Riftshards vs. Savage Mace - Mathcraft

[owner_strike]

edxs' Graph (http://forums.heroesofnewerth.com/showpost.php?p=1561068&postcount=19)

For damage normalized by gold costs, see Macrohard's Graph (http://forums.heroesofnewerth.com/showpost.php?p=1577158&postcount=51)

[Rilian]

It all comes down to your attack damage without the items.

Late game Riftshards will normally out-DPS a Savage Mace.
Before then, when your DPS is low-medium, Savage Mace will be better...

The Math.

Riftshards (MAX) - +75 damage, 10% to proc 2.4x Critical Strike
Savage Mace - +80 damage, 35% chance to deal an additional 100 damage.

Let's look at 100 base damage

Potential damage on a single strike:

Riftshards - (100 + 75) x 2.4 = 420 Damage
Savage Mace - 100 + 80 + 100 = 280 Damage

Factoring in proc'ing chance:

Riftshards - (100 + 75) + (100 + 75) x .24 = 217 average damage per hit.
Savage Mace - (100 + 80) + 100 x .35 = 215 average damage per hit.

Conclusion
As you can see, even with just 100 base attack, your average output with Riftshards will be slightly higher then with a Savage Mace.
Remember though, this is theoretical damage, based off the assumption that the items proc evenly (eg. Savage Mace will proc 35/100 times), which, in practice, does not happen.

[tuete]

i think riftshards is better at any time on full level

[`patrick`]

It all comes down to your attack damage without the items.

Late game Riftshards will normally out-DPS a Savage Mace.
Before then, when your DPS is low-medium, Savage Mace will be better...

The Math.

Riftshards (MAX) - +75 damage, 10% to proc 2.4x Critical Strike
Savage Mace - +80 damage, 35% chance to deal an additional 100 damage.

Let's look at 100 base damage

Potential damage on a single strike:

Riftshards - (100 + 75) x 2.4 = 420 Damage
Savage Mace - 100 + 80 + 100 = 280 Damage

Factoring in proc'ing chance:

Riftshards - (100 + 75) + (100 + 75) x .24 = 217 average damage per hit.
Savage Mace - (100 + 80) + 100 x .35 = 215 average damage per hit.

Conclusion
As you can see, even with just 100 base attack, your average output with Riftshards will be slightly higher then with a Savage Mace.
Remember though, this is theoretical damage, based off the assumption that the items proc evenly (eg. Savage Mace will proc 35/100 times), which, in practice, does not happen.
Your math is flawed.

You ignored armor, true strike and the +15 attack speed.
Which is obviously very disadvantageous for savage mace.
Also, savage mace is +88 damage not +80.

[Rilian]

i think riftshards is better at any time on full level

Considering most heroes have over 100 base attack by Lv 25, theoretically it's true.


Your math is flawed.

You ignored armor, true strike and the +15 attack speed.
Which is obviously very disadvantageous for savage mace.

Armor is assumed to be zero, and all attacks are assumed to hit. Better?
Also, the attack speed is almost negligible, especially when you consider that as base damage increases, as does the damage gap.
As I glazed over at the start of my post, when your damage is high enough, Riftshards wins outright. All due to the fact that Savage Mace's proc adds a static 100 damage to your total damage output, while Riftshards takes your total damage output and multiplies it by x2.4. Which means the more damage items you have already means the wider the damage gap.

[`patrick`]

In the best case +15 attack speed is +15% dps - that's not negligible.

All the things you ignore are what makes a savage mace good - the point at which riftshards takes over is much higher than 100 base damage.

(assuming average 25% armor reduction, which is much more likely than 0%, at the point you get a savage mace/riftshards; 1 attack/second)

(188*0.75 + 100*0.35)*1.15 = 202.4 DPS
[(100 + 75) + (100 + 75) * 0.24]*0.75 = 164.75 DPS

=> Without any other attack speed increasing items, at 25% average armor reduction, ignoring the true strike savage mace is better than riftshards 5, in terms of dps, until you have a base damage of >500.

I think thats a very different result than 100 :(

[yyr_]

When calculating damage from Savage Mace you assume it gives +123 dmg and 15 attack speed. Crit is amazing in that crit is actually a percentage increase to damage which can be worked out by proc chance*crit increase, so 1.4*20%=28% damage increase. So from here its pretty easy

X is your damage before you buy either.

(X+123)*1.075=(X+75)*1.28

So Savagemace is better if you have below 177 or so damage before your purchase either of them. (So most of the time)

I multiply by 1.075 as thats a rough estimate of how much it will increase your damage purely off your base agilities + attackspeed. I really cant be bother to complicate this equation by taking agility gains into account, so dont rage at this fudge factor. I have worked it out for a majority of heroes before and this is a fair estimate but by no means perfect.

EDITED: 1.24 to 1.28 making finally answer 177 pre-purchase.

[yyr_]

I am pretty sure Malle (or for me, Maide) made a graph for riftshards vs shieldbreaker which was pretty cool as well. Ill probably end up doing a full post on all DPS items later.

[F0Sh1ZZL3]

Hammerstorm with EP and lvl4 Riftshards with galvenize and ulti up is a fun thing to do! =D

[Futte]

Hammerstorm with EP and lvl4 Riftshards with galvenize and ulti up is a fun thing to do! =D

yeah, then riftshards would be the best item. but when you're the new FA, then go split-shot lifesteal and mace.

