Saturday, June 10, 2006

Pricing in Loot Systems

I've been reading this thread on Loot Systems in the Guild Relations forum. It's a pretty interesting read.

There are several issues to consider when designing a Loot Systems. One important one is pricing. There are two main ways of pricing items: fixed prices, or bidding. The problem with fixed pricing is that an item's true value really depends on a number of factors, including class, spec, amount of DKP, guild's progress in endgame, other options available, rarity, etc. At best, a fixed price can only approximate all these factors. For example, I do not have the Illumination talent, and as such, I value [Azuresong Mageblade] a lot less than a Holy paladin would.

As well, getting the price wrong makes your loot system act funny. For example, when my guild killed Onyxia for the first time, we set the price of Onyxia's Head at twice the amount of a Tier 1 set piece. Because no one had really built up significant DKP reserves yet, only one person was willing to take the Head (and I think he did it just so it didn't rot). Now, I'm not saying that the price was absolutely wrong, just at that specific point in time, people considered it to be overpriced. Now, when people have started building up DKP, more and more people are considering it reasonable.

The other option of setting prices is bidding. Bidding generally measures how much a player wants an item for her character. Items go to the player who is willing to pay the most for it. This is an amount which is closer to the item's true value at that point in time.

The problem with bidding is collusion. This is where an item is only valuable to a small subset of the players and that subset agrees to buy the item for a smaller amount than it's true value. They do this in order to save points for when they need to compete against the entire group of players. In WoW, collusion is really attractive because of set pieces. Set pieces provide a ready-made subset of players who know that their collusion will not be interfered with. So all the paladins might collude to spend very little on set pieces, saving their points to outbid the warriors for the uber-2H weapons.

In addition, a strong class-priority system on non-set items will also encourage collusion on those items. If paladins have priority on 2H weapons, there is a strong incentive to collude in order to maximize effectiveness when bidding on healing items available to multiple classes.

I think I prefer bidding systems, though. Allowing the player to determine the value of items for himself is the most effective option, in my view, and one that gives the most power to the player. As well, bidding systems tend to automatically counter problems with inflation, because value can also be expressed as percentage of your current total. If something costs 40 DKP, is it expensive or cheap? If you only have 50 DKP, it's expensive. If you have 500 DKP, it's cheap.

The problem then, is how to resolve collusion in a bidding system.

To resolve collusion, I think the best way is to separate out class sets from non-set items. When awarding DKP, give each player Set DKP, and Non-Set DKP. This way, colluding on set items does not give an advantage when bidding on non-set items. It also has the advantage of ensuring that players can save up for something special, while still improving themselves.

As well, I would not impose class priorities on non-set items, making collusion ineffective there. This has the added advantage of fitting into my philosophical beliefs on character improvement.

So my ideal loot system would feature Set DKP and Non-Set DKP, no class priorities, using a Vickrey Auction (single secret bid, winner pays price of the second highest bid) to determine the actual winner of the loot. I like the Vickrey Auction, because the optimal tactic is to bid your true valuation of the item.

No comments: