For multiple reasons, we have decided to let you in on the formula we use to determine how much damage is dealt is duels. This will help you better decide just what kind of equipment configuration you want to shoot for. Also, it the duels are borked, this will help you verify that they are and report it.
The classic duel damage formula is exceedingly simple and is show below:
So then, using this formula, when fighting an opponent of equal rank. you should do no more than 210 (120 x 1.75) damage and no less than 60 (60 x .75) damage. Keep in mind there is a 5% chance to land a critical hit (crit) that doubles your damage and a 5% chance to miss. This also assumes you and your opponent have the same number of votes.
This formula also gives you an easy way to see the difference that rank makes. Let's assume that two Scientologists attack each other and the random roll is 100 for both. They will both then do 100 damage to each other. But how much does that change when one of them is promoted? Let's look at this handy table:
Rank  Dmg Dealt  Dmg Received  % Adv 
Rubber Dog Turd  101  99  2% 
Pink Peep  103  97  6% 
Mullet Farmer  104  96  8% 
Mullet Vender  106  95  12% 
LOLcat  107  93  15% 
Awesome Acolyte  109  92  18% 
Gorilla Guerilla  110  91  21% 
Awesome Assassin  111  90  23% 
Theorist of Awesome  113  89  27% 
Incarnation of Awesome  114  88  30% 
Ninjas & Special Ranks  116  86  35% 
As you can see, when the rank difference is just a few ranks or less, the advantage is not too significant. But, once you hit around a 5 rank difference, the advantage because almost too much to overcome. You should also note that a 5 rank difference between LOLcat and a Scientologist (15%) will be slightly more than a 5 rank difference between an Awesome Acolyte and a Ninja (14%) because of how we add the rank constant. Also note that these numbers get closer if the roll is lower than 100 and get farther apart if the roll is higher than 100.
Now it is time to get a little complex. Do you have your math hats on? Adding equipment to the mix obviously copmlicates the formula just a smidge. Take a look:
Pretty self explanitory, right? To sum up, the attacker's damage and the defender's armor directly adjust the random roll. Then that adjusted roll is multiplied by the attacker/defender ratio that is shown. Then that is multiplied by the action multiplier, and finally, the rating wieght modifier is added.
A few notes. First, until a couple days ago, armor was divided by 4 instead of 2. It was decided that armor was way too crappy that way, so it was improved. Second, if the result of an action is a counter (attacker wild and defender defend), the forumula takes the defender's attack stats and the attacker's defense stats and then multiplies the result by 1.15.
Third, take note that your Rating statistic is applied to the top of the ratio portion of the formula if you are teh attacker and the bottom if you are the defender. Thus, a point of rating acts the same as a point of attack and a point of defense combined. Fourth, adding the average of the duelists equipment rating divided by 4 is to make it so that duels where people have very high health do not last for 50 rounds like they used to.
Monster duels differ only slightly from this formula. The rank mod in monster duels is always 75 and there is no addition of the average of duelists ratings. Other than that, the formula is the same.
Crit, resilience, hit, and dodge and not shown in the formula because they work differently. Everyone has a natural 5% change to land a critical hit or to dodge an incoming attack. For crit, subtract the defender's resilience stat from the attacker's crit. Add 5 to the result. Then take a random number from 1 to 100 and if it is smaller or equal to that result, the strike will be a crit, doubling the damage dealt. Note that if the defender's resilience is greater than the attacker's crit, the attack will never be a crit.
Dodge and hit work pretty much the same way. On an incoming attack, subtract the attacker's hit from the defender's dodge and then add 5. Then take a random number from 1 to 100 and if it is less than the result, the defender dodges the attack. Note that if the attacker's hit is greater than the defender's dodge, the defender will never dodge the attack.
One last thing to keep in mind is that we are always tweaking things to make them work better, so this may not be 100% accurate forever. However, I expect this to be very close to 100% accurate for quite some time.
