We're doing what Kalman noted earlier. The player chooses their primary path ie. tanking for warriors or enhancement for shamans. For the warrior they pay nothing for their non-primary path gear but only once the warriors that have it as their primary path already have the gear in question.

Some classes will obviously be recieving more overall loot than others but it shouldn't be a big issue.

Depending on how many prot warriors you have, I'd have to say that this is probably going to be a pretty rare scenario in 25 mans. We went through the entirety of Naxx with 4 prot warriors, 3 of which were usually online and in the raid. I'm pretty sure thats more than most guilds, and there were very few fights in which they had to take off their tanking gear and put on dps gear. The only ones that come to mind off the top of my head are Loatheb and Sapphiron, one of required gear that was mostly crafted. That leaves Loatheb, a fight for which a prot warrior isn't of much use, geared or not they'd probably be replaced by another class before the boss is farm status.

In most cases its the other way around, dps warriors have to put on some tanking gear and pick up an add. In those cases, their gear usually isn't that big of a factor. The job can be done with gear that's outdated for those slots. Ie, BWL gear should be enough to offtank pretty much anything in Naxx. That can be compared to instance blues being good enough for offtanking the first tier of raid content released.

We are still working on the details of the system, but what I have in mind is something like this. Players classify themselves. There are 3 pools: DPS, tanking and heal. You decide how to split your DKP. For a resto druid with some DPS ambition it might be 70% heal 30% DPS.

Then we use bidding. Bidding helps this system immensely because one feature, good or bad, of a bidding sytem is that the lesser loot is typically disporportionally cheap. That is, an item that in a fixed point system might be only 10% cheaper could easily cost only a quarter of the points since the competition for lesser loot is typically minimal. That way, hybrids, even though they do not get discounts on their second set, effectively still get a good amount of loot for the off spec. They just do not get the best loot, and honestly (and please note the class in my profile) they should not.

The advantage of the system is that players are not pigeonholed like it happened a lot in the past. You classify yourself, and you also have to compete based on that. If you are a priest and put 80% of your points into DPS be prepared to compete with warlocks for raid spots. It also tells the guild how much of each role you have. In the past, there was a strong connection between class and role, but now that is not the case, so the guild needs to know how many healers they have, not how many priests/druids/pallies/shamans.

A player can change their point allocation, but only infrequently. There are some more wrinkles but they are beyond the scope here.

Another advantage of the system is that once I got off my ass and wrote some simple tools for it, a raid will take less than a minute to input. All you need to do is copy a file to a directory on the server and done. No manual processing will be required. As I said above, to me that is more important than getting every single facet of the system right.

I like Kalman's thinking on the issue, I will be advocating a declared role with last-priority discounted off-role system. (But then again, I'm Feral McGee).

I liked a system a previous guild I managed used (that might be why I designed it), which was inflationary but zonal DKP. If something isn't working, there's effectively a reset every new raid zone anyway, so you can tinker with numbers to noone's lament (there's carry-over for populating the new zones for a few kills, so vets have some incentive, but can't lord over newer members new kill loot). Not directly related, but if people start finding a way to game whatever system you do come up with, "Hey, guys, we can fix this in [new-BWL], without disrupting the standings at the moment."

Personal experience shapes our opinions of this to a great degree. 9/10 of the warriors my guild has ever had have been rogue-wannbe lootwhores who wanted two huge piles of purps yet never wanted to actually tank anything. When raiding on my warrior alt I considered "needing" two sets of gear to be a privilege rather than a burden. So I don't see how rigging the system so hybrids get twice the loot for the same price is more "fair" or desirable in any way.

Perhaps you should work on the problem of having a bunch of shitty warriors. Rather than screwing half the players in the guild who have more than 1 set of raiding loot.

Are you suggesting that one can't engineer a DKP system that solves all of society's ails? Surely with the correct formulae we can summon Azathoth from beyond the Outer Wastes and rain down sufficent loot to appease everyone. I'm fairly sure that the first step is bringing a donkey to market.

I really like the idea of just picking a spec and getting stuff in the other one/two at reduced price as long as no one with that spec goes after it. I'm not sure I'm a fan of the whole percentage split, since that just seems to be complementing something that would be best implemented in the keep it simple fashion.

I might bring that up to my guild, currently we are only looking at doing a zero sum for 25 mans with 3 bid levels. (low/med/high with associated point values).

Between cross-class issues and working out how to fit in Karazhan loot (it was bad enough last time with jealousy between people getting Leggings of the Black Blizzard etc for free vs those paying serious DKP to pick up equivalent in BWL), we've said screw it and gone to loot council.

A percentage split, besides considering the relative amounts of weaponry (and trinkets!) that different classes need should also consider whether damage/PVP gear (especially for pure healers and pure tanks) is included, or whether those classes have to pay extra compared to DPS classes to end up with both PvE and damage/PvP gear. I say damage/PvP because you never know when a PvE fight's going to come along when you want your priests to be able to do damage too...

The reason why the "when nobody wants it" method is problematic can easily be seen from the last 2 years of raiding. Simply put, there is not an awful lot of decent loot that nobody wants. How much DPS gear will a holy priest get if he has to wait for all mages, warlocks, moonkins, shadow priests and elemental shamans to get it first? Pretty much none that is worth anything.

That will then again lead to healers being pigeonholed into only healing without any ability to do decent damage. How often have you heard senior holy priests complain that they could not get DPS gear because they were put even behind some new mage? And now imagine this much much worse because those guys will also be put behind new shadow priests that chose the DPS track. Under those conditions, who would chose the healer track?

The best way to avoid all that requires the healers to have some form of GUARANTEED DPS loot. Not just loot that nobody wants, because that will never be guaranteed with turnover and guilds quitting older dungeons. The only way to attain that is to give them some DKP that they can spend on DPS loot, even outbidding the actual DPSers. Just not too often.

That is the reason for the percentage split. Which btw is not all that hard to set up and administer.

Thanks. It has a bunch of interesting side effects too. For example you can ask applicants to list their desired split. If someone tries to get in via the good old "I am a total healbot, I love healing and will do nothing but heal 24/7 if you only let me in" line then he better back that up with a 100/0/0 or at least 90/10/0 split. And he will be stuck with that split for a while so he cannot spec shadow next week and collect DPS loot.

Therefore, you can recruit much more accurately and it is also a convenient way to keep track of what exactly every member currently wants to do with their character.

It is also possible to apply the same principle to a fixed point system, but you might have to lower the cost for off-specs a little, otherwise they might not get enough loot to be viable in those situations that you need them.