Nezralix

If you don't care, then I'm obviously not going to convince you of it. I'm not even sure why you feel strongly enough to repeatedly complain that you don't see the point.

 

dinesh

Perhaps. But using an addon like Pawn (or, to a lesser extend RatingBuster) will much more quickly tell you exactly which item has "better" stats. And it doesn't require that you be comparing items of comparable item levels either.

For my own gear, "whether stats [of my current gear] are budgeted appropriately" does not, at any point in time, enter into my decision process about what other gear is an upgrade. I will happily wear an "inappropriately weighted" piece of gear if it ourperforms other gear and gear sets I have.

Perhaps that is unique to my class, since our stat modeling, while complex and interactive, is seemingly well suited to "equivalency ratings" that can be directly compared piece to piece.

Perhaps I missed it above, but are you familiar with addons like Pawn? In case not, here is a sample tooltip it creates. Used alongside an addon like EquipCompare, you can mouse over any item and immediately look at the calculated values for both the new piece and the piece you're currently wearing, and immediately see if an item is an upgrade, and by "how much". No need to even consider stat weighting, appropriateness of distribution, or anything. Though a pie chart might be easier to digest than the output of RatingBuster, for example, I don't think the same is true of the output of Pawn.

I'll grant you that one thing which a pie chart might be more helpful for is seeing that an item, which may be an upgrade for you, was actually designed with stats for another class for which it is likely a much bigger upgrade. For me at least, though, this doesn't really require a pie chart to do - a quick glance at the item stats, looking for things which are not traditional stats for my class, will tell me the same thing.

It's clear to me that you want what you want. I have and had no intention of trying to dissuade you. But it's also pretty clear to me that what you want would, for me, be an inferior mechanic for achieving the goals you've expressed than other tools that already exist. Still, I might install and check it out, should it ever be developed.

Finally, I'll try one more time to explain to you, I'm not "complaining" about your request. I'm trying to understand it, to see whather it might also be useful for me. To each his own.

 

Belenos

I think this is interesting, if nothing else it may give us insight about how items are designed, and more understanding into WHY Pawn or lootrank.com may say an item is very good for us.

I like the bar graphs! You should be able to make a cool pie chart with a Javascript pie chart library, a quick search turned up this one: PlotKit - Javascript Chart Plotting | liquidx

Personally, I think it'd be slick if this became a plugin for Wowhead.com or something, so you could click a link to see any item's stat allocation.

Author of CasterWeaponSwapper: suggestions welcome by forum PM or to wikwocket@gmail.com.

 

dinesh

I'm going to shut up for a bit so as to stop overwhelming the thread with my blather, but at least in my case, the numbers Pawn uses for weighting are input by me, after being calculated in one or two very precision rogue gear simulation sheets available here on these forums. They give a much more detailed and precise insight into the value of each stat than any back of the envelope guess based on a pie chart could ever do.

Having said that, I believe the rogue sheets are much farther along than those for most other classes, both because we've been lucky to have great folks participating, and because of the nature of the interaction of our mechanics. YMMV for other classes.

 

Marek

Pawn or Lootrank say an item is good by simply multiplying the user-giver worth of a stat with its ammount on an item. So if the result is greater for one item than another, it really is better.

With this addon we could see which item has the most "raw" stats, which can also be achieved by entering the default stat modifiers in Pawn or Lootrank, for example 1 for strength, 0.5 for AP, 0.86 for spelldamage and so on. It is more LIKELY that an item with better distributed raw stats will also end up being better, but it is not given.
Item A: 40 AP, 20 Str -> better raw stats
Item B: 79 AP
Given these two items you won't make a decision based on the raw stats. You will wear the first as a druid and the second as a rogue.

A decision whether item A or B is better can be made using AEP or similar class specific models for stat evaluation.
So if you want to make a decision about the quality of the design of a specific item, you have to take class/talent specific weightings into account. In my opinion this IS an interesting topic because it is far from trivial.
An item with 2x of stat A and x of stat B might not be optimally itemized even if you value A twice as much as B. There is a point where you could trade in 1 A for more than 2 B because A became increasingly more expensive.

 

Margot

I could see this being useful if you didn't have wowhead and didn't know/have time to plan out your gear upgrades.

Let's say you want a max +heal set, but not to the detriment of other stats. You don't need a Pawn value or a knowledge of items at this point. If it's a well-budgeted +heal item, it's going to last you a while. You can do this across a whole tier/raid instance of items, and end up with whatever you were looking for. In this case, you might not have the highest absolute +heal value, but you will get a lot of mp5, spirit, etc along with it, even though your 'goal' was +heal.

Basically, it helps you look for 'bargains', which may not be the same as best in slot, so I can see why people would think it extraneous. At the same time, 'bargains' may give you the highest overall combination of stats.

 

Vykromond

The primary use case for me of an addon with this functionality would be for participating in beta tests of any new WoW instance or raid dungeon, I would use an addon with this functionality to quickly determine which items on the drop tables have been poorly itemized. This can be determined with some work either by eyeballing the item or through something like Pawn with a ton of custom settings for each spec/class, but being able to see at a glance how an item's stats are misproportioned or underbudgeted increases the quality and speed of the feedback I can give.

It would be interesting purely as an academic curiosity, too.

 

Marek

The problem is misproportioned stats ARE optimal in most cases.

For example, the badge ring [Angelista's Revenge] spents 921.2 points of item budget.
29 AGI (316.2 points spent)
58 AP (316.2 points spent)
126 ArmPen (139.9 points spent)
28 Stamina (148.9 points spent)

How would this item look if it was optimally designed for a T6 rogue?
From the Roguecraft 101 thread the weightings for rogues in T6 content are AGI=2.19 EP, AP=1 EP, ArmPen=0.38 EP.
This would give the ring a power of 169.39.

If we keep stamina equal and only adjust the DPS stats we can distribute 772.3 points over the three stats. Given the rogue weightings the ring would be optimal (or close to it) with the following distribution:
23 AGI (212.8)
43 AP (190)
222 ArmPen (368)
Only 770.8 were spent in this case, but the power of the ring became 177.04. Note that the distribution of AP and AGI shifted from an equal distribution (optimal for raw stats) to a stronger representation of agility.


I used the following formulas (taken from WoWWiki) for the calculation:
1. Raise each stat on an item to the power of log2/log1.5
2. The sum of all values is the item budget spent on these stats. If you want to compute the item level you will have raise this sum to the power of log1.5/log2 and do some multiplying with a quality factor.
3. For stat redistribution take the sum of the budget spent on the stats you want to redistribute.
4. Redistribute the stats in an optimal way. At this point I am not sure if an analytical solution exists. I wrote a simple program that raises one stat at a time that would be optimal at this point (needed item points divided by power increase)
Example:
Raising AGI from 20 to 21 would cost 14.57 points of item budget and increase power by 2.16 equals 6.747 item points/power.
Raising AP from 35 to 36 would cost 6.58 points and increase power by 1 equals 6.58 item points/power.
In this case raise AP by one instead of agility.
This algorithm is NOT optimal because it makes a local choice between the stats (greedy algorithm).


There are quite some scenarios where calculations like this would prove useful, for example:
1. With class/spec specific DPS formulas it is possible to calculate the potential DPS a class could do given a specific item/progression level. This would provide a metric for cross-class DPS comparision and reveal scaling issues and overpowered/underpowered talents and abilities. From a simulation point of view it is easily done by replacing the fixed EP values with the DPS formula of each class (I don't even want to try to approach this from an analytical point of view)
2. It is possible to reverse engineer the weightings Blizzard assigns to each stat, for example the badge ring would be optimal with EP values of AGI about 2 EP, AP about 1 EP, 0.204<ArmPen<0.205. I am not sure IF Blizzard has fixed weightings for each stat in mind, but a close look on class sets might reveal more.
3. It simply tells you if the item you have could be better

[EDIT]Fixed a calculation error on the stamina item budget, it is 2/3 not 3/4. Changes highlighted.[/EDIT]