Two Hand Blunt Max Weapon Damage (Monk)

[Treats]

Something is off on P99 when calculating maximum hits for Monk two hand weapon damage. Have not tested any other classes or two hand slashing.

55 Monk (Tranquil Staff 29/30)
245 Offense
153 Strength
16 Damage Bonus

Minimum damage = 17
Magic Number = 74
Maximum hit = 131.42

This is the hit distribution on P99, a lot of hits are over what max damage should be:

http://img594.imageshack.us/img594/3280/monk2hb.png

[Versus]

best graph ever

[kanras]

Max damage multiplier from 1 - 50 is determined by (STR + OFF) / 100 (simplified)
Max damage multiplier from 51+ is determined by (PWR / PWR_CAP) * DMG_TABLE (simplified)

[Treats]

Max damage multiplier from 1 - 50 is determined by (STR + OFF) / 100 (simplified)
Max damage multiplier from 51+ is determined by (PWR / PWR_CAP) * DMG_TABLE (simplified)

I assume PWR = some calculation of attack and strength. If Strenth is factored into this calculation it shouldn't be, but that would be a whole different thread. If you were to test this on a Paladin here on P99 would it look similar to this graph for the max hit with the Paladin's damage table?

http://users.erols.com/aburner/eq/Poly.jpg

54 Paladin using Polyphenomenal Axe (37/48)
171 Strength
210 2HS Weapon Skill
210 Offense (Unsure on Offense cap for Paladin, do not know skill at 54)

This is what I get using (Str + Off) / 100 * Weap Dam + Dmg bonus. Unsure how the minimum damage is 19. It uses a damage bonus of 15 to calculate the Magic Number.

Min = 16 (Unsure how it is 19 on graph)
Magic Number = 89
Maximum hit = 155.97

Using 18 damage bonus (Damage table possibly?) for Min and Max, 15 for Magic Number:

Min = 19
Magic Number = 89
Maximum hit = 158.97

[kanras]

I assume PWR = some calculation of attack and strength.
PWR is the portion of displayed ATK that relates to damage calculation after a hit has successfully landed. The other portion of displayed ATK is ACC which relates to your chance to hit. Yes, STR has an impact on PWR.

Unsure how the minimum damage is 19
If you look at live parses where high damage (15+) weapons are used, you'll see that:
MIN_HIT = DMG_BONUS + MAX(FLOOR((WPN_DMG+5)/10),1)

So for your poly axe, it's 15 + FLOOR((37 + 5)/10) or 15 + 4 = 19

54 Paladin using Polyphenomenal Axe (37/48)
All classes besides war/mnk/rog/rng still use the (STR + OFF) / 100 to calculate their max damage multiplier post-50.

[Treats]

PWR is the portion of displayed ATK that relates to damage calculation after a hit has successfully landed. The other portion of displayed ATK is ACC which relates to your chance to hit. Yes, STR has an impact on PWR.

Thanks alot for posting that, great info. Just gonna post this here about STR affecting attack then.

Here is where the graph of Paladin damage came from, a lot of info here:

http://www.angelfire.com/rpg/whitewind/gtxt/guides/maxdam.htm

Although you cannot see the graphs anymore at the bottom of the page, they talk about strength not affecting hit distribution.

SHAMAN MELEE
Thanks to the logs Arkand provided in the So, Stat Buffs Are Useless? thread, I was able to generate Damage Profiles for Arkand's 52nd level Shaman. They answered several questions I had. Thanks again to Arkand for putting them up.

When I first looked at them, I noticed that they all had huge, prominent peaks at 1. I was puzzled by this...until I saw a line in one of the logs that said "You bash a seafury cyclops for 1 point of damage!"

As a Shaman, Arkand's Slam skill will never increase. And it will never do more than 1 point of damage. Once I figured THAT out, I went into the logs and removed all the Bashes. After that, the logs made much more sense. =)

Here's his Damage Profile from his first test (where he cast "Strength" on himself):


Interesting things to note:
1) Shamans don't get a level damage bonus (which all of you probably knew, but I didn't )
2) The shape of this graph is roughly (very roughly) similar to the two graphs of my weapons and the graph of Copeland's Lamentation. The reason this particular graph is so "jagged" is because there are less data points overall. My first graph represents 3732 total hits. Copeland's graph probably represents around 5000 hits. This graph of Arkand's first test represents a total of 282 hits. As the number of hits increases, the graph becomes smoother and begins to look more like Copeland's.

