The loser of a battle is the first person to lose 80% of their troops in each troop category. So even the loser will keep 20% of their troops.That beginning to make some sense. Are you implying that there are some iterations to the battle? Any side starts doing damage 1st? Or is it simateneous. And then repeat.
People have been trying to figure out the algorithm for battle for a long time ... seems it is a "secret" so the developers aren't putting that info out though there has been some experiments that have noticed some "trends" during battle ... anyways as xyz said look at the details and you can see the specifics concerning the strengths of each side ...
As CIM says there is no proper explanation to this but as per my tries with my fellow players we came to know that 80% troops destruction results in loss irrespective of attacker or defender. But things which matters
1) Knowledge of both players
2) Hero equipments and Skill points of players
3) troop type of players
4) Last No. of troops players hav
BUT ABOVE ALL HEALTH OF TROOPS MATTERS MOST. If not done with good health stats then it is like sending your troops with some diseases to fight well trained army. Its bit complex to understand all this in few Days but if you try then it is possible that you may get some idea. I have seen many players breaking through 4Mil troops single handadly and have also seen Full Onslaught with 10Mil troops in it fail at hand of a huge Trapper and that to single Trapper. BUT the point is it depends upon many factors i listed above. Try to see battle report of players. Hope it helps
When two armies meet the first thing the battle system does is determine who has the bigger army. The formula for this is:
Square root of (attack * defense * health)
This also works to determine the tiers. Example 9 t1 units are the same as one t6 unit. Taking into account its comparable research.
When the system determines who has the bigger army the loser loses 80% of everything. So when the system determines that it moves on to decreasing the attack of the player who lost (overwhelming) and it can go up to 48% of the attack or so depending on how big of a difference between the size of the two armies was meaning the winner would lose less troops if his army was way bigger than the guy who lost.
Then the stats start to mater individually to determine what the winner will actually lose. You have attack and separately health and defense who work together as some sort of a health pool the formula for this is health * (defense/100). That's why you always go for health rather than defense when you have to choose. All you need to do then is divide attack of the attacker with health pool of the defender you should get the amount what Is lost on the winners side accurately.
Don't get me wrong you need all 3 stats in both defense and attacking situations cos the higher they all are the bigger chance of winning. However if you attack and you are overwhelmed and lose the only stat that maters is your attack which will take down more of your enemies troops.
Now enemy stat reducers biggest thing in the game no one really knows how they all work together cos you have a lot from research and boosts but what is known is that the ones from hero 50 percent literally half your enemies stats so it is absolute paramount to have those 3 skills on your hero maxed before everything else. After 60 percent or so they don't behave like that, they still reduce but a bit differently.
Also troop types mean nothing later on in the game but they still play a minor factor, when they mater a lot is the start.
"Also troop types mean nothing later on in the game but they still play a minor factor, when they mater a lot is the start. "
I disagree with this statement because when you have 2 players/OS where most of the stats are close to max out on both sides, then troop type match up is the big deciding factor. I have also seen many times in an OS where if you contribute the "wrong" type of troops you will take the brunt of the losses even in a big win and those T-6 are expensive to replace.
I was speaking in terms of player vs player. I can take any hit no mater what you send at me even a huge os with only troops that are good vs mine.
And yes in an OS the troops first to die are like you said, however there is always a chance your troops will get the damage but it is a bad idea to from an OS with two ppl. You need some t4 and t5 in there too at least so you avoid losing t6. If we lose t6 in an OS we consider it a bad OS. It is very expensive to lose say 800k t6 in one go. Couple of kvks ago we lost 14m troops in the initial OS consisting of t5 and t6 so I know how bad it is losing them.
Are you sure for this formula work? We did some tests in our clan, and formula didnt work for us
the formula is good, it determine the troop power... but it is not all who count in battle...
you have the health pool formula too who is important for determine the loss : life * ( defense / 100 )
along the saturation level : example with identical type of troop and same buff
troop attack 1000 defense 1000 : both loose 800
troop attack 1000 defense 1050 : attaquer loose 800 , defender loose 716
troop attack 1000 defense 1100 : attaquer loose 800 , defender loose 668
troop attack 1000 defense 1500 : attaquer loose 800 , defender loose 533
amount of troop in defense influence the loss ratio... make it difficult to crack a city with millions of t1 inside
for more detail about battle mechanic, take a look at troop secret revisited
btw, I like videos of this guy. his analytics seems really good