Retreat factor

Talk about everything related to CoCoNet or Colonial Conquest!

Moderator: Kroah

Retreat factor

Postby Kroah » 21 Mar 2008, 00:18

Monty wrote:japan's army retreats after exactly 250 losses (1,000/(2^(3-1)))
england's army retreats after more than 4000 losses (8,001/(2^(1))).
every 4 (1+3) played english men 1 japanese would be killed. so, after 1,000 men played by england, 250 japanese would have been killed.
every 2 (3-1) played japanese men 1 englishman would be killed. so, after 1,000 men played by japan, 500 englishmen would have been killed.

so the battle would end after 1,000 played men with the retreat of the japanese army. correct?

so the formula to calculate the minimum of needed attacking men would be:

a(X) !> (a(Y)*(A+T)*2^(A))/((D-F)*2^(D-F)) = (a(Y) * (1+3) * 2^(1)) / ((3-1) * 2^(3-1)) = a(Y) = a(X)

with:
a(X) = amount of the needed attacking men
a(Y) = amount of the defending men
A = attack factor of the attacking army
T = terrain factor
D = defensive factor of the defending army
F = fortification factor (0 or 1)
!> = has to be greater than
Last edited by Kroah on 21 Mar 2008, 00:21, edited 1 time in total.
Kroah
Site Admin
 
Posts: 430
Joined: 07 Feb 2006, 01:01
Location: France

Postby Kroah » 21 Mar 2008, 00:20

Monty wrote:i attacked a fortified german island (125,000 men) with 350,000 men. at 100(germany):250(japan) the battle was over with germany as winner...???
[…]
hmm, i think the island could have been java (terr:3).
but if i use the retreat-formula you gave, then japan wouldn't ever had come down to 250,000 men because they would have retreated after the loss of 11,000 men.
-> 350,000/2^(2+3) = 350,000/32 = 11,000 (and with an added random factor it would be even less men).
so something must be missing in the battle-formula...?

i wonder what that retreat factor is good for at all? it just seems to end battles faster. and your retreated army can escape to an adjacent country. but if there isn't an adjacent country of your colors they would be destroyed entirely, wouldn't they? because of that, in that case i would want them to minimize the enemy's army as far as possible and not to retreat and vanish without harming the enemy.
the retreat-factor sucks, don't you think so?

Monty wrote:you know, when i'm criticising those things, i'm always referring to the "feeling" of the ST-original, which is the only version i ever played.

i know it's not the aim to create an entirely new combat-mechanism, but to recreate the old one, but i feel that due to the elimination of the bugs it's almost like a new game in the combat department.

i already told you that i found those army-retreatments annoying.
on ST, when you failed in conquering a country, the returning men were of a much lower amount. maybe because of the bugged algo, but it "seemed" ok to me. it seemed that your army REALLY tried hard to conquer the target country and did not leave the battlefield before time.
now it seemed like they were a bunch of losers ;)

Monty wrote:I played with england and they're unbeatable. their only bottleneck is the army-costs but they can defend a country with 100,000 men against russia attacking with 2,000,000 men easily.
Kroah
Site Admin
 
Posts: 430
Joined: 07 Feb 2006, 01:01
Location: France

Postby Kroah » 21 Mar 2008, 00:32

Here’s the algorithm of an England vs Russia Army battle:

Example 1: England attacks a Russian area

England:
- Army:1000
- Offensive factor: 1
- Defensive factor: not used

Russia:
- Army:1001
- Offensive factor: not used
- Defensive factor: 3
- Fortification:yes
- Terrain:2

Steps:
- Remove the fortification value from the defensive factor: Defensive factor: 2
- Compute the maximum losses each side can sustain before retreat: MaxLosses = army / 2^factor
EnglandMaxLosses = 500
RussiaMaxLosses = 250
- Add a random value [0;3] to the offensive and defensive factor (for the example 0).
- Add the terrain value to the offensive factor.
Offensive factor: 3
- Init the remainingArmyToAttack for each side to their number of army.
- Battle loop: while no one retreat and at least one side remainingArmyToAttack > 0
- A man is killed to the defending side every offensiveFactor loop (here every 3 loops)
- A man is killed to the attacking side every defensiveFactor loop (here every 2 loops)

Finally, England losses 250*3/2 = 375 men and conquers the area with 625 men left. Russia retreats after 250 losses to an adjacent friendly area with 752 men left.

Example 2: Russia attacks an England area

Russia:
- Army:1000
- Offensive factor: 4
- Defensive factor: not used

England:
- Army:1001
- Offensive factor: not used
- Defensive factor: 1
- Fortification:yes
- Terrain:2

Steps:
- Remove the fortification value from the defensive factor:
Defensive factor: 1 (minimum is 1, England doesn’t need any fortification!)
- Compute the maximum losses each side can sustain before retreat: MaxLosses = army / 2^factor
RussiaMaxLosses = 62
EnglandMaxLosses = 500
- Add a random value [0;3] to the offensive and defensive factor (for the example 0).
- Add the terrain value to the offensive factor.
Offensive factor: 6
- Init the remainingArmyToAttack for each side to their number of army.
- Battle loop: while no one retreat and at least one side remainingArmyToAttack > 0
- A man is killed to the defending side every offensiveFactor loop (here every 6 loops)
- A man is killed to the attacking side every defensiveFactor loop (here every loop!)