Be that as it may, you can still see the rough outlines of the same shape that you see in my graphs and Copeland's graph.

In particular, there is a prominent peak at the Magic Number (16), a secondary peak at MIN (1), a distribution of points from MIN to Magic Number, and a smattering of hits from (Magic Number + 1) to MAX.

This suggests to me that the engine which drives the damage distribution of Shaman melee works just like the engine that drives the damage distribution for Fighter/Hybrid melee. This doesn't mean that Shamans can melee as well as Fighter/Hybrid classes. Shamans don't get Double Attack, nor Dual Wield, nor the Level Damage Bonus. All of these things will conspire to keep the total damage output of Shaman melee far below the damage output of Paladins and SKs, much less Warriors, Rogues, Monks and Rangers. It DOES mean, however, that when a Shaman makes a single successful hit, the mechanism which determines how much damage that hit causes to the MOB is the same mechanism (minus the Level Damage Bonus) that makes the same determination for Warriors, or Paladins, or whatever.
Here's the Damage Profile for Arkand's 2nd test:


This is the one that really got my attention.
Note that the primary peak is NOT the Magic Number. It's the MIN. Why? Because in this test, Arkand's 1-hand Blunt skill didn't get any higher than 150 *and* he didn't cast STR on himself to make up the difference. Note that the Damage Profile for Test 4 exhibits the same behavior -- which is also a test where he didn't cast the STR buff on himself.
Note that the damage distribution of a Warrior fighting red MOBs exhibits the same behavior.
Congratulations, Arkand. You've successfully proven that, at least in terms of Damage Distribution, fighting a blue with a low weapon skill does indeed simulate the damage distribution of fighting a red with a high weapon skill. =)
(However, since you didn't log your misses as well as your hits, we have no idea what your hit/swing ratio was, nor do we know how accurately it simulates fighting reds.)
Here's one last Damage Profile from Arkand. This is from his 3rd test, where his 1HB skill ranged from 154 to 160 and where he cast the 67-point STR buff on himself:



The conclusions that can be drawn from this graph are left as an exercise for the interested reader. =)

Just to add to this good post - when I did testing about a year ago on damage distribution and strength, I reached similar conclusions. This is a graph from an old test I posted here around September of last year, showing the % hit frequency for varying levels of strength:
Here is a chart summarizing some of the key variables in these tests:

It's really hard to draw any conclusion from that data as to whether strength affects a shift from minimum damage (1) to the magic number (I think it was 24 in that example). At best it's inconclusive I think. The only conclusive feature here is that strength extends the maximum damage figure a little further, leading to a tiny increase in average damage.

However, more recently I did a test using two "+ATK" type buffs: firefirst and greater wolf form on a level 51 druid. These buffs had the effect of raising ATK from 763 to 845, but changed no other visible stat. Here is a comparison of the hit distribution with and without these buffs (the test was performed against the same exact mob, a low blue, and all other factors, including level, strength, and weapon skills - 175 weapon, 200 offense - remained constant):


Quite a different result. Clearly the ATK buffs caused a huge shift from minimum damage amounts to magic number amounts. Here is a table also showing how this greatly swayed average damage amounts:


As you can see, greater average damage, less misses, less minimum hits, and more magic number hits.

I found the contrast of these two tests very interesting, and when combined with your results, they lead me to believe that, in essence, raising ATK does not always mean the same thing. The ATK variances in my old test were much greater than in the recent test, yet the results weren't nearly as striking.

It seems that if you raise strength, it will increase your maximum damage primarily, but will leave the rest of the distribution pretty much the same. However, if you raise your ATK rating via a spell other than strength, it will greatly affect all parameters of combat, in your favor.

[re: Arkand's 3rd test graph] The conclusions that can be drawn from this graph are left as an exercise for the interested reader. =)

Since you took the time to actually parse out that data, can you clarify whether the 1H blunt skill really changed during the test? If the test showing a distribution toward magic number hits was performed later in time, and there were skill increases along the way, that may explain a lot about why the distribution of the second graph is different - not because of STR, but because he gained skill. This aspect was not very clear in your post.

[This message has been edited by Tudamorf (edited 09-15-2000).]

