Page 1 of 2

Fight twice during combat phase

PostPosted: 21 Mar 2008, 00:40
by Kroah
Monty wrote:i (japan again) was in guiana and argentina.
france (cpu) was in brazil.
of course i cheated to see what was about to happen: france would attack guiana from brazil with 1,000,000 men leaving 100,000 back.
i had enough men in guiana to defeat them and i myself attacked brazil from argentina with 700,000 men.

then, in the combat phase, at first the brazilian attack against guiana was performed. they lost 100,000 men and 900,000 returned to brazil.
only then the argentinean attack was executed with the 700,000 men having to fight against 1,000,000 men (and not the expected 100,000).
of course japan lost quickly :(

in my sense of logic it must not be possibly for one army to fight in two battles in one round. you should have to decide either to conquer a new country and leave the original country somewhat vulnerable or to protect the original country with your army, but you can't have both.

maybe it was the same on ST, but we, the players, did not notice it because the amount of returning men was (due to the bug?) unimportantly small and did not affect the next battle.
what do you think about that?

PostPosted: 21 Mar 2008, 00:40
by Kroah
I agree with you. An army could not attack twice the same round.

I’ve checked the 800XL and ST version with CoConet and the three are identical. Like you said, this was not so relevant on ST because of the small amount of retreating army (because of the bug). The best solution would be to retreat the armies after the last round (at the end of the combat phase), but it’s not really easy to implement at this time.

Any idea?

PostPosted: 27 Mar 2008, 10:23
by Monty
example:
RUSSIA owns country A
JAPAN owns country B and C
moves:
RUSSIA attacks country B from country A with army R1, leaving army R2 in country A
JAPAN attacks country A with army J from country C


at first i thought it should be executed like that:
in round 1 (out of 9) of the combat phase army R1 attacks country B. if they lose, of course they want to return to country A immediately but instead "have to be hold" where they are, because while they are away, other moves have to be executed (e.g. JAPAN attacking country A from country C). at the end of the combat phase (in or after round 9) they are allowed to return to country A. if country A isn't of their colors anymore they may flee to a bordering country of their colors if available. if no such country is available they have to vanish.

but recently i began (again) to think that all "problems" derive from that retreat-factor.
i mean it's not illogical that the returning army R1, that finds their origin country under attack, fights together with the defenders R2 against the attackers.
the problem is that because of the retreat-factor:
- the returning army is almost as big as they were when they left
- the returning army is returning too early, because their battle for country B is broken off early (due to retreat). if they fought to the least man the battle would have lasted a few rounds before they could return home. in practice, i never witnessed the situation that the attack of army J began before army R1 returned home, though it would be logical as both armies R1 and J make their attacking move in round 1...pascal, can you explain why they are always back before JAPAN's attack begins?

conclusion:
i would like to test the game with retreat factor switched off ;)
plus the combat moves should be executed in a realistic timeframe (if i haven't seen it wrong and they aren't already)

PostPosted: 27 Mar 2008, 23:14
by Kroah
Monty wrote:in practice, i never witnessed the situation that the attack of army J began before army R1 returned home, though it would be logical as both armies R1 and J make their attacking move in round 1...pascal, can you explain why they are always back before JAPAN's attack begins?


Army battles (not navy) of the first round are all played randomly. I've just checked and it works fine (Germany vs Russia). Army J could attack before army R1 returns home (50%). Whether you had bad luck or an unknown bug.

Of course, it the attack comes by navy, it's another story: the attack takes place when the round is equal to the distance separing both areas.

Monty wrote:i would like to test the game with retreat factor switched off ;)

I understand your frustration. I've thought how to resolve this issue, but only few ideas come to mind:
- we could let the army attacks until round 9 where it will retreat. The retreating army will be lot smaller, but this is like doing 9 combat phases in 1...
- (your idea) we could let the army attacks as is, but the army will retreat at home at the 9th round (not bad, but hard to implement yet).

Monty wrote:plus the combat moves should be executed in a realistic timeframe (if i haven't seen it wrong and they aren't already)

By "realistic timeframe", do you mean the random move orders within a round? I don't understand your remark "if i haven't seen it wrong and they aren't already", sorry.

Sincerly, if we find an idea to have smaller retreating armies, the ramdom order of the moves don't annoy me anymore, and you?

I think to do moves of the type "friend to friend area" first. This will make chain moves useful and less random (to attacks). What do you think?

