I've been trying over the last few weeks to break apart the medal reward system into a repeatable (if only general) imput vs result system to try to and understand what is required to succesfully get the most out of my attacks on average. Obviously there's no way to guarantee a medal from each attack but there must be a maximum average that can be achived, this is my goal. Below are my thoughts so far on what the system seems from my experiances, please read and reply with comments or any contradictions to what follows that you have observed. Links to combat reports would be most appreciated. Also please excuse the rambling I'm very bad when talking in long strings.

I'm beginning to think that the medal system is much more complicated then simply what level of valley, flat, or npc you attack as most posts I've read on these forums would seem to indicate. I'm thinking that combat rounds have a significant effect on medal chances and type of medal received. for instance, if you attack a level 1 barb town with 10 ballista and some transports combat will finish in roughly 60 rounds and you will get nada, I did this for about 4 days straight for hours a day and got nothing, waiting for 5 different barb towns around several cities of mine to regenerate their units and hitting them over and over. Later I started adding in a few men to die on the gates, started with 50 and then slowly I started sacrificing more fodder as time went on going up in increments of 50 over a week. I noticed that I started getting medals around roughly 500 to 600 and increasing dramatically the higher I pushed this until a point then the medals stopped completely. I noticed that below roughly 19 combat rounds the medal rate dried up dramatically until I was once again receiving nada after dozens and dozens of attacks. Then when I reduced the number of units and broke back into the low to mid 20 rounds range the medals once again began flowing at an expected rate of a third to a fourth of the attacks (this is average with strings of several attack producing medals and then several rounds with fruitless attacks). I then expanded into level 2 and 3 barb towns and found the exact same thing. I then went on here and looked at all the "how to medal grind" posts I could find for methods that were posted where there were follow ups where people claimed success ( looking for at least 4 to 5 post replies figuring that for every person who posted at least 3 others tried it with success and didn't). in every instance it seemed that the numbers claimed with the techs necessary/hero attack level ranged an attack rounds of the low 20's to upper 20's some just barely into the 30's on one or two I found (this seemed to result in a medal occasionally but was much more towards the fourth of attacks made range and less to the third range that seems average to me, your results may vary).

From all of this I’ve come to think of the medal system as a multi layered math system with a random component. First the level of the area attacked is taken into effect by calculating the range of medal that is possible, there seems to be a single medal that is most likely from a particular level with slightly lower chances for both the medal above and below to drop. This then sets the possibilities of an attack. Then the combat rounds are taken into effect: Too few combat rounds and the force that was defending was (obviously) mowed under by a much larger force, this obviously rewards on a VERY limited basis or not at all in extreme cases of less then 15 rounds (being as to accomplish combat in less then 15 you'd have to be able to completely demolish whatever was defending by sheer ranged dps in less then 3 to 4 rounds after you calculate the time necessary to close range and attack). When you look at the inverse you’re looking at a range function usually associated with very slow mech units in massive numbers out ranging and decimating a much smaller force and slowly creeping across the battlefield. In the range of 40 combat rounds or longer your medal chances are severely limited if not completely annulled (you won but the enemy never had a chance hence no reward for an obvious win). I'd also like to state that it seems that if the total losses for combat are high enough, then this can over-ride the combat round problem or I would say that this factor is weighted much more heavily in determining medal drop chances. Hence if combat took 60 rounds but both sides lost an exorbitantly large number of units then the fact that the combat took so long is merely a fact that there were simply that many people that needed killing. Also it is possible or seems likely that the range of combat rounds that will return a positive response is shifted up and down like a slider depending on the total number of troops assembled. In fact if I was setting up a system like this I would take total calculated forces for attackers and divide by total calculated forces for defender, if this was roughly = to 1 or 2 at the outside. Then I would then make a total function that would be directly related to the maximum number of combat rounds to work with and still achieve reward. This way you would be looking for a successful attack to end within 2/3rds to 3/4ths of the total combat rounds possible when this value is maxed and scaling downwards the fewer numbers of troops that you have occupying the field with a minimum value set at the total number of rounds it takes (on average) to close range and attack, roughly 5 to 10 rounds on the absolute minimum. You could also weight this minimum to the type of units that were on the field and link it to their relative calculated speed of movement but that's getting complicated...

So for instance lets say you have 100k archers attacking 100k archers, simply you'd divide and get (exactly) 1, this would then prompt a totaling function where 200k would be maximum (example here I’m sure that considering that defending units can be much higher then offending that this is different but for these purposes these numbers will work, also this would be vs. a NPC town and there would be a slightly altered mechanic or simply a "cap" number for maximum defenders where past a certain point the maximum number of combat rounds is allowed where there is still a chance for medal rewards). This would then function to determine just how equally well the sides were matched,

I also would like to state that I believe that the type of medals that can drop are weighted toward a calculated loss vs. kill ratio. I've noticed this trend in my combat and a lot of people seem to get certain medals more ophen when they win but bled a lot more units then their enemy to manage it. The inverse is also true where they killed a lot more units but did receive roughly a quarter of the enemy's total losses. This as well seems to be a flexible statistic weighted by combat outcome. After all these factors are calculated I think the numbers are plugged into a weighted equation with a random component for a score. This would be seen as A+B+C = x.

Where A would be the combat rounds with a curved score reaching maximum possible at a mid 20 to 30 range and falling off on either side and a maximum score of roughly 250 (keeping in mind that this mean can be shifted depending on total number of forces, or not at all in the case of severly out numbered attackers or defenders). B being a function of combat losses total directly proportional to the value imputed with a maximum of 550. C being your random luck factor with a maximum score of 800. if X is = or greater then 1200 then you get a medal determined randomly from the possibilities for the level of the location attacked and (since barb towns seem to produce a normally higher level of medal then a valley or flat at the same level) the type weighted towards one medal or another by the ratio of losses to kills in combat. If else then you’re boned.

This is as far as my thinking goes I'd like to see if it's possible to start nailing down more exact numbers in terms of troop size, composition of forces, combat rounds occured and reward recived. I'm limited to what I can feasably attack in a givien period of time and what I have in range so any input would be appreciated. Thank you.