It seems that if you raise strength, it will increase your maximum damage primarily, but will leave the rest of the distribution pretty much the same. However, if you raise your ATK rating via a spell other than strength, it will greatly affect all parameters of combat, in your favor.

I think this is more easily explainable by the simple fact that the two ATK buffs essentially compensated for your 175 weapon skill.

Consider: We have a 39th level Warrior with a STR of 100 and a 1HSlash skill of 200. We also have a 39th level Warrior with a STR of 200 and a 1HSlash skill of 100. Assume that the Offense skill for both is maxed. This means that their MAX damage for the same 1HSlash weapon will be the same. Give these two Warriors the same 1HSlash weapon and throw them at the same blue-conning MOB. Who will be more effective against it?

The smart money will bet on the Warrior with 200 skill and 100 STR.

The graphs from Arkand's logs show two things:
1) Low weapon skill (for appropriate values of "Low") will produce a damage distribution where the MIN has a higher peak than the Magic Number.
2) A sufficiently large STR buff can compensate for low weapon skill. In other words, 140 Weapon Skill + 60-point STR buff seems to be equivalent to a 200 Weapon Skill and no STR buff.

Consider the Damage Profile of Arkand's 1st test compared with his 4th test:
FIRST TEST
FOURTH TEST

Things to note here: There is significant overlap of ATK values in both tests. In the first test, Arkand's ATK rating ranged from 770 to 780. In the fourth test, Arkand's ATK rating ranged from 767 to 788.

Note, however, that the Damage Profile of the first test has the prominent peak at Magic Number. The Damage Profile of the fourth test has the prominent peak at MIN. Why?

Because in the first test, Arkand cast STR on himself. His skill ranged from 141 to 143 in the first test, but the 67-point STR buff compensated for that low skill. In essence, the 67-point STR buff artificially boosted his weapon skill to 200.

In the fourth test, Arkand did not cast STR on himself. His ATK rating was simply a reflection of his Weapon Skill + unbuffed STR + Offense Skill. And during the fourth test, Arkand's weapon skill ranged from 161 to 168. But he had no buffs to compensate for this low Weapon Skill, so the Damage Profile has the big peak at MIN.

This suggests to me that STR buffs (and almost certainly ATK buffs as well) act primarily as a boost to Weapon Skill if the weapon skill is low. Once the weapon skill hits a value that is considered to be "not low" (and who knows how that value is calculated?), the effect of the buff diminishes.

It is known that a 67-point STR buff on a Warrior with maxed Weapon and Offense skill for her level won't have a significant effect on her damage output aside from moving the value for MAX up the appropriate number of points. Compare the peaks at MIN and Magic Number on my first Damage Profile vs. the peaks at MIN and Magic Number for Copeland's Damage Profile. If STR did more than move the MAX and compensate for low weapon skill, then Copeland (with 93 more points of STR than me) would (in theory) have less hits at MIN compared to his hits at Magic Number than myself. We would see a shift of his hit distribution away from MIN and towards Magic Number.

But he doesn't. The total number of his hits for MIN is 80% of his total number of hits for Magic Number. The total number of my hits for MIN is 78% of my total number of hits for Magic Number. They're roughly equal, despite Copeland having 93 more points of STR than I do. Which suggests that if you were to somehow buff my STR by 93 points, it wouldn't somehow give me a weapon skill of 293. So there appears to be a point at which a STR buff stops acting like a boost to Weapon Skill. My guess is that it's around 200.


Are the lower strength buffs just added 2-3 extra points on max which may only hit a few times in a night?

Unfortunately, it certainly looks that way. The 5-point 1st level STR buff that Shamans get is completely, utterly ineffective for any purpose at all that I can tell except for giving you five more pounds of carrying capacity.

As the buffs add more to STR, their value becomes highly dependent on the DMG rating of the weapon. See my post named Some Preliminary Analysis on How STR + DMG Affects Combat for more details.

allow me to quote ye a moment.

"Congratulations, Arkand. You've successfully proven that, at least in terms of Damage Distribution, fighting a blue with a low weapon skill does indeed simulate the damage distribution of fighting a red with a high weapon skill."

So we can assume that fighting a red with a high level skill and a 67 point strength buff will simulate me fighting a blue yes?

You shoulda quoted the very next line, Valantor, 'cause I answered that very question.