It's really nice that you guys try to break down the numbers, but people should have a feeling for when they should go mace or riftshards.

Normally i buy Mace and Geo on agi carries, to save the modifier for lifesteal.

Riftshards is if, i have had a hard game, cus it's much more easy to assemble. Or if i'm melee without crit, because it's needed from time to time.

[ElementUser]

Crit is amazing in that crit is actually a percentage increase to damage which can be worked out by proc chance*crit increase, so 2.4*10%=24% damage increase.

It's 28% for level 4 Riftshards. 20% to do 2.4x damage, not 10%. Also, the formula is wrong like Malle informed me when I made the mistake a long time ago.


Your formula is not correct; consider the scenario where you have 100% chance to crit with a crit multiplier of 1 (i.e. you always do your normal damage). Your formula gives the output 2, which is obviously false.

A correct formula is
AvgDmgIncrease = CritProb * CritBonuswhere CritBonus = CritMultiplier - 1

Thus, with Riftshards 4 you have an average damage increase of .2 * 1.4 = .28 which is equivalent to a damage factor of 1.28

For the above mentioned test scenario we then have an average damage increase of 1 * 0 = 0, or a damage multiplier of 1, which is correct.

EDIT: If you want to view it as ElementUser did, you have to take into account how often you do your normal damage as well: 20% to crit for 2.4 times the damage means you have 80% to do your normal damage. Sum these up and you have a damage factor of 1.28:
.2 * 2.4 + .8 * 1 = 1.28

[Rilian]

Looks like I was slow and the big boys finished the math...

[ElementUser]

Wow I just noticed, all of you listed wrong numbers for Riftshards.

It's 20% chance (not 10%.......) to do 2.4x damage at level 4, resulting in an average DPS increase of 1.28x or 28%.

[Rilian]

I went off the tooltip for the item on the site... it doesn't say anything about 20%...

So much outdated stuff on the main site...

[Tyrando]

ye dont trust the site, Madman still has his old skill lol

[Puzzles1]

+100 damage is true damage, Savage Mace makes you not miss your attacks, the interrupt can stop channels AND it goes through BKB.

[ElementUser]

+100 damage is true damage, Savage Mace makes you not miss your attacks, the interrupt can stop channels AND it goes through BKB.

The damage is physical:


<item
name="Item_Weapon3"

icon="icon.tga"

cost="0"
components="Item_Halberd Item_Halberd Item_Voulge"
autoassemble="true"
allowdisassemble="true"

actiontype="passive"

damage="88"
attackspeed="0.15"
truestrike="true"
>
<onattackimpact>
<chance threshold="0.35">
<condition test="not_target_type building">
<damage effecttype="Physical" amount="100" />
<popup name="bonus_damage" a="result" />
<applystate name="State_Stunned" duration="100"/>
</condition>
</chance>
</onattackimpact>
</item>

[CheshireKat]

situational really, if they have an agi carry that's likely to hit up a wingbow savage will be better dps wise even with 2342343 doombringers as the evade will block to much damage.

Mostly riftshards is luxury except for a few people like darklady.

[edxs]

Ignoring Savage Mace's secondary effects, it is always superior when attack damage is less than 96.429.

Otherwise, the region north of this curve is where level four Riftshards takes over:
http://i41.tinypic.com/121xrhy.png

Equation, if anyone needs exact numbers:
http://i43.tinypic.com/30m50tc.png

[ElementUser]

Care to explain the derivation of the equation? :)

(Particularly the fractions)

[edxs]

Care to explain the derivation of the equation? :)

(Particularly the fractions)

It's just the curve where DPS with riftshards minus DPS with savage mace is zero, given the input variables damage and attack speed.

The fractions come from dividing out the 0.28 from riftshards, though I suppose the fractions could be simplified to integer coefficients. I didn't bother. It's nicer to know where the asymptote is.

[Johnson]

In the best case +15 attack speed is +15% dps - that's not negligible.

All the things you ignore are what makes a savage mace good - the point at which riftshards takes over is much higher than 100 base damage.

(assuming average 25% armor reduction, which is much more likely than 0%, at the point you get a savage mace/riftshards; 1 attack/second)

(188*0.75 + 100*0.35)*1.15 = 202.4 DPS
[(100 + 75) + (100 + 75) * 0.24]*0.75 = 164.75 DPS

=> Without any other attack speed increasing items, at 25% average armor reduction, ignoring the true strike savage mace is better than riftshards 5, in terms of dps, until you have a base damage of >500.

I think thats a very different result than 100 :(

1. In general, +15 AS is not equal to +15% DPS

2a. ~175 AS is 1 attack per second with a 1.7 sec attack CD, and +15 AS brings you to 10/9 attacks per second
[ (188*0.65 + 288*0.35) ]*1.111 = 247.78 DPS
[ (175*0.80 + (175+175*2.4)*0.20) ]*1 = 259 DPS

2b. Your math was wrong, riftshards is better for the given conditions (e.g. 1 attack per second and 100 dmg without the item in question)

3. Your attack speed and base damage matter for the calculation of DPS

4. For average damage per attack:
[ (x+88)*0.65 + (x+188)*0.35 ] = x + 123
[ (x+75)*0.80 + ((x+75)+2.4*(x+75))*0.20 ] = 1.28x + 111

x = 43 damage without the item in question before riftshards is better for the given condition

Conclusion -
Riftshards at max level is essentially always better than savage mace when calculating average damage per attack.

