Updated.

I’m roleplaying as a designer again. This is my “Honor System” for WoW.

This is a rough plan of how I would try to design the whole “honor system”. It tries to address all the problematic issues that the players are bringing up but still mantaining a focus: to be fun.

So far, these are the issues:
1- Prevent to assign penalties if the attack is a mistake
2- Allow an high level character to defend itself if attacked by a low level opponent without incurring in a penalty
3- Not reward a zerg of players killing a poor guy wandering alone, even if on the same level
4- Avoid to penalize support classes and incentivate the players to group
5- Make a dishonorable action have a greater impact if done by a player high on the honorable ranks

I’m not a mathematician so I’ll try to explain how the general system should be built.

The first part is about how we assign the points and when. In this system I won’t address external goals like accomplishing a mission. I’ll just include player-killing. So, the honor points should be assigned when an enemy DIES. This is the first step. Attacking someone doesn’t imply a gain or loss of honor. The honor changes only if the enemy actually dies. So we solve the (1) point.

The second step is to determine if the honor we are going to change after a kill will be a penalty or a reward. My idea is to simply determine this part by “tagging” the enemy. It means that who starts the attack is responsible for it. If an high level player starts an attack to a low level player it will incur in a penalty when he kills it. Instead if it’s the lower level player to start the attack, the high level player will be able to defend himself without incurring in a penalty (neither a reward since the low level player isn’t worth much). Following this idea we solve the (2) point.

The second point also leads directly to the third. This point is only partially solvable and my first plan isn’t playable as I expected. Basically two players level 55 attacking a single enemy level 55 won’t receive a penalty because the system would never be playable and fun. Still, the reward will be scaled and splitted (so not as high as if the fight was more “fair”). Nothing can work better than this and on the other side we have a positive effect: players will be encouraged to play in groups for a better protection. So the (3) point is only partially solved, the “balance check” will be done by comparing the “average” level of a party with the target to determine if the result is a reward or a penalty. But the number of players won’t directly affect this check. To prevent “zerging” the reward should be “scaled” between all the groups that took an active role when damaging the target. This means that the reward will still be very low when “zerging”. Best design scenario possible, I think.

For the fourth point we delve more into the math. Support classes won’t kill directly the opponents. This means that if the honor is assigned “per kill” these classes will be penalized. It’s not all. Because we should also push the players to play together. Casters will more likely die first considering their weakness. If we reward only who is alive we’ll finish again to penalize these classes. How to prevent all this and reward the group? The idea is to reward group survival. If a group survives an encounter (51% alive) it will gain a bonus multiplier (with a diminished return softcapped at 2.5). The more a group survives the more the bonus builds up (and the same group will be worth more if killed). If the group flees from a PvP battlefield OR is defeated (only if 50% or more of it dies) the bonus is lost and resets to zero. This means that the reward (or penalty) will be equally distributed for all the group. Both to those alive and those who died. On the same time the reward will depend on the “group performance” as a whole and NOT tailored and measured for each participant. This solves the (4) point.

The last point is about how we calculate the penalty. Not only the “honor points” should behave in a similar way to a normal exp table. But the final result should be adjusted a last time before being assigned. This last modification depends on the level of honor of the player. This means that as a player climbs up in the honor ladder, the honor points should slow down, progressively. Accumulating more honor should become harder as you go up. At the same time a dishonorable action should have the opposite impact. It should increase as the honor increases. So, while the reward should be achieved as an inverse proportion, the penalty should work as a direct proportion. This solves the (5) point.

Ahh. I’m tired. I wrote and read too much on this and other forums for a lot, a lot of hours. I hope it will be somewhat useful. Really hope.

—
I wrote down an algorithm to explain the plan I posted above:

1- A battle begins.

2- The server registers which *group* does the first attack and “flags” that group as “attacker” and the other as “defender”. Every other group joining the same battle executes this point again to be flagged properly.

3- The fight takes place. The server registers who damages who, based on the groups and not on the single character. Also registering the total damage (in %) for *each group*.

4- Someone dies. The server considers each group that damaged the target and starts the following procedure for each:

a- A “balance check” is done by calculating the average level of the two groups (attackers and defenders).

b- The two average levels of the two sides are compared to measure the “fairness” of the combat. Determining if there will be a reward or a penalty.

c- The reward/penalty status is once again modified by checking if the winning group is flagged as “attacker” or “defender”.

d- The amount of “honor points” (to add or detract) will be calculated based on the “value” of the target and the % of damage dealt by the group to this target. then split equally between every member of the winning group.

e- For each player the amount of honor is adjusted again by considering the rank (as described at the (5) point of the plan I posted above).

– |a to e| Are repeated for each group who damaged the target.

This is the whole algorithm. The only part that it is missing is the modifier that I desribed at the (4) point of my plan that I let out to not make things too complex.