits

How's this possible? I would say that my army in the alternative similation is stronger at any time.

It's still approximate, not a guaranteed exact number of losses.
10.07.13 20:43
Fadito

If you use this simulator http://settlersonlinesimulator.com/d...fsimulator/en/
and follow at the bottom "Example simulation" (click) - "Round 1" - "Normal Strike" - "Caused damage" (click) you can find that all your militias and soldiers will attack a ranger. but whether a recruit, militia or soldier is attacking a ranger does not matter since a ranger only got 10hp and is killed in 1 hit...

So this 2 armies are in this case basically the same. For the rest, like Sto_Helit wrote, "not a guaranteed exact number of losses". Just the minimum/maximum out of 2200 random calculated fights.
10.07.13 22:09
MrBranch

Without meaning any offense, but I think the "approximate, not a guaranteed exact number of losses" has nothing to do with it.
I'm getting the same simulation results, no matter how often I'm simulating. Anyone with other results?

When my assumption is correct that all of us get the same simulation results, then we are facing an ordinary calculation bug (even when Fadito is right when stating "these 2 armies are basically the same").

There is one thing I don't understand though: why is Recruit #1 attacking Guard Dog #1. Isn't the first attack done on the Roughnecks?
Or does it work differently for Cavalry than for Guard Dogs?
10.07.13 23:52
Fadito

Quote:

Originally Posted by MrBranch

I'm getting the same simulation results, no matter how often I'm simulating. Anyone with other results?

No one. :)
The calculation is random, but once made it is stored. If you want a new one try 173 recruits and fill it with a random combination of 27 militia/soldier/elites. If it is a fresh calculation you should see lower simulation counts - e.g. "The simulation was run 100 times". If you then simulate again the simulation count goes up and the result can still change. But once it reached the 2200 it is stored and just shown again.

Quote:

Originally Posted by MrBranch

There is one thing I don't understand though: why is Recruit #1 attacking Guard Dog #1. Isn't the first attack done on the Roughnecks?
Or does it work differently for Cavalry than for Guard Dogs?

I'm not sure but think that the simulator attacks the bandits always as given in the input boxes from top to bottom. Whether it is correct - i have no idea...
11.07.13 06:25
Stecker78

Tried one with:
165R, 30M, 5S

Now min. loss is 137R, max. loss 165R.
That seems to be more logical.

-----------------------------

But, when attacking with 171R, 24M, 5 S:
min. loss 136R, max loss 167R.

The alternative solution earlier mentioned seems to be a strange unlikely result indeed. No more then coincidence perhaps...
11.07.13 09:44
Gerontius

Quote:

Originally Posted by Stecker78

The alternative solution earlier mentioned seems to be a strange unlikely result indeed. No more then coincidence perhaps...

How many people do you need to tell you it's random and coincidence before you'll accept that it's random and coincidence?
11.07.13 18:19
Stecker78

Zero if you read my last sentense. And since you did quote it yourself, I think I have to recommend you some English courses?

And yes, I wrote perhaps....I simply don't believe in coincidence.
11.07.13 18:27
peck_ed

Okay guys, lets try and tone this down please

Thanks.
11.07.13 21:42
Tripi
I really would suggest you try http://www.settlersonlinetools.com/en/combat_simulator/ rather than the simulator you are talking about. A little information (some previously stated) about them both:

http://settlersonlinesimulator.com/d...fsimulator/en/
- Performs the calculations on the server
- Runs a maximum of 2,200 simulations (per unique combat setup)
- Stores any maxed simulation (2,200 runs) that it deems as a top example (top 50 or so garissons)
- Can quickly suggest a 'best option' for most known fights
http://www.settlersonlinetools.com/en/combat_simulator/
- Performs the calculations locally (within flash on the users system)
- Runs a maximum of 99,999 simulations (though this can be re-ran multiple times)
- Does not store any 'previously' ran simulation
- Will take time to calculate a 'best option' due to re-running the simulation and mathematically testing all permutations
- Can also suggest best tactic for max rounds, min rounds, cheapest value losses etc
- Trickier to use. Gives a lot more information though
There are pre-run simulations on the one you were talking about that 'show' less losses if you send less troops (rather than more troops); simply because on that particular run the result was lower for the lesser troop version. This does not mean you 'will' lose less by sending less, just means in a simple 2,200 simulations the lower troop simulation had luckier math results.
11.07.13 22:34
Gerontius

Quote:

Originally Posted by Stecker78

Zero if you read my last sentense. And since you did quote it yourself, I think I have to recommend you some English courses?

And yes, I wrote perhaps....I simply don't believe in coincidence.

You asked for comments from more experienced people and you got them, answers which were correct and logical. Yet you continued to argue the point. Why? If you're not going to accept explanations from people who know the game better than you, then why ask for them?