[yyr_]

Wow I just noticed, all of you listed wrong numbers for Riftshards.

The heroes page has the numbers wrong and we are not very good at things.

Thanks for clearing up the formula btw but i now need to change my post on engi's turret qqq.

[yyr_]

[ (x+75)*0.80 + ((x+75)+2.4*(x+75))*0.20 ] = 1.48x + 111

x = 25 damage without the item in question before riftshards is better for the given condition

Conclusion -
Riftshards at max level is essentially always better than savage mace when calculating average damage per attack.

You made the same mistake i did. That number i bolded should be 1.4. So your conclusion is incorrect. Check Elementalusers post if you want to see why.

[ElementUser]

Yeah that or just replace ((x+75)+2.4*(x+75))*0.20 with 2.4*(x+75)*0.20.

It's also more simplified and easier to see.

[Johnson]

You made the same mistake i did. That number i bolded should be 1.4. So your conclusion is incorrect. Check Elementalusers post if you want to see why.

http://img132.imageshack.us/img132/2252/riftshards.jpg

Shockingly, I read the whole thread before posting and verified my numbers in game before doing the math...
I don't think I have any mods installed.



[ (x+75)*0.80 + ((x+75)+2.4*(x+75))*0.20 ] = 1.28x + 111


In the " ((x+75)+2.4*(x+75))*0.20 " term:
(x+75) is your damage with riftshards on a non-critting hit
2.4*(x+2.4) is the additional damage from the crit
0.20 is the average chance of occurance

So, the issue may be that I am misunderstanding what a critical strike means in this game.
Is the damage from a crit -
a. regular attack damage + (crit multiplier)(reg attack dmg)
b. (crit multiplier)(reg attack damage)

?

[yyr_]

Element can you move this thread on Engineers Turret weapons into mechanics with the other weapon threads please :D Link (http://forums.heroesofnewerth.com/showthread.php?t=115533) <3

[yyr_]

Shockingly, I read the whole thread before posting and verified my numbers in game before doing the math...
I don't think I have any mods installed.

You multiplied by had a 0.2*(damage+(damagex2.4))

What you calculated there is actually 3.4x crit as you can move the damage on the left hand side into the "damagex2.4" as they are technically like terms. Making it 3.4damag*20% chance. The 2.4 already compensates for your original damage. Make sense? im not very clear when trying to type in the morning.

[Johnson]

4. For average damage per attack:
[ (x+88)*0.65 + (x+188)*0.35 ] = x + 123
[ (x+75)*0.80 + ((x+75)+2.4*(x+75))*0.20 ] = 1.28x + 111

x = 43 damage without the item in question before riftshards is better for the given condition

Conclusion -
Riftshards at max level is essentially always better than savage mace when calculating average damage per attack.

If b. is correct above, then..

[ (x+75)*0.80 + ((x+75)+1.4*(x+75))*0.20 ] = 1.28x + 96
or
[ (x+75)*0.80 + 2.4*(x+75)*0.20 ] = 1.28x + 96

x = 96.43 damage without the item in question before riftshards is better for the given condition

Agrees with

Ignoring Savage Mace's secondary effects, it is always superior when attack damage is less than 96.429.

[Johnson]

You multiplied by had a 0.2*(damage+(damagex2.4))

What you calculated there is actually 3.4x crit as you can move the damage on the left hand side into the "damagex2.4" as they are technically like terms. Making it 3.4damag*20% chance. The 2.4 already compensates for your original damage. Make sense? im not very clear when trying to type in the morning.

I edited my post because this occurred to me. So crits in HoN are option b then?



In the " ((x+75)+2.4*(x+75))*0.20 " term:
(x+75) is your damage with riftshards on a non-critting hit
2.4*(x+2.4) is the additional damage from the crit
0.20 is the average chance of occurance

So, the issue may be that I am misunderstanding what a critical strike means in this game.
Is the damage from a crit -
a. regular attack damage + (crit multiplier)(reg attack dmg)
b. (crit multiplier)(reg attack damage)

[roarp]

Go on with the maths, but now try to think 2 items in a more practical sense. What makes people buy savage Mace > riftshaft is the mini-stun and the truestrike it provides, the 88 damage (and 100 true damage proc) is just a nice bonus for all the money you have spent. Buy mace when people start getting evasion items (wingbow, snake bracelet) or with evasion skills (ie. zephyr). Straightforward.

I have no idea what you guys arguing about. If you want pure damage whn enemy has no evasion, no channels, get rift, you obviously will have >100 base damage by the time you make rift 4 (i hope you are not getting it for your first item lol) and rift4 surely hits harder than mace in terms of dps by that time. If they have wingbow, well....good luck with rift4 for the insane high evasion rate from that item.

[ElementUser]

Element can you move this thread on Engineers Turret weapons into mechanics with the other weapon threads please :D Link (http://forums.heroesofnewerth.com/showthread.php?t=115533) <3

I can't, it's in Strategy. You'll have to report that thread to be moved to Mechanics if you want.


If b. is correct above, then..

[ (x+75)*0.80 + ((x+75)+1.4*(x+75))*0.20 ] = 1.48x + 96
or
[ (x+75)*0.80 + 2.4*(x+75)*0.20 ] = 1.48x + 96

Yes that looks better

[edxs]

The coefficient of pre-riftshards damage (written 'x' most everywhere here) is 1.28, not 1.48. The 96 constant damage is correct.

