Paladin Tanking Item Points - PTIP [gear ranking system]

All things related to the expansion

Moderators: Fridmarr, Worldie, Aergis

Postby cordelia » Tue Sep 23, 2008 4:33 am

Noradin wrote:The way you combine modifiers is not correct.
Incrementing modifiers are combined additive.
Decrementing ones multiplicative.
(e.g.: DS (+.15) and Kings (+.1) combined result in a modifier of +0.25. You'll get 1.25 times the strength you would get without them not (1.15*1.1=1.265) times.)
Conversions from one stat into another or into damage from an ability and from hit to crit or the like are multiplicative as well. (If you have Kings, DS, Redoubt and the +5% BV meta 1str on gear will give you 1*(1+.1+.15)=1.25 strength and 1*(1+.1+.15)*.5*(1+.3+.05) BV.)


Can someone else please confirm this? It means lots of changes to the formulas :P
User avatar
cordelia
 
Posts: 630
Joined: Tue Oct 09, 2007 3:33 am

Postby Rehlachs- » Tue Sep 23, 2008 4:37 am

cordelia wrote:Err, that's a mistake - fixing it now. lol. And, as I said, gems are confusing and complicated - suggestions welcome on how to deal with them.


aye ^^
I don't want to harm you. if I sound too harsh in any way.. I will accuse the English language.

Look up improved Divine Spirit. 10% of Spirit is added as SpellPower. However, each point of Spellpower is only worth 0.86 ItemPoints. So, only 0.1*0.86 item points per spirit is added by the buff. There was a typo. I forgot to modify iDS by Kings. That's fixed now.


sure, I know Imp DS *hinting at my first post*.
my current problem in understanding the calculation is, that Spirit got a modifier for other stats - in that case spellpower because of ids.
but Agility contributes to other stats, too (you already included them at other calcs of your posts). so, Agility improves Dodge and Melee Crit. Int improves Spell crit. and so on.
I'm missing all those modifiers in the final PTIP calculation - hence the confusion.
Rehlachs-
 
Posts: 690
Joined: Tue Nov 13, 2007 1:45 am

Postby cordelia » Tue Sep 23, 2008 4:57 am

Rehlachs- wrote:
cordelia wrote:Err, that's a mistake - fixing it now. lol. And, as I said, gems are confusing and complicated - suggestions welcome on how to deal with them.


aye ^^
I don't want to harm you. if I sound too harsh in any way.. I will accuse the English language.

Look up improved Divine Spirit. 10% of Spirit is added as SpellPower. However, each point of Spellpower is only worth 0.86 ItemPoints. So, only 0.1*0.86 item points per spirit is added by the buff. There was a typo. I forgot to modify iDS by Kings. That's fixed now.


sure, I know Imp DS *hinting at my first post*.
my current problem in understanding the calculation is, that Spirit got a modifier for other stats - in that case spellpower because of ids.
but Agility contributes to other stats, too (you already included them at other calcs of your posts). so, Agility improves Dodge and Melee Crit. Int improves Spell crit. and so on.
I'm missing all those modifiers in the final PTIP calculation - hence the confusion.


Remember, PTIP only modifies the item's base item points for tankadin Talents and Buffs. 1 pt AGI = 1 Item Point, regardless of what AGI does or does not do(crit, armor, dodge). After Kings, 1 pt AGI is the equivalent of 1.1 Item Points. 1 pt SPI = 1 Item Point. After Kings and iDS, 1 pt SPI = 1.19 Item Points. Kings and iDS are both Buffs that are applied to the base Stat, so thus, the value of that stat, DUE to tankadin Talents and Buffs, is modified. Whether or not spirit is useful is not under consideration at all. PTIP only tries to quantify the effects of Tankadin Buffs and Talents. If we had a talent to increase dodge by 10%, then AGI would have a different PTIP. The dodge component of AGI, after kings, would also be multiplied by 1.1. (I guess I will eventually need to change PTIP due to Toughness, but for now, Armor multiplier = 0 for PTIP, b/c it's kinda complex, and will need to be figured out later).

You're confusing modifications that belong in the APTIP section with the ones that belong solely in the PTIP section. APTIP is where the effect of AGI is broken down into its threat(crit), avoidance(dodge), and EH(armor) components, and the true value of AGI to tankadins is reached. But APTIP is currently very very rough - note all the brown and smelly numbers in the threat section. Thus, PTIP puts a foundation down to simply evaluate an item's relative worth. APTIP will eventually be a better indicator of an items utility, but mucho more theory crafting is needed there.

And no, don't worry about sounding too harsh. It's in this thread for critique and suggestions. That's the purpose of it.
User avatar
cordelia
 
Posts: 630
Joined: Tue Oct 09, 2007 3:33 am

Postby Rehlachs- » Tue Sep 23, 2008 5:33 am

so the final PTIP is just the actual item value of a certain item after you added the given modifiers?

I can't think of a proper usage for that number, but at least you would be able to calculate how much of the item value flows into threat or EH (for example, Legplates of the Mobile Phone is iLvl 200, 35% of its value is contributed to threat, 60% to survivability, 5% to wasted stats)
Rehlachs-
 
Posts: 690
Joined: Tue Nov 13, 2007 1:45 am

Postby Songblade » Tue Sep 23, 2008 5:42 am

cordelia wrote:Look up improved Divine Spirit. 10% of Spirit is added as SpellPower. However, each point of Spellpower is only worth 0.86 ItemPoints. So, only 0.1*0.86 item points per spirit is added by the buff. There was a typo. I forgot to modify iDS by Kings. That's fixed now.


Actually, Imp Divine Spirit is only 6% in WotLK IIRC.
Image

[Insert witty comment here]
User avatar
Songblade
 
Posts: 552
Joined: Mon Jan 07, 2008 4:46 pm
Location: Interdimensional Limbo

Postby cordelia » Tue Sep 23, 2008 6:28 am

Rehlachs- wrote:so the final PTIP is just the actual item value of a certain item after you added the given modifiers?

I can't think of a proper usage for that number, but at least you would be able to calculate how much of the item value flows into threat or EH (for example, Legplates of the Mobile Phone is iLvl 200, 35% of its value is contributed to threat, 60% to survivability, 5% to wasted stats)


PTIP, yes. And I agree, the usage of the number is vague. At best, it's some kind of raw item ranking, in terms of tankadin friendly stats. APTIP will hopefully provide a much better ability to rank and evaluate items, but PTIP has to come first.

There's an old adage that says, when you can't decide between two similar items, pick the one with the higher iLvl. Well, in this case, pick the one with the higher PTIP. It's a better pure indicator of item value for tankadins than iLvl is.
User avatar
cordelia
 
Posts: 630
Joined: Tue Oct 09, 2007 3:33 am

Postby caboom » Tue Sep 23, 2008 6:50 am

Nice ideas here.
Suggestions:

Clasify gem slots according to color, giving most points to blue and lest to red or yellow? Maybe you can base that on most popular gems used by paladins, but since the patch is coming no one actually knows what would be best to gem :P

About armor maybe you can divide armor by item ilvl, or consider item ilvl as armor score since all armor pieces of the same ilvl have the same amount of armor.

Finally maybe you could make a list of the new, reitemised, 3.0 badge tank armor that ppl could use to decide on what to spend badges on, and it would also show how your system works.

Gl
Image
User avatar
caboom
 
Posts: 562
Joined: Tue May 27, 2008 4:06 am
Location: Romania

Postby cordelia » Tue Sep 23, 2008 11:23 am

caboom wrote:Nice ideas here.
Suggestions:

Clasify gem slots according to color, giving most points to blue and lest to red or yellow? Maybe you can base that on most popular gems used by paladins, but since the patch is coming no one actually knows what would be best to gem :P


I thought about that, the problem is 20 Str gems have the most PTIP of any gem, @ 39.6. Str, of course, is mostly threat (1.36 APTIP), and only partially EH (0.588 APTIP). The question really depends on how people are going to gem. If I assume 10 Str/15 Sta, 10 DR/15 Sta, and 30 Sta, the red socket comes out at higher value, which is not necessarily correct, since Threat is not an issue, and most ppl want STA. I guess I could assume 10 Dodge R/15 Sta for Red slots. That should correctly value Blue slots best. The other issue is whether I should be using Rare gem values or Epic gem values.

About armor maybe you can divide armor by item ilvl, or consider item ilvl as armor score since all armor pieces of the same ilvl have the same amount of armor.


The problem comes from comparing pieces of different ilvl, and thus, different armor. I can't simply divide armor by ilvl, because then the extra armor from the higher ilvl piece is devalued. I think my best solution might be to use lvl 80 blue gear armor values as a BASE. And then add or subtract armor levels for all armor based on that, using the 0.07 ilvl points per armor point valuation.

Finally maybe you could make a list of the new, reitemised, 3.0 badge tank armor that ppl could use to decide on what to spend badges on, and it would also show how your system works.


Yes, I'm frantically trying to avoid manual data entry of individual items. I may have to resort to this, unfortunately. If anyone knows of existing spreadsheets with items, I would appreciate pointers in those directions.
User avatar
cordelia
 
Posts: 630
Joined: Tue Oct 09, 2007 3:33 am

Postby Kelaan » Tue Sep 23, 2008 12:37 pm

For now, I'm going to simply add up current PTIP from Stamina and BV, and then add in Armor*0.07. This will provide PTIP.EH values that have a HUGE base, due to base armor, I think, but I don't know what else to do.


I think that starting from an arbitrary base health/stamina would make sense. I suggest using full Tempered Saronite gear as a baseline... just because it's something everyone can get to. I realize that that still makes the calculations favor armor or health in varying amounts, but at least it's a somewhat more identifiable baseline that will apply to more than people-in-T6.
User avatar
Kelaan
 
Posts: 4037
Joined: Thu Jan 03, 2008 12:01 pm

Postby ferthadin » Wed Oct 29, 2008 8:23 am

I've come up with a similar system for ranking gear. But after much debating (in my head), i decided that I feel like adding in all the stat modifiers(kings), and talents (stamina, block value, etc.) does some bad things to the final ranking values that I don't like.

Block Value, and Strength are the best example. Once you figure in all the multipliers for block value, the value of Strength and block value go WAY up. This causes items heavy in block value, particularly, to go WAY up in value, pushing other (potentially better items) off the radar. Block value gains so much value when you put in the 30% from talents, that it even dwarfs stamina in item budget.

Other stats, like defense, parry, dodge...they gain nothing from talents/BoK, so their value goes way down. Do we really want to push these stats off the radar when comparing gear? I don't think we do.

I came up with the following modifiers for my sheet. I'm calculating strictly the "taking hits" benefit of these stats, and disregaurding threat completely (that's something I plan to just do by eye, between items that are close in rank). You'll notice strength comes in VERY low. But I feel like that's appropriate..for one item value of strength you really don't get that much "taking hits" benefit (a mere .5 block value). Where 1 item value of block value itself, gives you 1.534, 3x as useful)

http://wotlk.wowhead.com/?items=4.-2&fi ... :0.65:0.07

You'll notice I stuck these into wowhead, which may do away with spreadsheets for me altogether. I also throw these same values in-game into pawn.

What my numbers give you...is a fair comparison of the "taking hits item value" of an item. The item itself, all things are equal, it's not saying that block value helps you more, or parry helps you less (which may or may not be true, depending on the situation). It's a straight item value comparison.

I find that by engineering all the modifiers into the equation, you're looking at your stats from the item..and not the item itself. And while you ARE looking at the stats you'll get from the item, and that seems like it should be right..IMO it's wrong, strictly from a comparison of comparing gear. It's something I've been throwing around in my head for a while now..and I'd invite you to consider the same. Do you really want your equation to push items with block value up above all others (30% modifier). That's what it's doing, whether that's your intent, or not.

I'm not advising that this number is the end all be all way to rank items. But for me..this gives me a baseline for comparison. And from here, I can compare items similar in value, and choose the item that suits me best, for my current needs, at my current levels. And especially at my current gear level (not yet even uncrushable, not going to be uncrittable in the expansion), I really don't want to de-value things like block rating, and defense (which is what your equation is doing, with all the modifiers).



As a side note, I played a part in inspiring honorshammer's spreadsheet, with numbers I ran on prot warriors (i was his old GM)
ferthadin
 
Posts: 24
Joined: Tue Sep 30, 2008 6:34 am

Postby halabar » Wed Oct 29, 2008 8:45 am

What might help here now is to post the results from a small sampling of gear, to show what the results might be. Say, just do the shields.
Amirya wrote:... because everyone needs a Catagonskin rug.

twinkfist wrote:i feel bad for the Mogu...having to deal with alcoholic bears.
User avatar
halabar
 
Posts: 9379
Joined: Fri Jun 08, 2007 8:21 am
Location: <in the guild that shall not be named>

Postby ferthadin » Wed Oct 29, 2008 8:49 am

halabar wrote:What might help here now is to post the results from a small sampling of gear, to show what the results might be. Say, just do the shields.


it's easy for anybody to do this, comparing anything you like. Just enter these ratings into the wowhead weight scale, and go searching to compare anything you'd like.
ferthadin
 
Posts: 24
Joined: Tue Sep 30, 2008 6:34 am

Postby ferthadin » Wed Oct 29, 2008 8:57 am

halabar wrote:What might help here now is to post the results from a small sampling of gear, to show what the results might be. Say, just do the shields.


Do this using the talented shield block value weights in your equation, and you'll see all items in your list with block value gaining value over everything else (at a rate of 30%).

Compare two items with equal item value...

#1 budgeted Str, Sta, Defense, Dodge
#2 budgeted str, Sta, Defense, Block value


Item #2 will end up with a higher rating, simply because of the block value conversion rate..is this really the side effect we want from figuring in the talent bonsus? De-valuing dodge?
ferthadin
 
Posts: 24
Joined: Tue Sep 30, 2008 6:34 am

Postby Widdox » Wed Oct 29, 2008 9:24 am

We need to make sure we add something in for diminishing returns on dodge and parry. Cause at some point parry ratin will be better than dodge rating. I know it is on my gearing.
Image
Widdox
 
Posts: 350
Joined: Thu Nov 01, 2007 8:33 am

Postby ferthadin » Wed Oct 29, 2008 9:26 am

Here's an oversimplified example of my point. For this comparison I'm just assuming 1 Str = 1 Point, both peices have the same amount of Str/Sta/Armor, so it doesn't really matter, the key thing i'm trying to compare here, is the effect of figuring the 30% increase of block value into your gear sheets.


Item #1:
1000 armor
50 str
70 sta
50 dodge/defense/block rating/parry (pick any of these stats you like)

Item #2:
1000 armor
50 str
70 sta
77 Block Value (50 x 1/.65)


Item #1 calc: (1000 * .07) + (50) + (70 * .67) + (50) = 216.9
Item #2 calc: (1000 * .07) + (50) + (70 * .67) + (77 * .65) = 216.95 (rounding error)

Here are the calcs using actual item value conversions, both items get the same rank..the way blizzard budgeted them.

Item #1 calc: (1000 * .07) + (50) + (70 * .67) + (50) = 216.9
Item #2 calc(modified via Block value talent): (1000 * .07) + (50) + (70 * .67) + (77 * (.65 + (.65 * .3)) = 231.965

Now look, 2 items with the exact same blizzard budgeted item value, and item #2 gets a much higher rank (~15 points). By figuring the 30% increase of block value into our occation, we're no longer comparing the item value of an item, we're putting in our own rankings, and saying "I think block value is more important".

Now..if we're making a sheet that's trying to calculate the exact stats we get from an item, because we're trying to figure out what our dodge%, block value, hitpoints, etc. are going to be from a given set of gear. We definetely need these conversions. But for gear comparison, I think we leave this stuff out, otherwise we're introducing artificial preferences on certain stats over others.
ferthadin
 
Posts: 24
Joined: Tue Sep 30, 2008 6:34 am

PreviousNext

Return to WotLK

Who is online

Users browsing this forum: Exabot [Bot] and 1 guest

Who is online

In total there are 2 users online :: 1 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: Exabot [Bot] and 1 guest