Finally, Russia retreats at home after 62 losses with 939 men left. England losses 62*1/6 = 10 men and stays at home with 991 men left.
Kroah
Site Admin
 
Posts: 430
Joined: 07 Feb 2006, 01:01
Location: France

Postby Kroah » 21 Mar 2008, 00:33

Monty wrote:Attacker: Russia
Offensive factor: 4

Defender: England
Army: 1000
Defensive factor: 1

Some (surprising) intermediate results for this specific battle:

- Fortification factor: does not matter
- Terrain factor: does not matter
- Random offensive factor: does not matter

Best case scenario for the attacker (random defensive factor = 3):
4001 men needed to conquer the country
Russia loses 250 men
England loses 500 men

Worst case scenario for the attacker (random defensive factor = 0):
16001 men needed to conquer the country
Russia loses 1000 men
England loses 1000 men

i don't know yet what to make out of it :)
i think it's maybe a bit unfair that russia needs a 4 to 16 times bigger army to conquer an english country, while their army is only 3 times cheaper.
but this is just a first impression. and of course nobody said that it has to be fair...
i think i'll wait and see what you and the others are coming up with concerning the balancing.
Kroah
Site Admin
 
Posts: 430
Joined: 07 Feb 2006, 01:01
Location: France

Postby Kroah » 21 Mar 2008, 00:33

I try to understand the goal of your test. Correct me please if get wrong.
You take the best scenario for Russia (no terrain, no fortification, null random offensive factor) and you compute how many men are needed to conquer the English area defended by 1000 men.

If so, I agree it’s unfair to send at best 4 to 16 times the defending army to conquer the area in 1 turn. Nevertheless, Russia is not a major to conquer an area in 1 turn, England pay the price for this.
Something to analyse, is the ‘army ratio ($100.000)’ statistics. It shows for the whole game the ratio between the kills and losses according to the army cost: (kills/losses*100)/(cost/100,000) (without taking account vanished armies). Needless to say, the highest, the better. The result is strange: among several games, Russia always had the best ratio (~45) and England the worst (~15). It seems the bad factors doesn’t handicap Russia too much according to the price.

Any idea?
Kroah
Site Admin
 
Posts: 430
Joined: 07 Feb 2006, 01:01
Location: France

Postby Monty » 22 Mar 2008, 05:06

ah, i think i have to explain it a little bit more extensivly:
when i play the game with my friends, everyone does his moves unwatched by the others. but when the combat phase starts, of course all of us watch the battles together. the guy who controls the mouse at the combat phase has to scroll to all areas anyone wants to see. naturally everyone wants to watch his own territories to see if some are attacked. because of that it's not very likely for an attacker to be able to minimize a defending army secretly step by step without being detected. therefore it has always been a good thing to conquer a country in one step in winter and then instantly to fortify it in spring.

additionally, before you told me how the battle mechanism works, i always thought that my attack power was better the bigger my army was. or, figuratively spoken, i thought a small army could be "surrounded" by a big defending army and therefore be massacred easier. so i always preferred to build up huge armies and to conquer another country in one step.

Kroah wrote:I try to understand the goal of your test. Correct me please if get wrong.
You take the best scenario for Russia (no terrain, no fortification, null random offensive factor) and you compute how many men are needed to conquer the English area defended by 1000 men.

not exactly. because of my style of playing ("one-step-conquering") i wanted to see if i can figure out a formula that can calculate in each case the needed amount of men to successfully "one-step-conquer" every wanted country (taking all given factors into account).

because there are those 2 unknown random factors, there have to be 2 extreme cases:

- the best case for the attacker (the defence power of the defender is lowered by 3 points while in the same moment the offense power of the attacker isn't lowered at all (+0 points))
- the worst case for the attacker (the defence power of the defender isn't
lowered at all (+0 points) while in the same moment the offense power of the attacker is lowered by 3 points)

as the probability of those 2 extreme cases to occur is only 1/16 you can say generally that if the amount of your army is closer to the worst case than to the best case you can expect to win the battle.

e.g. if season is winter and you are standing near moscow, having something around the best case amount of men i would try my luck to conquer moscow. it's only a small chance but if it's successful it's worth the risk.

if it's another season i would prefer to use those men to conquer more other countries around to gather more money.


anyway, the formula above is wrong, but the one in the excel sheet i sent the other day is right. based on that excel-thing i have now made a chart where you can see how much bigger than the defending army the attacking army has to be.
you just have to collect

- amount of defending army
- defensive factor of the defender
- terrain factor
- your offensive factor
- fortified (yes/no)

and then you can read off the multiplier (best case/worst case).

e.g. france attacking japan in fortified ottoman empire
-> the french army has to be 1,75 (4,00) times bigger than the japanese

of course that chart is top secret, but if anyone wants it, tell me ;)

Monty wrote:Some (surprising) intermediate results for this specific battle:

- Fortification factor: does not matter
- Terrain factor: does not matter
- Random offensive factor: does not matter

there are battle constellations where some factors do not play a role because others dominate. russia vs england is such a constellation.

Kroah wrote:Something to analyse, is the ‘army ratio ($100.000)’ statistics. It shows for the whole game the ratio between the kills and losses according to the army cost: (kills/losses*100)/(cost/100,000) (without taking account vanished armies). Needless to say, the highest, the better. The result is strange: among several games, Russia always had the best ratio (~45) and England the worst (~15). It seems the bad factors doesn’t handicap Russia too much according to the price.

Any idea?

i still have to analyse those statistics closer...

cheers monty
Monty
CoCoNet Beta Tester
 
Posts: 70
Joined: 24 Jun 2007, 23:03


Return to CoCoNet Discussion

Who is online

Users browsing this forum: No registered users and 24 guests

cron