Battles & Numerical Advantage

Tarafından yazıldı Laveley, 07.03.2015 at 11:37

There are various situations were split your stack or keep it together mays be advantageous. But thats not the point. The point is that knowing that your troops will fight potentially better when in numerical advantage is crucial to make the right decision between this two options.

Lavely, can you delete your topic... people don't seem to realize the original point you brought up has now been addressed, and the last 20 posts seem to represent an escalating flame war... it was a good topic. Give it a good death now that it has fulfilled it's purpose

Sure, error margins means nothing. I've lost all my hopes on Ireland education system.

Laverley, you can't absolutery prove the influence of the size of battle, but the Following test would reveal its existance:

First do 20 battles. 10 of them will be 8 tanks vs 8 infantries, rest will be 16 tanks vs 8 infantries.

Watch the battle full and take note of all the rolls there. Do this for the 20 battles. At the point you finish watching I expect you to have around 90-150 rolls from one side and more than 300 in total from both sides. Around 600 if we combine both tests.

So, divide the rolls by frecuency. ¿How many times did the tank rolled a certain value? ¿How many times did the infantry rolled a certain value?


Make Four graphics. Two will show tanks and infantries frecuency on 8 tanks vs 8 infantries. Rest two will be the same but for 16 tanks vs 8 infantries.

I don't know the fórmula for calculate the error involved in one test, but let's say that any value shouldn't bypass the other by more than 10 percent the number of rolls made for that certain unit in total ( if your tanks rolled 125 rolls in total, then any number from 1 to 8 shouldn't be higher than other value number by 12.5).

If you can effectiverly prove that certain values in the 16x8 test does bypass by more than 12.5 its values in the 8x8 test. Then you'll be proving the existance of such size effect.

Tarafından yazıldı clovis1122, 16.03.2015 at 15:07

Sure, error margins means nothing.

70 units with 1/1 vs 10 units with 1/1. no critical, no strategy, no city bonus.




Now 60 vs 20, same unit:



Edited: Same units, but now with 5 critical:

only 1 out of the 7 battles had a extra -1



Same units, but with 7 critical, now 3 out of 7 had extra damage





Here is the mind fuck:

Tunder's caclulation proves the stat of the unit change the dynamic of battle more than a critical chance.

Tarafından yazıldı Tundy, 16.03.2015 at 22:03

Tarafından yazıldı clovis1122, 16.03.2015 at 15:07

Sure, error margins means nothing.

Ironically you quoted this part... XD.

Listen, what I am asking for is the rolls of the test, not the Result of the battles.

Edit: nvm yeah as critical says. Stats does influences. But... We are talking about size of troops here? XD.

I am aware that if one unit have more attack or critical or HP, it will perform better than the other unit in different sizes. Example one tank against one militia, but since the tank have more attack and critical, then in 20 tanks vs 20 militias the tanks will perform better. This effect is made cause the tanks have more opportunities to roll, which means more chances of getting higher roll than militia, plus critical hit...

But what are we testing here is how different are the rolls from a 8 tanks vs 8 infantries situation, and 16 tanks vs the same 8 infantries. ¿Shouldn't they perform equal? Laverley and Laochra says it does not, and for some unknow reason the tanks get better rolls while larger the size. I say it does not.

The only exeptions I've found is the one where the 8 tanks dies first than the infantries, making the battle to go a bit further which would benefit the 16 tanks for size effect. But if we assume the 8 tanks would die, then ¿Shouldn't said effect be bigger as the number of troops grows? If I send 100 tanks against 8 infantries, According to them, I would get better rolls than with 16 tanks.

Btw wort testing. Compare the rolls from 30 tanks vs 1 infantries with the rolls of 16 tanks vs 1 infantries. It will be shorter than the last test purpose.

If in 10 battles, the rolls from the 30 tanks vs 1 infantry are way better than the rolls from the 16 tanks vs 1 infantry, then the overstack does matter. Take a 10 percent from the total rolls you've got as error margin.

Hi.

I though this discussion was over and just now saw the other posts after my last reply.

In fact, i did the 20 battles experiment 16tanksx8inf and 20 in 8x8 and took note of all the rolls. Unfortunately, i did it like 2 weeks ago in that weekend i said i would do it... and now that i'm searching the data i cant find it x_x

I may have deleted it.

Anyways, the results just confirmed laochra theory (the guy that first replied, he is laochra right?), the difference is given in the crits. The infs seems to crit less in the 16x8 than they do in the 8x8 scenario, thus the impossibility of destroying 8 tanks like frequently happens in 8x8.

Now, i dont want to "prove" anything here. I dont need to "prove" the earth is round when i say it to someone, even though i havent the ways to prove it. If i say to someone that the earth is round and he replies it is flat and refuses to accept my point until i "prove" it, i will just give my back and walk away. I dont want to change nobodies opinion here. My only interest is in myself and the info i want and i'm pretty convinced by my tests that there is indeed a numerical advantage in battles (due specially to the quantity of crits)... if somebody disagrees or think i'm lying, just to the tests by yourself and take your own conclusions.... this whole discussion is just a waste of calories imo.

