That thread isn't there anymore, but if it's the one I'm thinking of, this is a modification of the nurfed formula.php that uses the values from that thread, in place of nurfed's nutty values and shadow items. I also had to modify a line in apply_formulas.php to make it work (don't remember which line, but you'll see it in your php error when you go to Apply forumulas to new/edited items on the admin page).

On a default new nurfed dkp site with this formula install the result is this: http://dkp.subversatile.net

(note, there's a few items that Nurfed usses "assigned" dkp for, so those aren't updated in my adjusted site)

Unfortunately, this is just another (albeit, better) solely stat-based score calculation. "Quality" is inherently subjective, so the only way to apply score multiples based on the quality of budget use is to go through and arbitrarily decide this for each item, in which case you might as well arbitrarily assign each item's score and forget the whole calculation scheme anyway.

Heh, 10% is exactly what I've been working with.

Incidentally, people who want to play with this can loot at pages 5 and 6 of this:
http://www.sigilguild.net/hamlet/Sig...perimental.xls

You may not find me funny but at least you find me informative. But yeah, as stated, Nurfed DKP is highly subjective consider some stats will be rather different for a class depending on talent spec and already established gear. (Crit becomes more powerful depending on your current attack power value, Spell Crit is less valuable on an SM/DS Warlock while it is much more powerful for an Ice Mage).

Essentially my primary displeasure with NDKP is the fact that for all the mathematics involved, it will always be wrong unless they charge you based on class, spec and current gear, which is so wildly improbable to properly judge. Not to say it isn't "fair" but its so much effort to be fair and just fall short.

That would be why we ended up saying we'd price based on estimated stat budget and accept that some items suck. I don't particularly want to meet the druid that wears Rune of Metamorphasis outside of WSG, so I don't mind rotting them for nexus crystals.

Nexus crystals have a real value, most of the items you'd end up reducing the price on because they suck don't.

Now, situational items that may apply to raiding? That's what progression mods are for. We found subtracting 10-20% of the cost made progression items (we define them as resist gear for the most part) cost sufficiently less than equivalent tier gear that it would always be available at downgrade (free) if it made it to that round.

If you use a Full/Up/Down pricing tier, you don't need the prices to exactly reflect each item's subjective worth to a class, you simply need them to reflect where an item falls in terms of total budget, and let people pick the items that fit their spec. Decisions will be made to take items at full if they're in high demand (which is good, it helps to equalize out the old timers vs the new folks) and at upgrade if they are marginal, and at downgrade if they are situational or "just for giggles". At the end of the day, each player has an equivalent amount of statbudget spent on their highest price items, and assuming they're competent, they get the same benefit. Trust your players to decide "is this worth it", don't try and make your pricing system make items "bargains" if noone can think of a reason to take them. If you absolutely have to make something a bargain (we decided that we'd rather players took situational resist gear in advance of the fight showing up so we don't have to farm it later) then do it categoricly. (We mod the following, as an example: Items with a combined resist of 16-25 get 10% off their price. Items with 26+ combined resist get 20% off. Quest items get 10% off due to the extra labor required to put it together, and guaranteed drop items like nef's head get 25% off.) None of that is really needed, but it certainly helped smooth the way for the move to the new pricing system.

I found this thread interesting as I've also considered the possibility of building a way of automatically pricing items. It'd certainly save a ton of time when new content is released.

The question I had was, with the introduction of desirability coefficients to adjust the individual item value or a more general scale-coefficient such as the 10% mod proposed for class specific adjustments, how exactly does the system still end up saving time and energy? I'm assuming that the goal is to save time pricing items and to end up with a more uniform/consistent set of prices.

Basically, once you run the item pricing formula for all the items in a dungeon, you're left setting the value for the subjective coefficient. In order to do this, people will look at the prices of comparable items and decide that this item is either too high or too low, and tweak the coefficient till the resulting price is subjectively more accurate. Unless I'm very much mistaken, this process will be exactly the same amount of work as coming up with the price independently. I guess the ilvl/stat budget formula part will at least give you a rough approximation to start with and can save time that way. However at some point I think you might find yourself saying " the formula priced [X] at 150, but it should really be 165, so lets go calculate the subjective coefficient that makes it 165... ". At some point, when you already know what the price should be, it becomes more efficient to simply dispense with the formula and say " the item is 165." and be done with it. The number of items for which the subjective coefficient is 1.0 seems like a good indication for how 'good' the budgetting formula is. If this number is not, say, 75% (handwaving) I'd suggest that you'd be saving time by just setting the prices directly.