Here it is:
"(However, since you didn't log your misses as well as your hits, we have no idea what your hit/swing ratio was, nor do we know how accurately it simulates fighting reds.)"

All I can say about how accurately "low weapon skill vs. blue MOB" simulates "high weapon skill vs. red MOB" is that the damage distribution for both situations is similar. Note that "damage distribution" does not equate to "overall effectiveness." If the hit/swing ratio for the "high weapon skill + red MOB" scenario is 50%, while the hit/swing ratio for "low weapon skill vs. blue MOB" is 60%, then no, the "low weapon skill vs. blue MOB" scenario will not be an accurate simulation of the "high weapon skill vs. red MOB" scenario, even if their graphs look identical down to the last point value.

Before we can say that it's an accurate simulation, we would need to know (A) the average hit ratio for a low-skill, blue MOB scenario; and (B) the average hit ratio for a high-skill, red MOB scenario.

Until then, all we can say is "The graphs look more-or-less the same."


Nowhere in your post did you mention the bonus damage Verant added to two handed weapons a few months ago. Are these tests from before that change? Or are the charts up to date and there never was a bonus damage added to 2hand at all? I wouldnt put that past Big Brother Verant.

Nah, no conspiracy. I just forgot.

The new Level Damage Bonus for 2-handers is calculated by:

2-handed Level Damage Bonus up to 50th level = (Level - 25)/2

At 51st level or higher, it's calculated by:

2-handed Level Damage Bonus post-50th = (Level - 7)/3

BUT!

Not only did they adjust the Level Damage Bonus for 2-handers, but they seem to have snuck in a hidden bonus for 2-handers that nobody can yet account for.

Look back up top at the graph of Qualtar's Damage Profile with the Polyphenomenal Axe.

Qualtar was 54th level during the time the Profile was taken. So his Level Damage Bonus would be (54 -7)/3 == 47/3 == 15.

The Polyphenomenal Axe is a DMG 37 weapon. So his Magic Number would be (2 * 37) + 15 == 74 + 15 == 89. And sure enough, the big peak of the graph is at 89.

So! His MIN should be 1 + 15 = 16. And his MAX should be 155.

But they ain't.

His MIN is 19. His MAX is 158.

Something shifted the two ends of his damage spectrum upwards by 3 points, while leaving his Magic Number unaffected. And so far we have no clue what that "something" might be.

[Treats]

Revisiting and bumping this for Kanras, found a few older posts.

PWR is the portion of displayed ATK that relates to damage calculation after a hit has successfully landed. The other portion of displayed ATK is ACC which relates to your chance to hit. Yes, STR has an impact on PWR.

My question -- Are the effects of STR for ATK on P99 too high?

Also, several of us have tested stats and their effect on damage & accuracy using logs of fights and found that no statistic influences accuracy (which was later confirmed by Verant) and that strength only slightly influences damage output at high levels.

STR does increase maximum damage and *therefore*
it increases average damage. However the effect
on average damage is *extremely* small (a few
percents for a +100 STR buff, with most formulae),
which means for most purposes you can ignore the
effect of STR on your damage output.

You must not be playing the same game, then. No amount of stats will significantly change your damage output - the highest strength I saw tested was 240 and average damage (i.e., damage/second) barely changed; less than a single point increase by moving from 60 strength to 240.

Lots of other info in the thread also:

http://paladinsofnorrath.yuku.com/topic/1829/AC-vs-StatsAlagarn#.UQcalKnB-uI

Here he also states that the effects of STR on ATK were changed at some point:

We all know that ATK is valuable. We ask and answer a lot of questions about ATK. We're seeing now that ATK from strength actually works the same as "pure" ATK (a change from days past). However, unless someone is at least a little mathematically minded, explaining that it makes you "hit harder" can be difficult to understand and unsatisfying without knowing why.

http://paladinsofnorrath.yuku.com/topic/1782/Value-of-ATK-graph-of-damage-distribution#.UQcaSKnB-uI

[quido]

Give us the full damage source across the board for review and commentary.

[Treats]

Could not find the original post, that was the best I could do.

Not sure where I could find a backup of ezboard, this is the original link with all of the data:

server5.ezboard.com/fpaladinsofnorrathgeneral.showMessage?topicID=522. topic

This is where I got the link from:

https://groups.google.com/group/alt.games.everquest/browse_thread/thread/986910de51191dff?hl=en&noredirect=true