|
|
Post by saltin on Sept 13, 2015 4:56:17 GMT
I have reached the -preliminary- conclusion that all critical skills (the ones that have description of "fatal blow") are lame ducks, basically worthless skills.
More tests need to be done but from what I can see the skill doesn't trigger that often and most importantly is always really low.
The skill seem to do slightly better on very strong generals but is still hardly noticeable.The skill additional damage might as well be invisible on lower ranking generals.
Does anyone seen this? If you want to look at this make sure that other skills are not interfering with your observations (like anti armor skill ect..)
|
|
kalima
First Lieutenant
Posts: 19 
|
Post by kalima on Sept 13, 2015 16:17:02 GMT
I have infantry leader on an eyeless general but has made little to no difference might just make the smallest difference to it.
|
|
|
|
Post by kanue on Sept 13, 2015 16:40:24 GMT
From my experience, I see the skill triggers more often than I expect (I have a low hope on 16% chance) yet the damages are barely noticeable. I, personally, do not like the skill much but all good tank generals have it so I just think about it as better than nothing.
|
|
|
|
Post by curiosity on Sept 13, 2015 17:23:19 GMT
I find them not noticable too. I find the skills that have your enemies not retaliate sooo much better.
|
|
|
|
Post by Eastern empire on Sept 13, 2015 22:45:30 GMT
Critical skills though as broken.
Trigger chance percentage following advertised law in harmony,that part working as should be,easy to test this.
Broken part is damage range when crit happens,quantity of hurts very low.If test and gather data and then plot numbers into excel to get estimated formulation one can see it follows this:
Crit damage = base damage mean x (1+ Gen. stars / 4)
For generals with high star number the damage will be a very small add over the mean (not even the upper range).For generals with low stars the skill is just not observing.
What meaning for this? average damage increase from crit often just 1 or 2 points extra from normal damage.
Guessing Crit skill just not working and should be left as only last choosing.
|
|
|
|
Post by kanue on Sept 14, 2015 3:56:33 GMT
For a 5 stars gen.
Crit damage = base damage mean x (1+ 5/4)
= base damage mean x (1 + 1.25)
= base damage mean x (2.25)
If 2.25x damage is really the case then it is not bad.
|
|
|
|
Post by saltin on Sept 14, 2015 8:29:16 GMT
K I did a quick test to have a few actual numbers to look at. Test conditions 1939 German conquest, reload 1st turn each time after getting a single sample attack value. Test subjects: Guderian German general (npc general not player owned) on a heavy tank against a regular Polish car.  Subjects background:
Heavy tank raw attack stat as can be seen is 30-46 , because it is commanded by Guderian who has 5 stars in the tank stat the heavy tank will do a pre-modification damage of an adfitional 30 points (6 extra damage per star,6X5=30 ). So Gudarian on a heavy tank raw damage now becomes 60-76,before terrain and technology modifiers, for simplicity we will ignore the midifiers for now.  Gudarian crit skill is at level 4 and so should trigger at a rate of 16% or about one and a half times per 10 attacks.  None of Guderian other skills should trigger as his health will not dip below 50%,and he will not benefit from plain fighting. Panzer leader (the crit skill) always pops an icon when it triggers,so we will note when that happens and its corresponding value. I have gathered a sample of 21 data point to give us a rough idea,here are the attack values:
Data points: (bold yellow are crits) 66,72,58,68,68, 74,58,66,61,62,69,66,73,58,68,67,64, 74,63,70,69. Ok so I am sure there is a better way of doing this by using a more scientific method and what not,mean and standard deviation but instead I am gonna keep it simple and just plug common sense values. -Total data points: 21 -Total data points without crits:19 -Crit occurance: 2 times out of 21 -Crit average value:74 -Average damage with no crits:1320/19= 69.47-Average damage with the crits:(1320+74+74)/21= 69.90Meaning overall the critical skill contribution to our average damage is that it increased it by less than 1 point. Also in EasternEmpire post I think the base mean damage is taken from the unit? In this case it's the average damage of the tank . So in the other formula it becomes: (30+46)/2=38 and then crit damage=38x(1+1.25)=85.5 before modifier . PS:does the terrain take 10% of that value too so it goes from 85-8.5=76.5 which gets rounded to 76? A number really close to our tested value of 74. |
|
|
|
Post by curiosity on Sept 14, 2015 10:35:48 GMT
K I did a quick test to have a few actual numbers to look at. Test conditions 1939 German conquest, reload 1st turn each time after getting a single sample attack value. Test subjects: Guderian German general (npc general not player owned) on a heavy tank against a regular Polish car. Subjects background:
Heavy tank raw attack stat as can be seen is 30-46 , because it is commanded by Guderian who has 5 stars in the tank stat the heavy tank will do a pre-modification damage of an adfitional 30 points (6 extra damage per star,6X5=30 ). So Gudarian on a heavy tank raw damage now becomes 60-76,before terrain and technology modifiers, for simplicity we will ignore the midifiers for now. Gudarian crit skill is at level 4 and so should trigger at a rate of 16% or about one and a half times per 10 attacks. None of Guderian other skills should trigger as his health will not dip below 50%,and he will not benefit from plain fighting. Panzer leader (the crit skill) always pops an icon when it triggers,so we will note when that happens and its corresponding value. I have gathered a sample of 21 data point to give us a rough idea,here are the attack values:
Data points: (bold yellow are crits) 66,72,58,68,68, 74,58,66,61,62,69,66,73,58,68,67,64, 74,63,70,69. Ok so I am sure there is a better way of doing this by using a more scientific method and what not,mean and standard deviation but instead I am gonna keep it simple and just plug common sense values. -Total data points: 21 -Total data points without crits:19 -Crit occurance: 2 times out of 21 -Crit average value:74 -Average damage with no crits:1320/19= 69.47-Average damage with the crits:(1320+74+74)/21= 69.90Meaning overall the critical skill contribution to our average damage is that it increased it by less than 1 point. Also in EasternEmpire post I think the base mean damage is taken from the unit? In this case it's the average damage of the tank . So in the other formula it becomes: (30+46)/2=38 and then crit damage=38x(1+1.25)=85.5 before modifier . PS:does the terrain take 10% of that value too so it goes from 85-8.5=76.5 which gets rounded to 76? A number really close to our tested value of 74. Wow, you did you math homework. |
|
|
|
Post by Erich von Manstein on Oct 4, 2015 18:20:42 GMT
The maximum attack of the heavy tank is 46, Guderian's star gives +28(my thread says that). 46+28=74. Correct.
So the trigger means the maximum attack of the unit+general star.
|
|
|
|
Post by saltin on Oct 4, 2015 22:31:48 GMT
The maximum attack of the heavy tank is 46, Guderian's star gives +28(my thread says that). 46+28=74. Correct. So the trigger means the maximum attack of the unit+general star. I think more needs to be done to understand this. Your formula for crit damage is simply the max damage from the tank added to the normal damage that the general has. Which would be strange because it would mean crit damage=normal max damage That isn't the normal definition of crit damage in most games. In any case it doesnt really matter which formula we use,the data clearly shows the skill to be weak. |
|
|
|
Post by Erich von Manstein on Oct 5, 2015 0:44:14 GMT
The maximum attack of the heavy tank is 46, Guderian's star gives +28(my thread says that). 46+28=74. Correct. So the trigger means the maximum attack of the unit+general star. I think more needs to be done to understand this. Your formula for crit damage is simply the max damage from the tank added to the normal damage that the general has. Which would be strange because it would mean crit damage=normal max damage That isn't the normal definition of crit damage in most games. In any case it doesnt really matter which formula we use,the data clearly shows the skill to be weak. But it seems like it is correct...... Simple but the result fits. But true, xx Leader is a pretty weak skill. |
|
|
|
Post by The Armchair General on Oct 5, 2015 2:37:36 GMT
Well it might be why the crit damage skills are broken in the first.
If the crits cannot get higher than the normal max damage then they are busted and not working.
|
|
|
|
Post by kanue on Oct 5, 2015 3:42:45 GMT
If crit damage=normal max damage then all crit must do the same damage given other factors are controlled. I don't think that is the case, though.
Also,I think we can't make any inference from just 2 observations of crit (from above) because the sample size is way too small. More test is needed.
|
|
|
|
Post by Suvorov on Oct 6, 2015 15:18:59 GMT
Perhaps someone with a rooted phone should change the stats of a tank to a number (let's say 10-10) then put a general with 1 star in tank on it, change the likeliness of the trigger of the critical damage skill to 100%, subtract the answer by 6 and look at the outcome.
So for example:
Stats of tank: 10-10
Likeliness of triggering crit damage skill: 100%
Number of stars in tank ability the general has: 1
Outcome: 19
19-6 (1 star in tank)=13
13- (stats of tank) 10=3
So, the critical damage is 3.
|
|
|
|
Post by Moreau on Oct 6, 2015 19:11:22 GMT
Sometimes it happens for me and it usally saves me through a misson
|
|