Thank you for you feedback,
Pascal

PostPosted: 27 Mar 2008, 23:37
by Monty
Kroah wrote:By "realistic timeframe", do you mean the random move orders within a round? I don't understand your remark "if i haven't seen it wrong and they aren't already", sorry.

no, nothing to complain about the random move order at this time, i was talking about the duration of fights, but as it seems i haven't seen it correctly or, as you assumed, have had bad luck.
Kroah wrote:Army J could attack before army R1 returns home (50%). Whether you had bad luck or an unknown bug.

i have never seen this behaviour.
whenever i played, R1 attacked country B and if they lost they were back in country A before J attacked that country.
because of that i assumed that the fight of R1 wasn't executed in a realistic duration (of 1 to x rounds). i thought that if the random factor picked the R1-move as the first to be executed, everything that relates to it (the attack, the fight and the return) is played before the next move of the random order is executed (here: J attack).
but obviously this assumption was based on incorrect observation due to bad luck because you have seen it playing in the correct way...

PostPosted: 27 Mar 2008, 23:49
by Monty
Kroah wrote:- we could let the army attacks until round 9 where it will retreat. The retreating army will be lot smaller, but this is like doing 9 combat phases in 1...

ah, you mean that every or many fights will then last the full 9 rounds?
ok, but how is it played in the bugged version? it must be played often to the 9th round, too?
i think i still haven't understood completely when a combat round is finished and the next one begins. is it based on time or killed men or...? could be that you already explained it but i have forgotten...

PostPosted: 28 Mar 2008, 02:39
by Kroah
Monty wrote:
Kroah wrote:- we could let the army attacks until round 9 where it will retreat. The retreating army will be lot smaller, but this is like doing 9 combat phases in 1...

ah, you mean that every or many fights will then last the full 9 rounds?
ok, but how is it played in the bugged version? it must be played often to the 9th round, too?


You can see the bugged version like the current one but each side attacks his self instead of attacking each others... We can say the losses of both sides are independent of the opposing army (!)

Monty wrote:i think i still haven't understood completely when a combat round is finished and the next one begins. is it based on time or killed men or...? could be that you already explained it but i have forgotten...

Here's a simple example for a lone army move for the whole combat phase:

- round 1:
(A) The battle begins, each side attacks until one side retreats or all men of both side have attacked:
* if the attacking side retreats, the battle stops, the army move is ended and the attacking army try to return at home.
* if the defender side retreats, the battle stops, the army move is ended, the area is conquered and the defending army try to retreat to a friendly neighbor area.
* (B) if all men of both sides have attacked, the army move will be resolved again for another battle at the next round with the remaining men. Start again at (A) for round 2.
* if at round 9 the army move isn't ended, the army move is ended and the attacking army try to return at home.

As you can see, time is not used.
If the army move isn't ended when the attacking side retreat (before round 9), there will be 9 battles instead of 1, so more kills for both sides.

On ST, (B) is:
* if all men of >one< (not both) side have attacked, the army move will be resolved again for another battle at the next round with the remaining men. Start again at (A) for round 2.
-> This is why lots of battles lasts 9 rounds on ST.

Here's a simplified example (ST version, same random through rounds):
England: 1000 men (factor 1, retreat 500, random 1 -> 1 loss every 1):
Japan: 50 men (factor 2, fortified, retreat 25, random 2 -> 1 loss every 4):
Round 1 stops when 50 Japan men attacks (50/4=12 losses)
England loss 50 men (50/1).
Round 2 stops when 38 Japan men attacks (38/4=9 losses)
England loss 38 men.
Round 3 stops when 29 Japan men attacks (29/4=7 losses)
England loss 29 men.
And so on...

As we can see, England kills itself at a high rate! Russia (and high factors majors) are really overpowered, they loss few men.

I really don't understand why ST version is done this way... I call this "the big bug" until someone explain me the reason of this new behavior (i've decompiled 3 different ST versions, same result).

PostPosted: 28 Mar 2008, 04:01
by Monty
thanks again for your detailed explanations, pascal.

Kroah wrote:Here's a simplified example (ST version, same random through rounds):
England: 1000 men (factor 1, retreat 500, random 1 -> 1 loss every 1):
Japan: 50 men (factor 2, fortified, retreat 25, random 2 -> 1 loss every 4):
Round 1 stops when 50 Japan men attacks (50/4=12 losses)
England loss 50 men (50/1).
Round 2 stops when 38 Japan men attacks (38/4=9 losses)
England loss 38 men.
Round 3 stops when 29 Japan men attacks (29/4=7 losses)
England loss 29 men.
And so on...