[Johnson]

The coefficient of pre-riftshards damage (written 'x' most everywhere here) is 1.28, not 1.48. The 96 constant damage is correct.

DAMN
I was wondering how you and I got different numbers. I need to stop trying to do math after work....
Thanks for the catch.

[yyr_]

The coefficient of pre-riftshards damage (written 'x' most everywhere here) is 1.28, not 1.48. The 96 constant damage is correct.

I have been sitting there looking at it trying to work out what went wrong, i was worried i was going crazy, thanks for confirming :D

PS: I have become bad at things all of sudden and cant find the mistake.

[Johnson]

Rofl.
edxs, yrr, and I (and ElementUser) are such nerds

[yyr_]

Rofl.
edxs, yrr, and I (and ElementalUser) are such nerds

Fixed :D that we are, that we are.

[Urizen]

Target armour is also a factor as depending on the disparity between the bonus damage on crits and the 100 true damage from savage mace, factoring in attack speed and truestrike things can change

[edxs]

Savage Mace doesn't deal any true damage.

[yyr_]

Target armour is also a factor as depending on the disparity between the bonus damage on crits and the 100 true damage from savage mace, factoring in attack speed and truestrike things can change

Not true damage. We ignore true strike because if they have evasion you dont need to math it, just by savage.. Multiply savagemaces damage by about 1.07ish and your done.

[FlamerFuel]

...based off the assumption that the items proc evenly (eg. Savage Mace will proc 35/100 times), which, in practice, does not happen.


Why will this not happen?

[Nedrapter]

Why will this not happen?


For the same reason that some ppl never win the lottery even if they played more than enough times, statistically.


On topic:
im not so good at maths and it's kinda tiring to follow up
the equations, so so far what's the damage point where Riftshards takes over?

[SiC]

Why will this not happen?

so you get a six every 6th time you roll the dice?^^
it's the thing with statistic, it will become true with great numbers, but you cannot rely on the % with the though of "I didn't get a procc the last 10 attacks, so now it should procc!"
for more see Wiki: Law of Large Numbers (http://en.wikipedia.org/wiki/Law_of_large_numbers)

btw: it's different in DotA



im not so good at maths and it's kinda tiring to follow up
the equations, so so far what's the damage point where Riftshards takes over?

afaik it's this post (http://forums.heroesofnewerth.com/showpost.php?p=1563027&postcount=29), so it's around 96-97 dmg

[edxs]

The practical cutoff is between 150-250 depending on attack speed.

[oberbreit]

Very interresting thread

I think Truestrike (100% no miss) and the fact Savage Mace can interrupt channeling spells should be mentioned too

Riftshards have "only" the dps boost

[PROc]

Very interresting thread

I think Truestrike (100% no miss) and the fact Savage Mace can interrupt channeling spells should be mentioned too

Riftshards have "only" the dps boost

Yea however riftshards can provide high burst damage rather than just sustained damage which can be very usefull when taking out a high priority target.

[yyr_]

The practical cutoff is between 150-250 depending on attack speed.

^This is pretty much the range. I would use 177 damage before you purchase either as the general rule though. 177 damage assumes you have +100 attackspeed before you purchase. For example, 75 agility, +25 from steamboots. Or 45 agility and a warpcleft (+55).

[thened]

This is both informative and useful. Thanks!

Ps. I Love you guys, I tried to keep up with the math but you lost me.

[yyr_]

':D'

[Targuil]

Don't forget the fact that savage mace is way harder to farm: Every part in savage mace costs more than any part in riftshards.

[MacroHard]

Introduction:


I'm tired of people postulating suggestions or ideas based on "feel", or half-ass mathematical analysis based on poor assumptions. There have been threads about Riftshards vs. Savage Mace in the past. Some people neglected speed factors. Nearly everyone failed to recognize that this game has an economy, so damage needs to be normalized by gold costs.

I will not make ANY assumptions. My calculations below factor in attack speed as well as the gold cost of the items (even though 5500 vs 5400 is close to negligible).

NOTE: ARMOR HAS NO EFFECT SINCE THE PROC FROM BOTH RIFTSHARDS AND SAVAGE MACE ARE BOTH PHYSICAL


The Acronyms and Variables:


SM = Savage Mace
SMd = Added damage output, including all factors, that SM adds
SMv = Value of SM = SMd / 5400
RS = Riftshards
RSd = Added damage output, including all factors, that RS adds
RSv = Value of RS = RSd / 5500
D = Damage BEFORE purchasing any of the items or components
S = Attack speed bonus BEFORE purchasing any of the items or components
(includes everything... so if you have 80 agility and +30 attack speed worth of items, S = 110)


The Basics:


SMd = 88+.35*100+{15*[(D+88)+35]}/(100+S)
SMv = 88+.35*100+{15*[(D+88)+35]}/(100+S)/5400
RSd = 75+(D+75)(.2)(2.4-1) = (.28D+96)
RSv = 75+(D+75)(.2)(2.4-1)/5500 = (.28D+96)/5500


The Method:


To solve the problem of which is better, we have to find the trend line where the values SMv and RSv are equal to each other. Since we have only two variables (D and S), we can simply solve for S in terms of D. The result will be a plot of S vs D.

