Chimpyang
03-22-2009, 00:10
Ok,
As wel all know, ETW has broken with past traditions on the effects of exp. and the important aspects. I believe that before we can make a mod that plays well in the battle area, we need to test and understand the mechanics of CA's game. The reason I have had a play tonight with it is that I was curious as to the effects of veterancy, something nobody has come up with an answer with yet. I will update this page as I do more testing, but here are initial resul
So....to be rigorous there are 2 main types of scenario for initial 1 unit vs i unit testing.
Unit : United Provinces Line Infantry (Large Unit size)
Map : Ceylon - it is one moderately gradiented hill where team 1 is above team 2. Weather is dry, middday and summer, no time limit.
Difficulty : Medium - once we establish the "comparative set", we can move onto effects of difficulty levels
Era : Early (does this make a difference, we need something to set a benchmark)
Procedural note : For every test I did not modify the formation, that is not the subject of the initial tests, experience shows that the AI will default to the triple rank (40 wide) when it decides to deploy in such a manner. Also, at the start of each battle, I clikced deploy and immediately paused the game (hotkey "p"), I order my men to advance and fire upon the enemy unit. No other orders are given afterwards
Variables : I have tentatively tested the following : Veterancy (effects of, AI's response to), Uphill/Downhill?
Test 1 : Veterancy.
1.1
I set my units onto max veterancy (9) and AI onto 0 - rationale is that if effects are subtle - we can see them best using the largest disparity offered. From deploying as team 1 (therefore approaching downhill) the results are:
1.1.1. AI marched towards me, firing match, enemy general not killed during firing, enemy broke at 20/120, my unit was at 56/120
1.1.2. Firing match again, general not killed, enemy broke at 22/120, friendly at 57/120.
It would seem that AI will only fire at you, will not charge no matter what - I was winning the firefight from an early stage. AI unit gained 1 exp each time during the firefight. Did that help them stay a bit longer?
1.2
I reverse the team so I am attacking uphill, enemy deploys in same formation. Same as test 1.1.
1.2.1 AI stopped momentarily as he entered my firing zone - bayonet charged me. Numbers were reduced when lines met but numbers unrecorded. Eventual outcome was 34/120 when the AI broke and 102/120 on my unit. General was killed almost instantly when his line hit mine.
1.2.2 AI again charged me - Same effect as last time, but enemy broke at 55/120, I had 107/120
1.2.3 AI changed tack and deployed in the same way but decided to fire upon me. Firefight lasted until AI had 78/120 and I had 99/120. Bayonet charge at that point - AI lost and broke at 45/120 but the general was not killed. I had 89/120 left.
The above seem to suggest that units with veterancy may have an advantage in H2H combat - even when facing a charge bonus? However, units recieveing a charge are often able to squeeze off another volley - aiding the defending process. AI seemed to only firefight when attacking uphill (however more tests are needed - 2 is not enough to rule out variables) whilst more options are available when attacking downhill. Unknown behaviour when flat surface. Whilst firefighting it would appear that my unit was more accurate than the enemy (however slightly) but is this the effect of the hill or veterancy? Only 1 test with me approaching uphill has any firefight - but when the AI charged, I was winning the firefight. Suggests veterancy helps but next test will confirm.
Test 2 - Testing the Hill effect
In previous TW games, the hill or high ground gave an advantage of greater range and better att and def bonuses. From test 1 it was observed that both lines always fired roughly at the same time, with sometimes even the lower line firing first so range is obviously not given a bonus in real gameplay terms. The ETW ai appears to wait until all of the enemy unit is in range before firing (again conclusion limited to infantry).
Test 2 was the same as test 1, just with no difference in veterancy - both sides had veterancy 0.
2.1 Going downhill at the enemy
2.1.1 To my surprise, AI deployed in a 10 wide 12 deep column formation - am I expecting a charge here? But as lines approached, it fairly quickly changed to the same line as me. Firefight on hill began in usualy place - however AI charged when it had 87/120 compared to 85/120. During the H2H combat my unit gained 1 exp and defeated the AI, wit hthe AI breaking at 41/120 and my unit at 69/120. Neither general died.
2.1.2 Same again but 72/120 for AI and 84/120 for me when AI charged. H2H went miserably for the AI again with it breaking at 37/120 leaving me with 73/120. However, it my unit for some reason volleyed first in the firefight - which may have given me an advantage Generals not killed
From the first 2 - it would appear that the hill confers only a melee bonus - firefights between identical infantry produce roughly similar casualties - will need further testing to confirm.
2.1.3 This time, to test out the hunch of whether the AI would be different if I were to approach uphill. Still deployed in column but changed to line again. Firefight was quite even - 85/120 when it charged my 95/120 unit - again my unit had first shot by luck it seems. Enemy broke at 26/120 and I had 60/120. General not killed
----------------------------------------------------------------------------------------------
From these very early tests I will go on to testing the following hunches later:
1. Hill will confer melee bonus only - more repeats with identical units to confirm - esp if AI behaviour is subject to be different to the one currently observed
2. Will the AI do different things if the units are equal but of max veterancy? easy to test - ratchet the exp up and repeat. This will be a good way to rid of the variable of increasing exp during battle.
3. Veterancy confers both a bonus in melee and shooting - compounded with "hill effect" on melee only in some of the test so far? Definitely more accurate - although more repeats to establish statistical truth and make sure the results are repeatable would be good.
4. Now we have a theory on the effects of veterancy, what about differing levels of veterancy? What would be the effect of a different of 1 level, also is tha assumption that veterancy here is linear correct?
Anything from the data that I have missed for the next round of tests? If you making suggestions, give a full logical reasoning otherwise I will probs ignore it. I need to be sure that we are not wasting time when there are holes in the testing plan. ALso if you spot any flaws in logic - please point out and hopefully I will be able to remedy it.
As wel all know, ETW has broken with past traditions on the effects of exp. and the important aspects. I believe that before we can make a mod that plays well in the battle area, we need to test and understand the mechanics of CA's game. The reason I have had a play tonight with it is that I was curious as to the effects of veterancy, something nobody has come up with an answer with yet. I will update this page as I do more testing, but here are initial resul
So....to be rigorous there are 2 main types of scenario for initial 1 unit vs i unit testing.
Unit : United Provinces Line Infantry (Large Unit size)
Map : Ceylon - it is one moderately gradiented hill where team 1 is above team 2. Weather is dry, middday and summer, no time limit.
Difficulty : Medium - once we establish the "comparative set", we can move onto effects of difficulty levels
Era : Early (does this make a difference, we need something to set a benchmark)
Procedural note : For every test I did not modify the formation, that is not the subject of the initial tests, experience shows that the AI will default to the triple rank (40 wide) when it decides to deploy in such a manner. Also, at the start of each battle, I clikced deploy and immediately paused the game (hotkey "p"), I order my men to advance and fire upon the enemy unit. No other orders are given afterwards
Variables : I have tentatively tested the following : Veterancy (effects of, AI's response to), Uphill/Downhill?
Test 1 : Veterancy.
1.1
I set my units onto max veterancy (9) and AI onto 0 - rationale is that if effects are subtle - we can see them best using the largest disparity offered. From deploying as team 1 (therefore approaching downhill) the results are:
1.1.1. AI marched towards me, firing match, enemy general not killed during firing, enemy broke at 20/120, my unit was at 56/120
1.1.2. Firing match again, general not killed, enemy broke at 22/120, friendly at 57/120.
It would seem that AI will only fire at you, will not charge no matter what - I was winning the firefight from an early stage. AI unit gained 1 exp each time during the firefight. Did that help them stay a bit longer?
1.2
I reverse the team so I am attacking uphill, enemy deploys in same formation. Same as test 1.1.