As we can see, England kills itself at a high rate! Russia (and high factors majors) are really overpowered, they loss few men.


ah, the needed losses to trigger the retreat are resetted to zero in each battle round and not added together? because if they were added the combat would end after/in round 3 (in fact after/in round 2 because JAPAN has defensive factor 3 and therefore retreats after the loss of 12,5 men, i think).

but apart from that, having the retreat factor the same on ST, wouldn't it be the same amount of men who retreat? my head is smoking...
i mean if random factor was 1 instead of 2, then J had 1 loss every 3 men. retreat still after 12,5 men. then in the first round the retreat amount would be hit and 38 men would retreat...but that wasn't the case on ST, wasn't it? or maybe i'm mentally blocked at the moment, i will look at it again later...(those statistics are to blame :wink: )

PostPosted: 29 Mar 2008, 04:08
by Kroah
Monty wrote:ah, the needed losses to trigger the retreat are resetted to zero in each battle round and not added together?

You're right, the needed losses are reset for each battle.
Example:

* Round #1
- battle #1 (major 1 vs major 2) <-- conquered
- battle #2 (major 4 vs major 3) <-- no retreat
- battle #3 (major 2 vs major 3) <-- retreated

* Round #2
- battle #1 (major 4 vs major 3) <-- no retreat

* Round #3
- battle #1 (major 4 vs major 3) <-- conquered

Each battle initializes its own variables, including the needed losses for retreat, the random factors, ...

Monty wrote:but apart from that, having the retreat factor the same on ST, wouldn't it be the same amount of men who retreat? my head is smoking...

Yes but... both the retreat factor *and* the 'power' (offensive or defensive) factor must be took into account:

- on ST, the japan uses its own power factor to compute the 'needed loses for retreat) and he attacks his self. Battle ends when all men of *one* sides have attacked (few losses for each round, but higher overall losses for all rounds) or one retreat or conquer.

- on 800XL, the japan uses its own power factor to compute the 'needed loses for retreat) (like ST) and he attacks his opponent. The opponent power factor is used to attack Japan. Battle ends when all men of *both* sides have attacked (higher losses in 1 round from opponent which could lead Japane to retreat, but lesser final losses) or one retreat or conquer.

There are other examples for wich battles are like the example i gave you in my previous post (huge army can't beat a little army of 50). This represents more case than expected in fact.

Monty wrote:i mean if random factor was 1 instead of 2, then J had 1 loss every 3 men. retreat still after 12,5 men. then in the first round the retreat amount would be hit and 38 men would retreat...but that wasn't the case on ST, wasn't it?

This is the case on both ST and 800 XL, but the configuration of power factors (for losses not retreat) aren't the same:

Code: Select all
             England -> Japan
army         1000        50
power factor    1         2
retreat       500        25 (fortified)
random          1         2
losses (ST) 1 every 1  1 every 4
losses (XL) 1 every 4  1 every 1


Don't know if it's more intelligible...
Pascal

PostPosted: 29 Mar 2008, 09:02
by Monty
hmm, i'm irritated that you still insist on "needed losses to retreat" = 25 men. wasn't the formula to calculate this for the defender given as:
(size of defending army)/2^(defensive factor-fortification factor)
= 50/2^(3-1) = 50/4 = 12,5 men ?

though, with random factor 2 for the defender the JAPAN-"kills per round-ratio" would still be 1 every 4:

1/(defensive factor-fortification factor+random factor) = 1/(3-1+2) = 1/4

JAPAN, using this on themselves (ST), would kill 12,5 men of theirs in the first round coincidentally (due to the random factor).

with random factors 0 and 1 they would kill even more of their own men (25 and 17) and reach the "needed losses to retreat" in the first round, too.

only with random factor 3 they would kill a smaller amount of men (= 10) and could enter round 2...

where am i (still) wrong? sorry for being stupid...:oops:

(btw.: isn't the attackers "kills per round-ratio" calculated as
1/(offensive factor+terrain factor+random factor)? so it could never reach 1 every 1 when random factor = 1...)