Basically it boiled down to this:
[indent]1) Set SMv = RSv
2) Solve for S in terms of D


The Results:


http://i39.tinypic.com/s4w4k5.jpg
All points above the curve favor Riftshards. All points below the curve favor Savage Mace.


The Conclusion:


As you can see from the graph, there are very few instances where Riftshards will provide more damage per gold spent. You'd have to hit for 190 at 100% increased attack speed BEFORE purchasing your item. This is a rare situation.

Also, Savage Mace has better intangibles... it provides a channel stopping mini stun, and has better reliability and consistent output (35% procs vs 20%). Finally, Savage Mace makes it so that you NEVER MISS... attacking uphills or against a hero with evade will no longer have an effect.

All this together means that Savage Mace generally gives more additional DPS per gold spent.


Exceptions:


If you have high attack speed or high base damage Riftshards is more viable. For example, if you are Soulstealer, your passive damage boost + damage from Assassins Shroud + 30% ias from steamboots = you may want to get Riftshards as your next major item instead of Savage Mace.


Hope this was helpful! If you have questions or private comments please send me a message and I'll try to get back to you.

[ElementUser]

Merged with the other thread lol

You must have missed it

[MacroHard]

Wow you act fast! I even named my thread the same title lol. Crazy...

[ElementUser]

Wow you act fast! I even named my thread the same title lol. Crazy...

Actually I renamed the thread to that since it looks a lot nicer

Thanks for your contribution though :).

[edxs]

stuff


You seem to believe Savage Mace's 100 damage on proc bypasses damage mitigation, this is false. Armor shouldn't factor into the comparison at all.

Damage per gold analysis is a good idea, though. I should have thought of that.

[MacroHard]

The Savage Mace's proc damage is true damage and bypasses armor and magic armor reduction. The calculation stands unless one of the code reading guru's says otherwise.

[edxs]

The Savage Mace's proc damage is true damage and bypasses armor and magic armor reduction. The calculation stands unless one of the code reading guru's says otherwise.

People constantly bring this up as fact, but it's been refuted countless times. It isn't true, and it hasn't ever been true. It's a separate instance of 100 Physical damage.

[MacroHard]

People constantly bring this up as fact, but it's been refuted countless times. It isn't true, and it hasn't ever been true. It's a separate instance of 100 Physical damage.

The changelog and item description says otherwise. Would a moderator mind stepping in and clarifying the issue?

If it is indeed physical there would be no added benefit against higher armors, and you may use the 0 armor trend line on my graph for comparison.

Later tonight I will do an in game test in practice mode... level 1 Soulstealer with savage mace (~130 damage) attacking a level 25 Pebbles with 6x Frostfield plates (~100 armor) and see if I can see the 100 damage boost.

[edxs]

http://forums.heroesofnewerth.com/showpost.php?p=1560012&postcount=17



Version 0.1.26
--------------

- Savage Axe's triggered bonus damage will no longer trigger on structures
* Fixed a bug causing Savage Axe's triggered effects to deal True damage instead of Physical
I was wrong saying it's never been true.

The description doesn't say anything about True damage.

[ElementUser]

Post #17

http://forums.heroesofnewerth.com/showpost.php?p=1560012&postcount=17

Darn edxs ninja'ed me :(. I am sad now

[Habile]

He ninja'd your OWN post too, at that!

[ElementUser]

At least it's impossible to set his post behind my own post chronologically :)

[hzzzln]

wait savage mace procs physical dmg? intended?

[MacroHard]

Thank you for enlightening me, edxs and ElementUser. I suppose the 0 armor line is the only one that applies. I will modify my post in a few moments.

*EDIT: My previous graph/post has been recalculated and changed completely.

[TheRoyalty]

What it boils down to is that it depends greatly who you buy the item against. Play against NH? Get Savage Mace. Enemy has Wingbow? Same. Etc etc.

[ElementUser]

Well now edxs' graph and Macrohard's graphs look similar lol

edxs' y-axis uses Actual Attack Speed while Macrohard's graph uses Actual Attack Speed - 100 (at least from what I see)

Hm still, something is bugging me about why the graphs look slightly different. I think it's where the vertical asymptote lies that's doing this

[edxs]

I believe he's using the curve where DPS/gold is equal.

edit: Yes, it looks like it. Using equal value moves the asymptote to 111.969. yyr's +100 AS point moves over to 209 pre-item damage.

http://i43.tinypic.com/vg2yog.png

[yyr_]

^ Where did i go wrong in my formula to get 177 at +100?

[edxs]

You didn't go wrong. That's the point at which Riftshards 4 overtakes Savage Mace in DPS (assuming +100 AS).

209 damage is where Riftshards 4 overtakes Savage Mace in DPS per gold (assuming +100 AS).

[yyr_]

You didn't go wrong. That's the point at which Riftshards 4 overtakes Savage Mace in DPS (assuming +100 AS).

209 damage is where Riftshards 4 overtakes Savage Mace in DPS per gold (assuming +100 AS).

Sweet, it was really bugging me that i was wrong, but all good :D cheers edxs.

[MacroHard]

Keep in mind that even though 209 is break point (at +100 AS), Savage Mace has two hugely beneficial extras: mini-stun and truestrike.

Even if I'm at 250 damage I'd still go SM for those two things alone.

Think about it... if you're hitting for 300+ (after buying either RS or SM), do you expect your enemy to stay and fight? Somebody with 2000 health can easily teleport away before you can get the kill. SM provides a beatdown AND makes teleporting an impossibility.