Btw, it doesnt take that long as i first predicted... you can do the same amount of tests in just one afternoon if you do it in a practical manner, and my sample was way over 1k cause battles even in 8x8 frequently have more than 10 rounds (and its 2 rolls per round)...

so yeah, my doubt is pretty much remedied. And i will continue to search the data of the tests in the myriad of junk of my computer, if i find it, i come here to post it.

Tarafından yazıldı Laveley, 18.03.2015 at 18:02

Now, i dont want to "prove" anything here. I dont need to "prove" the earth is round when i say it to someone, even though i havent the ways to prove it. If i say to someone that the earth is round and he replies it is flat and refuses to accept my point until i "prove" it, i will just give my back and walk away. I dont want to change nobodies opinion here. My only interest is in myself and the info i want and i'm pretty convinced by my tests that there is indeed a numerical advantage in battles (due specially to the quantity of crits)... if somebody disagrees or think i'm lying, just to the tests by yourself and take your own conclusions.... this whole discussion is just a waste of calories imo.

I think you'be misunderstand my position here.

For give an example, I do personally believe that there's something wrong when applying defensive bonus(trus, explaining infantries's weakness against offensive units) , but I cannot just go to one thread saying that it is. There is a line between what you think and what can be Rigorously proved, and you shouldn't try to affirm anything that can't be proved. Hence why I said: "Use the Word IMO when talking about lies or legends". It means "In My Opinion".

Basically, you aren't allowed to teach the world that the earth is round till you've proved it. But you do can say, in all your rights that "I think the world is round because.... ".

GL finding the tests.

BUMP. This is actually enough for me.

Tarafından yazıldı Tundy, 16.03.2015 at 22:03

You should do this test with real units. HP and CRIT have a big role in battles; plus I'm afraid there could be a bug somewhere.

Tarafından yazıldı clovis1122, 05.03.2015 at 17:21

You should use the word "IMO" when talking about lies or legends

Tarafından yazıldı clovis1122, 14.03.2015 at 18:01

my calculator

Tarafından yazıldı clovis1122, 05.03.2015 at 17:21

lies or legends

Tarafından yazıldı clovis1122, 05.03.2015 at 17:21

LIES OR LEGENDS

There is the official response.

/thread

This thread is great

Sorry to make this thread active again, but I have few elements to bring.

Tarafından yazıldı Laveley, 05.03.2015 at 10:12

Simple question: does numerical advantage matters in battle?

yes

Tarafından yazıldı Laveley, 05.03.2015 at 18:26

Results for 16x8:
Number of battles that tanks had:

ONLY 2 losses = 4 battles
3 losses = 22 battles
4 losses = 30 battles
5 losses = 27 battles
6 losses = 13 battles
7 losses = 4 battles
8 or more losses = NONE

Mean of losses in 100 battles = 4.35

Now, Results for 8x8 battles
Number of battles that tanks had

ONLY 2 losses = NONE, ZERO, NADA
3 losses = 6
4 losses = 26
5 losses = 26
6 losses = 18 battles
7 losses = 14 battles
8 or more losses = 10 battles, in other words, in 10 out of 100 battles that 8 tanks attacked the 8 infantries the tanks actually lost the battle.

Mean of losses in 100 battles = 5.38

As we can see here, the more tanks you send the less tanks you loose.

Tarafından yazıldı Laveley, 05.03.2015 at 18:26

Its clear from this results that the infantries performed better in the 8x8 scenario (destroyed more tanks) than they did in the 16x8.

No they didn't. They did the same performance.
The only difference is that you sent less tanks. So it takes few more rolls for your tanks to kill the 8 inf.
During these few more rolls, the 8 inf killed 1 or 2 more tanks.

Tarafından yazıldı Laveley, 05.03.2015 at 18:26

Why this happens? I don't know, what i know is: it seems to have nothing to do with criticals... if you put the battle log in very slow and observe the battles, you will see that the rolls for the units that are in numerical disadvantage seem to be lower and the ones with advantage larger in comparison with rolls in a balanced battle.

As explained above, nothing to do with critical hits, only the rolling system.

Tarafından yazıldı Laveley, 05.03.2015 at 18:26

This can happen due to various reasons: bad programming, bug, or simply a hidden factor that gives a bonus to the units that have numerical advantage and thus roll better numbers.

No bugs, black magic or other jewishery there, only maths. Not even maths, logic.

More attacking units -> more attack -> more damage to defense -> less defense on next roll -> less damage to attacking units on next roll
-> more remaining attacking units at the end of the battle

Tarafından yazıldı Htin, 16.03.2015 at 22:26

Tunder's caclulation proves the stat of the unit change the dynamic of battle more than a critical chance.

Sure, and what Laveley calculation shows is that number of units change the dynamic of the battle more than stat of the unit