1.2.1 AI stopped momentarily as he entered my firing zone - bayonet charged me. Numbers were reduced when lines met but numbers unrecorded. Eventual outcome was 34/120 when the AI broke and 102/120 on my unit. General was killed almost instantly when his line hit mine.
1.2.2 AI again charged me - Same effect as last time, but enemy broke at 55/120, I had 107/120
1.2.3 AI changed tack and deployed in the same way but decided to fire upon me. Firefight lasted until AI had 78/120 and I had 99/120. Bayonet charge at that point - AI lost and broke at 45/120 but the general was not killed. I had 89/120 left.
The above seem to suggest that units with veterancy may have an advantage in H2H combat - even when facing a charge bonus? However, units recieveing a charge are often able to squeeze off another volley - aiding the defending process. AI seemed to only firefight when attacking uphill (however more tests are needed - 2 is not enough to rule out variables) whilst more options are available when attacking downhill. Unknown behaviour when flat surface. Whilst firefighting it would appear that my unit was more accurate than the enemy (however slightly) but is this the effect of the hill or veterancy? Only 1 test with me approaching uphill has any firefight - but when the AI charged, I was winning the firefight. Suggests veterancy helps but next test will confirm.
Test 2 - Testing the Hill effect
In previous TW games, the hill or high ground gave an advantage of greater range and better att and def bonuses. From test 1 it was observed that both lines always fired roughly at the same time, with sometimes even the lower line firing first so range is obviously not given a bonus in real gameplay terms. The ETW ai appears to wait until all of the enemy unit is in range before firing (again conclusion limited to infantry).
Test 2 was the same as test 1, just with no difference in veterancy - both sides had veterancy 0.
2.1 Going downhill at the enemy
2.1.1 To my surprise, AI deployed in a 10 wide 12 deep column formation - am I expecting a charge here? But as lines approached, it fairly quickly changed to the same line as me. Firefight on hill began in usualy place - however AI charged when it had 87/120 compared to 85/120. During the H2H combat my unit gained 1 exp and defeated the AI, wit hthe AI breaking at 41/120 and my unit at 69/120. Neither general died.
2.1.2 Same again but 72/120 for AI and 84/120 for me when AI charged. H2H went miserably for the AI again with it breaking at 37/120 leaving me with 73/120. However, it my unit for some reason volleyed first in the firefight - which may have given me an advantage Generals not killed
From the first 2 - it would appear that the hill confers only a melee bonus - firefights between identical infantry produce roughly similar casualties - will need further testing to confirm.
2.1.3 This time, to test out the hunch of whether the AI would be different if I were to approach uphill. Still deployed in column but changed to line again. Firefight was quite even - 85/120 when it charged my 95/120 unit - again my unit had first shot by luck it seems. Enemy broke at 26/120 and I had 60/120. General not killed
----------------------------------------------------------------------------------------------
From these very early tests I will go on to testing the following hunches later:
1. Hill will confer melee bonus only - more repeats with identical units to confirm - esp if AI behaviour is subject to be different to the one currently observed
2. Will the AI do different things if the units are equal but of max veterancy? easy to test - ratchet the exp up and repeat. This will be a good way to rid of the variable of increasing exp during battle.
3. Veterancy confers both a bonus in melee and shooting - compounded with "hill effect" on melee only in some of the test so far? Definitely more accurate - although more repeats to establish statistical truth and make sure the results are repeatable would be good.
4. Now we have a theory on the effects of veterancy, what about differing levels of veterancy? What would be the effect of a different of 1 level, also is tha assumption that veterancy here is linear correct?
Anything from the data that I have missed for the next round of tests? If you making suggestions, give a full logical reasoning otherwise I will probs ignore it. I need to be sure that we are not wasting time when there are holes in the testing plan. ALso if you spot any flaws in logic - please point out and hopefully I will be able to remedy it.