PostPosted: 29 Mar 2008, 16:55
by Kroah
Monty wrote:hmm, i'm irritated that you still insist on "needed losses to retreat" = 25 men. wasn't the formula to calculate this for the defender given as:
(size of defending army)/2^(defensive factor-fortification factor)
= 50/2^(3-1) = 50/4 = 12,5 men ?

You're right, i've swapped the Japan offensive and defensive factor. So for the example, it's like if it was the USA (defensive factor=2, retreat=50/2^(2-1) = 25).
Monty wrote:though, with random factor 2 for the defender the JAPAN-"kills per round-ratio" would still be 1 every 4:

1/(defensive factor-fortification factor+random factor) = 1/(3-1+2) = 1/4

Not exactly. Like i said you, i've simplified some parts. In fact on ST, the fortification is only used to double the "needed losses to retreat". It's not used to compute the "X every Y" (1/(2+2)). On XL it does (1/(2-1+2)).

Monty wrote:JAPAN, using this on themselves (ST), would kill 12,5 men of theirs in the first round coincidentally (due to the random factor).

with random factors 0 and 1 they would kill even more of their own men (25 and 17) and reach the "needed losses to retreat" in the first round, too.

only with random factor 3 they would kill a smaller amount of men (= 10) and could enter round 2...

In the case of USA (my previous mistake):
- random factor 0 or 1: battle lasts 1 round
- random factor 2 (12,5 men): haven't tested, but i think battle lasts 9 rounds)
- random factor 3: battle lasts 9 rounds

In this example, 50% of the battles (random factor 2 and 3) are like i say "bugged".

Monty wrote:(btw.: isn't the attackers "kills per round-ratio" calculated as
1/(offensive factor+terrain factor+random factor)? so it could never reach 1 every 1 when random factor = 1...)

The ratio is calculated like you said, but the result could be "1 every 1" because terrain value may be 0 as the random factor.

Cheers,
Pascal

PostPosted: 29 Mar 2008, 17:24
by Kroah
Kroah wrote:
Monty wrote:(btw.: isn't the attackers "kills per round-ratio" calculated as
1/(offensive factor+terrain factor+random factor)? so it could never reach 1 every 1 when random factor = 1...)

The ratio is calculated like you said, but the result could be "1 every 1" because terrain value may be 0 as the random factor.

I must explain another thing here:

- on ST, the terrain is added to the defensive factor because the army attacks his self (bonus for defender because the defender kills less his self).

- on XL, the terrain is added to the offensive factor because the army attacks the opponent (bonus for defender because the attacker kills less the defender).

This is why i think the new behavior on ST is intentional (but bugged?). The developper has changed some combat rules. If the terrain is now added to the defensive factor instead of offensive, this is certainly because they know the army attacks his self. The result is the same: the terrain helps the defender.

But why on earth did they change the way kills are done (to the opponent -> to his self) ?!?

Pascal

PostPosted: 29 Mar 2008, 17:44
by Kroah
Kroah wrote:On ST, (B) is:
* if all men of >one< (not both) side have attacked, the army move will be resolved again for another battle at the next round with the remaining men.


One last major thing i forgot to say:
- on ST at round 9, the battle doesn't stop if one side has all his men attacked.

So if you attacks with a biiig army, it's here at round 9 you army is slew... by your own army! The battle doesn't stop when the 50 defender men have attacked. And generally, you kills yourself up to the "needed men to retreat", with a bunch of losses.

Does this make it clear?

PostPosted: 29 Mar 2008, 19:58
by Monty
omg, lots of unexpected information :shock:
as often, nothing is as it seems...:D

but great that you took the time to give examples (in the st-bug thread, too), because i'm of the "understanding-best-from-examples" type...:wink:

and just to make it completely clear: coconet is at the moment running the exact XL-rules, isn't it?

cheers monty (now resetting damaged brain and trying to boot with updated infos)

PostPosted: 29 Mar 2008, 20:08
by Kroah
Monty wrote:omg, lots of unexpected information :shock:

Yes, it's always difficult to give enough informations without overwhelming the reader by too much informations. Without talking about fundamental informations i forgot to say :wink: That's why i really like when you insist when something seems strange.

Monty wrote:and just to make it completely clear: coconet is at the moment running the exact XL-rules, isn't it?

Exactly. Actually i prefer to keep XL rules until we find a way to avoid the strange behavior (bug) of the ST.

Monty wrote:now resetting damaged brain and trying to boot with updated infos

lol, good luck!

Pascal