Secondly, in almost every game, there will be some form of evasion that comes up. Truestrike is useful every game. Got a ranged carry? Probably. Hate not getting that last hit kill because you miss your final attack uphill? I know it's happened to you.

Third, SM's proc is far more reliable. 35% vs 20% for that bonus really makes a difference. There are many cases that as a carry you can only attack two or three times between disables (since you are definitely going to be focused). It's important that when you attack you provide reliable and consistent output. There have been cases where I've attacked 6 or 7 times and failed to get a critical. While you can equally get several criticals in a row, the former is much more frustrating than the latter is satisfying.

The point of my post is that SM is not only more efficient in terms of DPS/gold (in most situations), but it's also better in general terms; HON isn't about who can do the most damage... it's a conglomeration of complex nuances.

[MechE_]

Keep in mind that even though 209 is break point (at +100 AS), Savage Mace has two hugely beneficial extras: mini-stun and truestrike.

Even if I'm at 250 damage I'd still go SM for those two things alone.

His post is lengthy, but I agree - Savage Mace has much better intangibles. High level players know this, that's why SM is much more popular in those games and rift shards is much more popular in -em games.

[SpyflueN]

His post is fail. "do you expect your enemy to stay and fight?".. rofl.. heard about stun, slow, bubbles? i guess not.. hon is not 1n1, its 5v5.

[CallofDuty5]

regardless, not every enemy is constantly disabled, and sometimes savage mace might be the only stun currently available

[svedlund]

ok, now to the most important question. if the 100+ dmg from SM is physical, is it included with the dmg output if you crit with crit-passive or crit-item? like
300dmg+100dmg * crit?

or

300dmg * crit + 100dmg?

?

and 2 more questions when Im writing my first post here:

1. Iron buckler, it says it reduces -10 from ranged heroes. But when Im arachna and hit scout with iron buckler HoN write "-20" everytime (got activated show in settings).

2. Barbed armor, does it give attacker (with 100dmg):

a, 80% of attackdmg, like 80%x100=80dmg back?
b, 80% of the dmg taken, like 80%x(100-armor-blacklegion reduction) = almost nothing
c, 80% of attackdmg - attackers armor, llike 80%x100 x )100% - attackers armor-reduction)?

[ElementUser]

It's the first situation.

And:

1) If the wielder is melee he blocks 20; if the wielder of the shield is ranged he blocks 10.

2) 80% of the damage taken (after reductions) by the wielder is redirected to the correct source.

[Habile]

It's the first situation.

And:

1) If the wielder is melee he blocks 20; if the wielder of the shield is ranged he blocks 10.

2) 80% of the damage taken (after reductions) by the wielder is redirected to the correct source.

Are you sure about Savage Mace and critical hits?


300dmg+100dmg * crit?(Attack Damage + 100Bonus) * Critical Multiplier


300dmg * crit + 100dmg?(Attack Damage * Critical Multiplier) + 100Bonus

Unless this changed since closed beta, Savage Mace procs are a completely separate source of damage from your actual attacks.

I'll retest in game, and update this post if I'm mistaken.
edit: It's the second. A flat 100 damage is added on top of the critical.

[ElementUser]

Oh I thought he meant the +damage he got from Savage Mace and not the +100 proc

~.~

Of course the +100 proc isn't affected by any multipliers that affects your attack (this is why illusions proc the full damage). I thought he meant Savage Mace's +88 damage since 88 is fairly close to 100.

The +88 does get amplified and (stressing again) that was what I was referring to - NOT the separate proc

[Habile]

Yeah, I had a feeling it was just a misunderstanding. I always like to retest things anyway, because changes happen.

[Marlesdk]

Puppet whipsplash/ Malikens ulti(splash).. How does that work with savage mace? does it splash the +100 true damage or is it only for one single target?
And ofcourse Runed axe?
Maliken should indeed carry riftshard 4 if its only one target who recieves more dmg when True strike procs!?

[fenrixx]

Puppet whipsplash/ Malikens ulti(splash).. How does that work with savage mace? does it splash the +100 true damage or is it only for one single target?
And ofcourse Runed axe?
Maliken should indeed carry riftshard 4 if its only one target who recieves more dmg when True strike procs!?

It's +100 physical damage, not true.

True strike means that it cannot miss.


Dunno abouy your question though.

[Kawaiiiii]

100+ proc DOES NOT apply to splash damage.
Therefore, Maliken with a Savage Mace and Ultimate enabled only does 100+ proc on the target he is focusing and not the others.


Critical damage DOES apply to splash damage.
Therefore, the extra damage that Maliken does with the Riftshards WILL apply to all the targets that are within range of his AoE.


However, after testing this in-game I found it pretty obvious if you read the descriptions of it.


'On attack impact: 35% chance to deal 100 bonus damage to target.'


10% chance to deal 2.4x bonus damage

[TheVeronicas]

Solution: Buy both

[FirstToStand]

The reason why Savage Mace will always continue to be more viable is because for a mere 1800 gold you can purchase a whopping 25% physical EVASION. Which is in practice far more potent than a flat -25% physical reduction.

[ilikepotatos]

The reason why Savage Mace will always continue to be more viable is because for a mere 1800 gold you can purchase a whopping 25% physical EVASION. Which is in practice far more potent than a flat -25% physical reduction.

I do not really follow this as I have never seen anyone just buy the snake bracelet people do not use it.

[`aNarchy]

edxs' Graph (http://forums.heroesofnewerth.com/showpost.php?p=1561068&postcount=19)

For damage normalized by gold costs, see Macrohard's Graph (http://forums.heroesofnewerth.com/showpost.php?p=1577158&postcount=51)

First of all Macrohard mentioned that soulstealer with shroud should get riftshards as a next major item. This pretty much rules out whatever **** this kid has said.

Second of all it is entirely situational on what you should build(regarding SM or RS) because Riftshards has a much easier build up, and does not need to be upgraded to level 4. You can merely leave your Riftshards at level 1 and build stat items + bkb.

[Tomate]

Double posting this here, I though I was writing here in the first place.

Not factoring the 15% attack speed is forgetting a huge component...

DPS = Damage X number of attacks per second.
DPS rift = (Base Damage + 75) X 1.28 X (attack speed)
DPS Savage = (Base damage + 88 + 35) X (Attack speed + (Base attack speed *.15))

Most interesting carries are at 0.72 (Silouhette) or 0.71 (Flint) base attack speed. Let's take 0.715 as the base attack speed for the math. If we first assume that the attack speed does NOT go up, let's find the break even point.

Between 349 and 350 base damage actually is the number.
Rift DPS = (349 + 75) * 1.28 * .715 = 388.0448
SM DPS = (349+88+35) * (.715 + .107) = 388.102 (Real number since I rounded the .107)
At 350, Rift becomes better.

Now, you will tell me, well, this is bullshit, I won't have 0.715 attack speed at the end of the game. True, so I figured I'd adjust the attack speed. I created 3 extra scenarios. End game attack speed of 1.0, 1.5 attacks per second.

At 1.0 attacks per second, a base damage of between 232 and 233 is there Rift takes over. Formula:
DPS Rift: (232 + 75) * 1.28 * 1 = 392.96
DPS SM: (232 + 88 + 35) X (1+0.107) = 393.07

At 1.5 attacks per second, the scenario of break even is at: 171 and 172 base damage prior to the item.
DPS Rift: (172 + 75) * 1.28 * 1.5 = 474.24
DPS SM: (172 + 88 + 35) * (1.5+0.107) = 474.13

SO, considering a level 17 flint with frostwolf, steamboots, geometer, what should he get? Well, let's check:
Base damage 39 + 16 levels gained * 2.9 agi gain/level + 25 damage from FWS + 26 from Geo = 136 damage. With those items, flint should be around that 1.5 attacks per seconds. As you can tell, for his DPS, he should still go with SM REGARDLESS of true strike.

This is why Rift is often a 2nd pick to Savage mace, factor in the fact that you can interrupt with Savage and have true strike, you have no reason going for rift until the ultra late game.

**EDIT** I tested and with 136 damage, you need 2.5 Attacks per second BEFORE the item to have a breaking point between SM and Rift.

[Tilexh]

Double posting this here, I though I was writing here in the first place.

Not factoring the 15% attack speed is forgetting a huge component...

DPS = Damage X number of attacks per second.
DPS rift = (Base Damage + 75) X 1.28 X (attack speed)
DPS Savage = (Base damage + 88 + 35) X (Attack speed + (Base attack speed *.15))

Most interesting carries are at 0.72 (Silouhette) or 0.71 (Flint) base attack speed. Let's take 0.715 as the base attack speed for the math. If we first assume that the attack speed does NOT go up, let's find the break even point.

Between 349 and 350 base damage actually is the number.
Rift DPS = (349 + 75) * 1.28 * .715 = 388.0448
SM DPS = (349+88+35) * (.715 + .107) = 388.102 (Real number since I rounded the .107)
At 350, Rift becomes better.

Now, you will tell me, well, this is bullshit, I won't have 0.715 attack speed at the end of the game. True, so I figured I'd adjust the attack speed. I created 3 extra scenarios. End game attack speed of 1.0, 1.5 attacks per second.

At 1.0 attacks per second, a base damage of between 232 and 233 is there Rift takes over. Formula:
DPS Rift: (232 + 75) * 1.28 * 1 = 392.96
DPS SM: (232 + 88 + 35) X (1+0.107) = 393.07

At 1.5 attacks per second, the scenario of break even is at: 171 and 172 base damage prior to the item.
DPS Rift: (172 + 75) * 1.28 * 1.5 = 474.24
DPS SM: (172 + 88 + 35) * (1.5+0.107) = 474.13

SO, considering a level 17 flint with frostwolf, steamboots, geometer, what should he get? Well, let's check:
Base damage 39 + 16 levels gained * 2.9 agi gain/level + 25 damage from FWS + 26 from Geo = 136 damage. With those items, flint should be around that 1.5 attacks per seconds. As you can tell, for his DPS, he should still go with SM REGARDLESS of true strike.

This is why Rift is often a 2nd pick to Savage mace, factor in the fact that you can interrupt with Savage and have true strike, you have no reason going for rift until the ultra late game.

**EDIT** I tested and with 136 damage, you need 2.5 Attacks per second BEFORE the item to have a breaking point between SM and Rift.

Remember that geo illusions can crit, but cant get savage proc.

[yourstruly]

This all seems very confusing to me.

Everything else is symmetrical, exept the effect of attack damage itself, which scales for riftshards and stays constant for savage mace. And proc effects from other items or abilities.