My second question is a little more general and on the topic of upgrades. As I understand the EJ loot system, it seems to be designed around the idea that people with equivalent 'power' levels have paid the same amount of points. This seems to be the purpose of the upgrade system.
Firstly, how does the ilvl/formula approach upgrade prices? Simply subtracting the value of two items because they happen to be in the same slot neglects some very significant power scale improvements.
Obviously a warrior in full wrath, with a good dps kit, 2 different situational tanking weapons, a full epic NR/FR/SR kit, etc is vastly more powerful than a warrior who has only 8/8 wrath and a single weapon. However a straight-forward slot-based upgrade system would have both wars paying the same price - the first war having gotten a significant amount of his 'power' for free.
Obviously if a player loots an item that is a clear linear upgrade, and DE's or permanently banks the old item in that slot, then the price-difference is proportional to the upgrade in power. However, if the item is situational/sidegrade is there a way of figuring out the increase in power from the ilvl/stat-budget formulae? I.e. is there an automatic way of computing the relative value of two items based on the stat budgets and therefore coming up with a reasonable upgrade price automatically?

This question is more in the 'food for thought' category. I'm not sure there is a good answer here. I was just intrigued by the vague idea of subtracting the stats for item1 from item2 and running it through the ilvl budget calculator and coming up with a delta price or something similar. No idea if this would be useful at all though.

Most ilvl pricing schemes use the estimated stat budget as the base price of the item, with an adjustment to weapons so that their base dps is not "free", although the system still works if you give mellee/hunters the dps for free. If you want to be really wacky when adjusting weapon costs (and closest to correct) you could determine what the budget for dps on an item was (hyz posted the curves in another thread, they're messy) and add the statbudget to the dps budget to price weapons. We short circuited it by doing that a few times and taking the best fit slotmod for those weapons rather than setting up an additive formula. I may go back to the additive formula for the expansion, if I find a neat way to implement it. In any event, the goal is to plug in a set of obvious values (in our case itemlvl, quality, and slot) and recieve a consistent price. Our system does that, so we're satisfied with it. Note it's not designed to do anything but price stuff in a consistent manner, that will stay consistent even if blizzard goes back to monkey with stuff.

The full/up/down buying system is designed to prevent useful sidegrades from being disenchanted because a player wants points for a future purchase. At any given time in our system, you can glance at total spent and have an estimate of how many items a player has taken at full, which in turn corresponds to their tier in progression pretty well. There's some exceptions, because occasionally a player grabs an item at full that they could have taken at down/up, but those events are healthy for the system because they do a good job of equalizing the gap in earnings between long time players and newer members. Each full purchase you make that isn't required, moves your break even date forward (the point at which you paid for all the epics needed to fill out a character sheet, and could start going up/down on any item at all). Eventually, assuming recruitment stopped and everone played/lived forever, the break even dates would all end up on the same day. That's a good thing.

The best thing about this pricing is that while it's not the best on an individual level (You could certainly argue that two ilvl 81 hats give different value to different classes) it's easily accurate enough overall. It's a time saver, and argument avoider. There is no value in arguing about pricing individual items "exactly" because the entire pricing structure is arbitrary to begin with. Who cares if cenarion is better than prophecy or might is better than nightslayer? It's really not worth the time it takes to type, much less the time of your officers to nitpick the exact price of each. Keep reminding yourself -- DKP is an arbitrary system that is implemented to save time and reassure members, not to provide perfect distribution, or even to provide the best raid usage. We're here to see that two identical attendance players can expect to loot similiar amounts of hats/chests/swords as their preference dictates, not to determine which items are least likely to be looted and discount them, and which items are most likely to be looted and inflate them. We're not a grocery store out to make a profit on our dkp sales. If someone want's a terribly stat allocated 2hander, that's their business, the price is on the tag. I don't need to cut them a deal because it's terrible, or gouge them if it's amazing. It's a 2hander, a ilvl 63 epic two hander, and that's the price.

If the pricing is balanced so that for each class to be in optimal gear or "character power or raid worth" would be paying the same in RP then resist gear would naturaly be factored into this and would result in classes that had to loot heavily in this area having lower armor costs. This is my understanding anyways.

Not to beat a dead horse...

Edit: I was beating a dead horse. Don't post while drunk kids, do it while bored at work.

REBIRTH.

Anyway I'm struggling to come up with an ilevel-based system that scales nicely. If you do 5*(ilvl-60)*SlotMod with Efficiency and Quality set at 1, it seems really messed up. A Tier 1 hat costs 30 DKP, a Tier 2 one costs 80, and a Tier 3 one is 140. I don't think the cost is upgrading in tune with the increase in power there.

Anyone else thought about this some more since June and come up with a better formula for pricing?

We simply use ilvl*slot modifier. Made one-handers a 90% modifier and 2handers a 150% modifier, and dropped trinkets to 54%. We do custom price certain items though because sometimes an item is just not very good and not worth what its ilvl is.

Why would you drop trinkets?

Looking at it from a rogue perspective, outside of weapons, if you take all of the gear from MC, BWL, and TAQ, and look at the relative upgrade most rogues will get out of it, the Jom Gabber, Badge of the Swamguard, and Drake Fang Talisman will stand out above the rest. And then for casters, ToEP is the weapon equivalent in MC.