Welcome to the Forum Archive!

Years of conversation fill a ton of digital pages, and we've kept all of it accessible to browse or copy over. Whether you're looking for reveal articles for older champions, or the first time that Rammus rolled into an "OK" thread, or anything in between, you can find it here. When you're finished, check out the boards to join in the latest League of Legends discussions.

GO TO BOARDS


New Aura Items: An Analysis by the Numbers (Long)

12
Comment below rating threshold, click here to show it.

Constrictor

Recruiter

02-02-2012

First off, a bit of background on how I come up with gold values for stats and make other calculations:

Stat values with base shop items are easily calculated: I take the total cost of the items and divide by the stats they provide.

Example: Health is (475 Ruby + 1110 Belt) / (180 Ruby HP + 430 Belt HP) = 2.6 gold value per point of health on average.

Stat values without base shop items such as cooldown reduction are derived using rune values. This is more complicated but it is calculated using expected values of quints.

First an average expected quint value is calculated using average flat quint gold values provided across all base shop stat values provided. Next, we divide the expected quint value by the amount of stat that is provided.

Stat values for unique abilities such as actives that can not be easily quantified I just generally price at their combine cost. Example: I add 425 gold value to a complete Phage to account for the new slow on hit proc chance.

Once we have the gold values of stats, we can compute what an item is actually worth. What is the gold leverage you are getting out of a completed item? Typically for a solo item with no gimmicks such as actives or unique abilities, the gold efficiency is somewhere around 1.3.

To put it another way, once you complete an item with a recipe, you are usually getting 30% bonus of stats.

Most items adhere loosely to this rule. If an item has some other benefit such as an aura or acquiring stats or gold over time, the base efficiency is usually below the 1.3 baseline and above the 1.3 baseline once it is stacked or the aura affects multiple targets or gold over time is acquired.

Now, on to the analysis of our aura items.

Strictly numbers:

Aegis of the Legion: 1.09 (solo), 1.49 (duo), 2.61 (5 man)
Soul Shroud: 1.11 (solo), 1.63 (duo), 3.19 (5 man)
Will of the Ancients: 1.14 (solo), 1.8 (duo), 3.76 (5 man)
OLD Starks: 1.23 (solo), 2.24 (duo)
OLD Emblem: 1.17 (solo), 1.54 (duo)

NEW Emblem: 0.77 (solo), 1.10 (duo)
NEW Zeke's: 1.11 (solo), 1.57 (duo)
NEW Iron Solari: 0.91 (solo), 1.23 (duo), 2.05 (5 man)


Comments, Notes, Analysis:

Aegis of the Legion: Useful aura stats for entire team both during and after laning phase

Soul Shroud: Aura stats fall off outside lane phase and aura stats are not useful for all teammates, such as people with blue aura or champions that cap their own CDR or do not rely on CDR. This is the reason you do not see this item built very often.

Will of the Ancients: Incredibly useful aura stats. All champions can use AP and Spell Vamp. Even after the nerf, this remains an incredible value.

OLD Starks: Solid item, but realistically only useful for a couple members of the team. Still a great pick, but not OP.

OLD Emblem: Decent item for when you want more lifesteal or are bottom duo lane. Nothing special, but not horrible.

NEW Emblem: Extremely gold inefficient both solo and duo. If you simply MUST use the Iron Solari item, skip this item and go straight from a heart of gold to the completed solari.

NEW Zeke's: Best item of the three new aura items, but it's not as good as old Stark's. Aura provides niche stats which maybe a couple of teammates can make use of. The shift to a support role item means you will probably not benefit much from the aura stats as the item holder. I would advise against building this as a support (get Aegis instead). I would also advise against building this as an AD champion as it is no longer as useful as old Stark's.

NEW Iron Solari: I calculated the gold value of the activated ability at 500 gold value per 2 people making use of the shield. Weak item in my opinion, especially when compared to Aegis. Avoid this item and get Aegis instead. I would only get this item after completing Aegis as a second defensive aura item. This will have the added benefit of making the activated shield work better on your teammates due to the aegis defensive aura.



In summary, the new aura items do not compare well with the gold standard aura items: Aegis of the Legion and Will of the Ancients. They are not gold efficient and should probably be avoided if at all possible.

Iron Solari AFTER Aegis can be useful for a niche build. Zeke's can also be marginal if you have a heavy autoattacker based team for some reason.


Comment below rating threshold, click here to show it.

GlyphTsen

Senior Member

02-02-2012

Are you including the active on Iron Solari in it's cost? I know that would likely be difficult, but if you hadn't, I think that would potentially make it valuable enough to be on par with the old items.


Comment below rating threshold, click here to show it.

FriendlyRaid

Senior Member

02-02-2012

I like this analysis.

People can quibble about some of your numbers and values, but regardless of your numerical baseline this pretty much confirms what people first thought of the new items, they're just not as good as the old ones and probably not worth buying.


Comment below rating threshold, click here to show it.

Constrictor

Recruiter

02-03-2012

Quote:
GlyphTsen:
Are you including the active on Iron Solari in it's cost? I know that would likely be difficult, but if you hadn't, I think that would potentially make it valuable enough to be on par with the old items.


Yes. I gave it a value of 250 gold per person affected by the shield.

If you get the shield up on all 5 people and you're level 18 and the shield absorbs its full value per person, it's worth more than the 1250 gold I assigned it, but I wanted to give it an 'average' value - considering it is not worth as much if you are lower level than 18 and if part of the shields are wasted (not used in the 5 seconds they are up).


Comment below rating threshold, click here to show it.

axesandspears

Senior Member

02-03-2012

@Constrictor: I don't know what formulas you were using to calculate cost-efficiency for the item's bearer, but you have Will of the Ancients wrong, so I am guessing that most of your values are off. Soulshroud is definitely off, but I'm not going to bother doing the math on that one again. Soulshroud is worth less than 100 gold more than its cost, lets leave it at that.

You should be using the largest block items, not the smallest as those are overpriced to make stacking them for an early boost before building major items a less viable strategy.

Will of the Ancients:
Ap (needlessly large rod): 1600/80*80=1600 gold value.
Spell Vamp (hextech revolver upgrade): 330/15*20=440 gold value. Previously was 550 gold value.
Total Gold Value: 2040 gold value.
Total Cost: 2100 gold.
Gold efficiency for the bearer himself/herself: 97.14%

With double-stack aura:
ap: 1600/80*110=2200 gold value.
spell vamp: 330/15*40=880 gold value.
Total Gold Value: 3080 gold.
Gold cost: still 2100.
Gold Efficiency for bearer himself/herself with double-stack aura: 146.67%.

This is why aura-stacking is broken btw.

I am assuming that you used the smaller block items in this case for calculations. Which is why all of the items are comically more valuable than they should be. This suggests that any item you evaluated as less than cost-efficient is much lower cost-efficiency than your calculations suggest.


Comment below rating threshold, click here to show it.

Constrictor

Recruiter

02-03-2012

If you read the post I explain how I obtain gold values. For stats with multiple base items, I average it out.

I don't want to bias the value based on one specific base item. I feel it's better to average it out.

Also, spell vamp is 'valued' the same as life steal (check the new quints (and masteries) that were released, it's 2% for either stat), which means Riot values them the same.

Since we know the value of lifesteal is 37.5g per point, we know spell vamp is 37.5g per point.


Comment below rating threshold, click here to show it.

Rysan Marquise

Senior Member

02-03-2012

I disagree with the specifics of the gold values determined, especially for spell-vamp (which I think should be valued at about 66% of lifesteal), but the comparisons are still close enough.

I also believe that aura items should generally be valued at aurax2, but you included that. Given those metrics, the new aura items are replacing a niche and very powerful item with a set of not nich, pretty weak items.

It really seems quite unfortunate, but its what we have.


Comment below rating threshold, click here to show it.

axesandspears

Senior Member

02-03-2012

Quote:
Constrictor:
If you read the post I explain how I obtain gold values. For stats with multiple base items, I average it out.

I don't want to bias the value based on one specific base item. I feel it's better to average it out.

Also, spell vamp is 'valued' the same as life steal (check the new quints (and masteries) that were released, it's 2% for either stat), which means Riot values them the same.

Since we know the value of lifesteal is 37.5g per point, we know spell vamp is 37.5g per point.


Ap Quintessances provide more ap than Ad Quintessances provide ad.
Ad Marks provide more ad than Ap Markr provide ap.
Does this not prove that runes are not all designed to provide an equavalent value of stats?

Lifesteal and Spell Vamp do not have the same value, and you do not pay the same amount for those stats in game. This is why these calculations based on the items themselves are used instead of trying to calculate a value based off of runes.

Furthermore, the smaller block items are specifically designed to be overpriced because otherwise stacking 5 Longswords would give you more ad than a B.F. Sword for roughly the same price and people would just build Longsword items because they can get the benefits immediately and do not save up for a B.F. Sword.

When you use the largest base items to calculate gold efficiency and calculate the non-block stats based on the most basic items that provide those stats you get a much more accurate picture of what those stats are worth. Furthermore, when calculated by the largest block items (which give you the smallest gold/stat values), about half of all build items are evaluated as worth their cost and the other half are evaluated as worth less than their cost. Using averages or the smallest block items causes most items to be over-valued, and it becomes more difficult to compare the cost-efficiency of items.

Since using the largest block items for calculations gives you the lowest cost-efficiency calculation and half of the items are on either side of the value that you get. It can be concluded that this gives the most accurate and reasonable calculation of each item's gold efficiency.

Using averages and smaller block items to calculate cost-efficiency would ultimately suggest that most items in the game need nerfs due to excessive cost-efficiency that are not needed in reality.

What I am basically telling you is that you unintentionally biased your efficiency calculations by averaging out the stats based on the block items and by assuming that Spell Vamp and Lifesteal have the exact same value when their very mechanics would suggest that they do not.

Btw, the largest block item method of calculation would prove your point that the new aura items are not really worth their cost more clearly than your own method did.

Quote:
Rysan Marquise:
I disagree with the specifics of the gold values determined, especially for spell-vamp (which I think should be valued at about 66% of lifesteal), but the comparisons are still close enough.

I also believe that aura items should generally be valued at aurax2, but you included that. Given those metrics, the new aura items are replacing a niche and very powerful item with a set of not nich, pretty weak items.

It really seems quite unfortunate, but its what we have.


Using the Hextech Revolver calculation, Spell Vamp is worth 22 gold per 1% spell vamp. Which is roughly the 66% value that you are looking for.


Comment below rating threshold, click here to show it.

Constrictor

Recruiter

02-06-2012

Quote:
axesandspears:
Ap Quintessances provide more ap than Ad Quintessances provide ad.
Ad Marks provide more ad than Ap Markr provide ap.
Does this not prove that runes are not all designed to provide an equavalent value of stats?


That's because AP costs less gold per point. Flat AP quints are the most cost effective in the game at a value of just over 100 gold per quint (average is at 79 gold per quint), but they are within a reasonable range. AD quint gives a value of 85.5, which is also above average. I really don't see your point here.

Quote:
axesandspears:

Lifesteal and Spell Vamp do not have the same value, and you do not pay the same amount for those stats in game. This is why these calculations based on the items themselves are used instead of trying to calculate a value based off of runes.


They do have the same value according to Riot. If you feel like spell vamp is over or undercosted by riot, then you should take advantage of it by either stacking the stat or avoiding items that have the spell vamp stat. Riot values spell vamp at the same value as lifesteal when they build items and provide the stat in runes/masteries. That much is clear.



Quote:
axesandspears:

Furthermore, the smaller block items are specifically designed to be overpriced because otherwise stacking 5 Longswords would give you more ad than a B.F. Sword for roughly the same price and people would just build Longsword items because they can get the benefits immediately and do not save up for a B.F. Sword.

When you use the largest base items to calculate gold efficiency and calculate the non-block stats based on the most basic items that provide those stats you get a much more accurate picture of what those stats are worth. Furthermore, when calculated by the largest block items (which give you the smallest gold/stat values), about half of all build items are evaluated as worth their cost and the other half are evaluated as worth less than their cost. Using averages or the smallest block items causes most items to be over-valued, and it becomes more difficult to compare the cost-efficiency of items.

Since using the largest block items for calculations gives you the lowest cost-efficiency calculation and half of the items are on either side of the value that you get. It can be concluded that this gives the most accurate and reasonable calculation of each item's gold efficiency.


Really, I think you're making too much of this issue. Using the average cost per point is an attempt to avoid bias towards or away from early game items. The difference is also fairly small. I would rather not color my item purchasing decisions too much towards big stat items or away from them.

Quote:
axesandspears:

Using averages and smaller block items to calculate cost-efficiency would ultimately suggest that most items in the game need nerfs due to excessive cost-efficiency that are not needed in reality.

What I am basically telling you is that you unintentionally biased your efficiency calculations by averaging out the stats based on the block items and by assuming that Spell Vamp and Lifesteal have the exact same value when their very mechanics would suggest that they do not.

Btw, the largest block item method of calculation would prove your point that the new aura items are not really worth their cost more clearly than your own method did.


I didn't do the analysis to prove a point. I did the analysis to figure out where these items stand and came to my conclusion based on the analysis.

Quote:
axesandspears:

Using the Hextech Revolver calculation, Spell Vamp is worth 22 gold per 1% spell vamp. Which is roughly the 66% value that you are looking for.


Using Hextech Revolver's combine cost to derive the cost of spell vamp and assuming a perfect 1.0 gold efficiency rating is flawed. You get a gold efficiency bump for completing items. This is easily seen when you look at some lower level items that provide only stats. Therefore, assuming 330 gold is supposed to buy 15% spell vamp across the board is wrong.

Using a derived value for spell vamp based on the average value across known stat values I get a 39 gold per point cost. Using the implied cost based on comparing spell vamp to lifesteal I get a 37.5 gold per point cost. Both of these values are nearly double your assumed cost, and I feel are much more accurate to use.

Using my derived gold value, I place hextech revolver at a 1.19 gold efficiency ratio, which is pretty much in line with other early game simple stat items such as zeal and fiendish codex.


Comment below rating threshold, click here to show it.

axesandspears

Senior Member

02-06-2012

Quote:
Constrictor:
That's because AP costs less gold per point. Flat AP quints are the most cost effective in the game at a value of just over 100 gold per quint (average is at 79 gold per quint), but they are within a reasonable range. AD quint gives a value of 85.5, which is also above average. I really don't see your point here.


That's because you were too stupid to actually look at what I wrote.
For Quints: AP>AD
For Marks: AD>AP
So obviously runes do not have any correlation to the value of the stats they represent. They are just free bonuses that you can use to help you in game


Quote:
Constrictor:
They do have the same value according to Riot. If you feel like spell vamp is over or undercosted by riot, then you should take advantage of it by either stacking the stat or avoiding items that have the spell vamp stat. Riot values spell vamp at the same value as lifesteal when they build items and provide the stat in runes/masteries. That much is clear.



Show me where a Riot staff member has actually said this. It appears that you pulled your assumption that Lifesteal==Spell Vamp out of your ass based on your own earlier flawed reasoning.


Quote:
Constrictor:
Really, I think you're making too much of this issue. Using the average cost per point is an attempt to avoid bias towards or away from early game items. The difference is also fairly small. I would rather not color my item purchasing decisions too much towards big stat items or away from them.


Wriggle's Lantern for example using the largest block items (the ones that give the LOWEST VALUE FOR CALCULATIONS) is worth 1872.5 gold value for a 1600 gold costing item.

Executioner's Calling is 1366.6666 gold value (1350 actual cost) using the same large block formula.

Using the smaller block items or even an average of the block item values to calculate stats is completely idiotic because by those calculations, just about every item in the game appears to be significantly underpriced unless has an extremely poor cost-efficiency rating.

As I just showed you above, small items are about average or unusually cost-efficient using the largest block method as well, so clearly this method gives you a more accurate picture of the cost-efficiency for all items than either of the methods that you used.

Now using the largest block method, the smaller block items appear less cost-efficient for their stats than standard items, but that is because they are less cost-efficient than almost every other item.

Quote:
Constrictor:
I didn't do the analysis to prove a point. I did the analysis to figure out where these items stand and came to my conclusion based on the analysis.



Using Hextech Revolver's combine cost to derive the cost of spell vamp and assuming a perfect 1.0 gold efficiency rating is flawed. You get a gold efficiency bump for completing items. This is easily seen when you look at some lower level items that provide only stats. Therefore, assuming 330 gold is supposed to buy 15% spell vamp across the board is wrong.

Using a derived value for spell vamp based on the average value across known stat values I get a 39 gold per point cost. Using the implied cost based on comparing spell vamp to lifesteal I get a 37.5 gold per point cost. Both of these values are nearly double your assumed cost, and I feel are much more accurate to use.

Using my derived gold value, I place hextech revolver at a 1.19 gold efficiency ratio, which is pretty much in line with other early game simple stat items such as zeal and fiendish codex.


Usually the combine cost from Hextech Revolver is used because it is the closest thing to a block item and only the spell vamp comes from the combine. This gives the 22 gold/1% spell vamp ratio.

If you evaluate the spell vamp from Hextech Revolver's total cost:
1200-(1600/80*40)=1200-(800)=400
400/15=26.6666 gold/1% spell vamp.

So even if the commonly used value is an under-estimate, it is not off by nearly as much as you claim. Furthermore, using the 22 gold/1% spell vamp ratio, every spell vamp item except Hextech Revolver is still cost-efficient, so there isn't an issue with using it.


12