You don't need to care about attackspeeds, base attack times, armors or anything else, because they work just the same for both items, only relevant thing is the attack damage.

Thinking pure damage, riftshards is better than savage mace when your autoattacks hit for 350 or more. Other items with proc effects can alter this, but usually do so in the favor of savage mace (since the proc cannot crit, but generally does benefit from attackspeed). Attackspeed, amor reduction, base attack time, all irrelevant.

EDIT: Geo indeed is better for rift than savage

[Lethe]

it's obvious geo is a better fit for rift than savage. Images get crit, but images with mace get nothing.

[GhettoJesus]

Talking about geo + riftshards tho isn't butterfly/eaglehorn or whatever its hon name + geo = better & more reliable dps for illusions by a longshot. As stats actually raise the damage and atk speed of the illusions. If you grab the illusion rune with a stat rather then damage set up you get tankier higher dps illusions. Enough to cause some mayhem. Altho this is a sort of build i'd really only go with MQ or slither. Also isn't a charged hammer a lot more reliable then riftshards once you've actually got some damage(70 atk speed, 28 dmg 20% 200 magic dmg proc). I've found that stacking several charged hammers practise mode can get you to the 3.3333 / 4 atks a second region which results in cheesey damage for sil and flint. But in a real game a blacksmith or empath could give you the extra atk speed to reach that 4/1 sec cap.

[AtheistGod]

Most interesting carries are at 0.72 (Silouhette) or 0.71 (Flint) base attack speed. Let's take 0.715 as the base attack speed for the math. If we first assume that the attack speed does NOT go up, let's find the break even point.

Those numbers aren't what you think they are.

Ignoring frames the comparison for when Riftshards is better is:
(Damage+123)*(Attackspeed+.15) < (Damage+75)*Attackspeed*1.28

For damage =
<96: never
150: 2.73 AS , unlikely unless you are using EP
175: 2.03 AS
200: 1.67 AS , almost always
225: 1.45 AS
250: 1.3 AS , you'll have this much AS by level 6 on an agility hero with no items

For AS =
1.5: 218 Damage, Only really occur if you have a 20 minute Doombringer with no Steam Boots.
1.75: 193 Damage
2: 177 Damage (this is mentioned earlier in the thread)
2.25: 165 Damage
2.5: 156 Damage
3: 144 Damage

Savage Mace works better than Rift Shards as a first item, but by the time you reach into the upper 100s for damage, Rift Shards will give better damage output.

Base Attack Time is irrelevant unless you are taking frames into account.

[Tomate]

Those numbers aren't what you think they are.

Ignoring frames the comparison for when Riftshards is better is:
(Damage+123)*(Attackspeed+.15) < (Damage+75)*Attackspeed*1.24

For damage =
<125: never
150: 6.825 AS, basically never
200: 2.69 AS
250: 1.865 AS
300: 1.51 AS , basically always

Base Attack Time is irrelevant unless you are taking frames into account.

Where the hell do you get your numbers... you do not just add .15 to number of attacks...

If you want the whole formula disregarding any side effects such as: Illusions propagating 1 but not the other 1, salvo effect on total damage and blablabla...

here's the formula:
(Damage + 123) * (Attack speed + base attack speed * .15) < (Damage + 75) * 1.28 * (Attack speed).

Redo your work, your numbers are just way off the chart and make absolutely no sens.

[AtheistGod]

Where the hell do you get your numbers... you do not just add .15 to number of attacks...

If you want the whole formula disregarding any side effects such as: Illusions propagating 1 but not the other 1, salvo effect on total damage and blablabla...

here's the formula:
(Damage + 123) * (Attack speed + base attack speed * .15) < (Damage + 75) * 1.28 * (Attack speed).

Redo your work, your numbers are just way off the chart and make absolutely no sens.

The only thing I messed up was 1.24 instead of 1.28. Base Attack Time is IRRELEVANT if you aren't taking frames into account, and you obviously .aren't.



DPS = Damage * Attackspeed / BAT * crit modifers * damage reductions.
Since BAT is constant regardless of items and enemy damage reductions aren't changed by either Savage Mace or Riftshards we can factor both of those out.

With frames you just change the (Attackspeed/BAT) to 1/(.05*ceiling(20*BAT/Attackspeed)). This complication prevents you from factoring the BAT out.


Those numbers aren't what you think they are.

This is the source of your problems. You are acting like level 1 damage matters. We aren't asking how much damage you need at level 1 with no attackspeed for Rift Shards to beat Savage Mace. I don't care that a level 1 with Doombringer and Savage Mace out damages a level 1 with Doombringer and Riftshards.

You also incorrectly applied the SM attack speed but that is a more minor issue.

Off-topic:
Did you come from LoL? The way you look at it and the error you made seem like they could be due to how LoL displays AS and LoL not having +AS at level 1.

[antics]

.png of effective buying is below
http://i54.tinypic.com/2nhm5oy.png
The shaded area is when it is effective to buy savage, the y axis being attack speed, and the x axis being damage.
Please note that these are the values of attack speed and damage BEFORE buying the item.

Ignore the terribly formed equations in the top of the frame; MSMath simplification blows more dick than a 50cent hooker.

[V83]

i buy both!! when i carry =D

[vipertail67]

the savagemace got a ministun as well. when u hit an enemy in the frame of his attack animation that animation will be canceled and results in a reset of that animation. that would mean incomming dps reduction. something that riftshard doesnt have.