Info: Combat (GL)

[Xin Yu]

bump

[SlowThinker]

post 122

Another possible info to be integrated:
Surprise attack bonus
Veteran status of Missiles(?)
City Battle Healing

[Marquis de Sodaq]

Thanks for the links, ST.
Thanks rah for the sneak attack tests, as well.


I'll get around to updating the main text someday soon.

[Thoddy]

In wich way are the effects of terrain and veteran status are added to the outcome of an combat.

I made a little experiment:

both units has 4a,1d,1hp,1fp

; defense = multiply by 50% to get % of normal combat factor defended at.
; (so 3 x 50% = 150%, meaning a strength 2 unit defends at 3).

I create a 400% terrain
Desert, 1,8, 0,1,0, yes, 1, 6, 5, no, 1, 5, 3, no, ; Drt

I expected no differences between the outcomes of a non veteran and a veteran combat, but

1st test
Without the veteran bonus the outcome was a little better than 1:1 and the winning unit was mostly damaged.

2nd test
If I made both units a veteran the winning ratio for the attacker has been significantly increased and if the attacker wins, the damage has bee decreased in comparision with the results from the first test.

Solutions

[SlowThinker]

The defender has an advantage: both players throw a dice but a defender wins draw results.
If attacker=1, defender=1, then the real values are attacker=7/8 and defender=9/8.
Read DaveV's and my posts in Modifiers for Attack/Defense

In your example non-vets are 3 7/8 vs. 4 1/8, vets are 5 7/8 vs. 6 1/8.

[Boco]

Thoddy,

What were your results? I think I may need to modify my calculator spreadsheet - it was predicting virtually the same result for both situations (48 v. 49% chance of victory).

[Shaka Naldur]

very interesting

why are there so many deleted posts?
is this thread already in the great library?

[SlowThinker]

Shaka Naldur,
Posts and suggestions that were included into the main text are deleted so that don't have to read 9 pages of text.

Thoddy,
try to increase hp of both units from 1 to 14 (IIRC this is the maximum). Results should be more stable and should correspond to the Boco's calculator.

[Thoddy]

@Boco
I run about 50 combats and the vet combat results were not the expected ones. So I ask for possible reasons.

Originally posted by SlowThinker
Thoddy,
try to increase hp of both units from 1 to 14 (IIRC this is the maximum). Results should be more stable and should correspond to the Boco's calculator.

I will test with more hp but this will last some time because of 3 week holidays.

[SlowThinker]

time of healing

I have tested the defense bonus against barbs in cities. See Strategies against Barbarian Attacks .

Originally posted by Marquis
Healing away from a city or fortress (by skipping that unit's turn) occurs immediately, before the next unit is available for an order. Barracks healing occurs at the end of the player's turn.

After testing I claim healing is always executed when your cities are processed by the Civ2 engine.
An example: You are attacked by AI, you see your unit got in red. Then cities are processed and if your damaged unit didn't move last turn (i.e. before the AI attack) you see your unit got in yellow.

[Thoddy]

Good news,

The calculator doesnt need any modifications.

By setting hp to 10 I got the expectedt results.
Seems the random factor at the first tests has well meant to me.

[SlowThinker]

Sneak attack bonus

Result of tests:
Sneak attack bonus works only if human player is attacked (no matter if by another human or AI).
Edit: It don't happen in Multiplayer/Hotseat mode because Civ2 engine considers other human players as AIs.
The effect: Attack factor is doubled (or Defense factor is halved, but I think it is the Attack factor).

[rah]

Thanks for the clarification on that.

RAH

[SlowThinker]

Now I agree that it works in SP too - post was edited.

[SlowThinker]

choice of the defender

In River War scenario I lost a key unit (Sirdar, 4d, 2h, 2f) because it was stacked with a unit less resistible to attack (Maxim Gun, 6d,1h, 1f). The attacker was Nubian Infantry (4a, 1h, 1f): it was a Veteran (6a, 1h, 1f). Sirdar is 2x stronger against Nubian Inf. attack than Maxim Gun, but the defending unit was Maxim Gun and it lost the combat.

So I ran some tests. The result:

For each unit in the stack Civ2 engine counts 'the defensive ability':
Defense*remainingHP/totalHP
The unit with the highest 'defensive ability' defends.
If more units have an exact 'defensive ability' then the first unit that came on the square defends.

Notes:
Vet status affects the Defense (x1.5), I suppose other modifiers too but I didn't test them.
Initial HP, FP are completely ignored! Only percentage of remaining HP (remainingHP/totalHP) is taken into account.
Unit role (0-attack, 1-defense) is ignored.
Other stats (Attack, Movement) are ignored. If you have a phalanx and legion stacked and the legion stands longer on the square then legion will defend

So the choice of defender is buggy . There are 2 defects:

1. HP, FP stats are completely ignored.
2. For a combat outcome the Defense has a bigger significance than Percentage of remaining HP. Civ2 takes a healthy Warrior equal as half-wounded Phalanx but Phalanx is considerably stronger.