DPS: Monk vs. Warrior vs. Rogue vs. Ranger

[bcbrown]

Originally Posted by Snaggles [You must be logged in to view images. Log in or Register.]

11x2+11 (damage bonus)/18 Delay = 1.8333
30x2+11 (damage bonus) /40 (delay) = 1.775

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

Blam Stick
44.125 / 40 (delay) = 1.1 ratio

SBoZ
21.0875 / 18 (delay) = 1.17 ratio

Your calculations are different than Snaggles, and I understand why, but both calculations come up with the same result: SBoZ is the better weapon. Is there any example where your calculations come up with a different result than Snaggles' would?

[DeathsSilkyMist]

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

Your calculations are different than Snaggles, and I understand why, but both calculations come up with the same result: SBoZ is the better weapon. Is there any example where your calculations come up with a different result than Snaggles' would?

Read my previous edited post.

[bcbrown]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

Read my previous edited post.

Back-running your calculations, I think a 32.34/40 weapon would be equivalent to SBoZ, meaning SBoZ is better than a 32/40, but worse than a 33/40.

Using (2*dmg + bonus)/delay, 32.34/40 comes out to a 1.89 ratio, very slightly better than the 1.83 ratio for SBoZ under this approach. 31/40 comes out to 1.825, 32/40 comes out to 1.875, 33/40 is 1.925.

So under your approach, these would be ranked best-to-worst as 33/40, 11/18, 32/40, 31/40. Under Snaggles' approach, the ranking would be 33/40, 32/40, 11/18, 31/40.

To back-run the equivalent 40-delay weapon as SBoZ under DSM's approach:
3b: 46.40 / 40 = 1.16
3a: 81.8 + 11 = 92.8 average damage * 0.5 miss rate = 46.40
3: (154.6 + 9) / 2 = 81.8
2c: 32.34 * 2 * 239 / 100 = 154.58

Does that look correct, DSM?

[DeathsSilkyMist]

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

Back-running your calculations, I think a 32.34/40 weapon would be equivalent to SBoZ, meaning SBoZ is better than a 32/40, but worse than a 33/40.

Using (2*dmg + bonus)/delay, 32.34/40 comes out to a 1.89 ratio, very slightly better than the 1.83 ratio for SBoZ under this approach. 31/40 comes out to 1.825, 32/40 comes out to 1.875, 33/40 is 1.925.

So under your approach, these would be ranked best-to-worst as 33/40, 11/18, 32/40, 31/40. Under Snaggles' approach, the ranking would be 33/40, 32/40, 11/18, 31/40.

To back-run the equivalent 40-delay weapon as SBoZ under DSM's approach:
3b: 46.40 / 40 = 1.16
3a: 81.8 + 11 = 92.8 average damage * 0.5 miss rate = 46.40
3: (154.6 + 9) / 2 = 81.8
2c: 32.34 * 2 * 239 / 100 = 154.58

Does that look correct, DSM?

I'll trust your numbers ratio-wise. These are using unweighted dice. On High AC targets, we need to lower the damage values by some percentage as the average skews lower. We can just make it up for now and say you divide that max damage + min damage by 3 instead of 2:

Blam Stick

1. 225 Offense Skill + ((2 * (231 Strength + 21 ATK from 1x AoB and Firefist) - 150) / 3) = 343.
2a. 350 - 65 (Minusfactor from damage table) = 278
2b. (285 / 2) + 100 = 239
2c. ((30 Weapon Damage * 2) * 239) / 100 = 143.4 max damage before damage bonus
3. 143.4 + 9 (min damage) / 3 = 50.8. average damage
3a. 50.8 + 11 (damage bonus) = 61.8 damage average * 0.5 miss rate = 30.9
3b. 30.9 / 40 (delay) = 0.77 ratio

SBoZ

1. 225 Offense Skill + ((2 * (231 Strength + 21 ATK from 1x AoB and Firefist) - 150) / 3) = 343.
2a. 350 - 65 (Minusfactor from damage table) = 278
2b. (285 / 2) + 100 = 239
2c. ((11 Weapon Damage * 2) * 239) / 100 = 52.58 max damage before damage bonus
3. 52.58 + 9 (min damage) / 3 = 20.52 average damage
3a. 20.52 + 11 (damage bonus) = 31.52 damage average * 0.5 miss rate = 15.76
3b. 15.76 / 18 (delay) = 0.87 ratio

As you can see, the difference between Blam Stick and SBoZ is now 0.1 instead of 0.07. Not a huge leap, but SBoZ gained ratio. So with these formulas you can see the specific gains for different scenarios roughly speaking.

[Snaggles]

Ultimately as long as the system has some degree of common sense and is used consistently, I wouldn’t knock any someone tweaking their own.

Many of the spreadsheets either assume an optimistic offhand swing ratio or uses outdated damage bonuses for 2h’s. This makes dual wield look far better than it is. People rarely look at the cell calculations too…which is troubling. I’ll let Excel do my math but I’d like to know how it’s doing it, lol.

[bcbrown]

Seems to me that DSM's approach is indeed more accurate, but that the (2*DMG + bonus) / delay is so much simpler with only minimal loss in accuracy (32.1 SBoZ equivalence at 40 delay instead of 32.4 in this scenario) that I understand why that's the popular rule of thumb. And that last example also helped me understand why SBoZ gets even better at high AC targets.

I have some follow-up questions to help clarify some of the mechanics I still understand poorly:

Using min damage value of 9 for a level 60 based on some log data for Frostwrath.

Could you clarify what min damage means here? I think what you're saying is that the minimum pre-bonus hit is 9, which means the log data here would be showing a minimum actual hit of 20 (with 11 damage bonus). Is that correct?

The damage table thing you linked shows { 285 Max Extra, 23 Chance, 65 Minusfactor }. The calculation you did uses the Minusfactor value. What do the other two numbers do?

Is the calculation you provided documented anywhere on the wiki?

And lastly, as an FYI, if you click the line number in Github it updates the URL to to link directly to that line: https://github.com/EQEmu/Server/blob...tack.cpp#L5285

[DeathsSilkyMist]

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

Could you clarify what min damage means here? I think what you're saying is that the minimum pre-bonus hit is 9, which means the log data here would be showing a minimum actual hit of 20 (with 11 damage bonus). Is that correct?

Yes. I never hit for 12, which would be the lowest damage value if the minimum damage was 1 at level 60. Lowest damage value I saw was 21 on Frostwrath (21 - 11= 10) and 46 on my 46/44 2handed weapon (46-37 = 9). Since the difference was 1 damage lower on the 2h weapon, I am assuming that is the min value. I don't know right now what the exact formula is for determining minimum damage. These damage bonus values of 11 and 37 are from this table:

https://lucy.allakhazam.com/dmgbonus.html?setcookie=1

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

The damage table thing you linked shows { 285 Max Extra, 23 Chance, 65 Minusfactor }. The calculation you did uses the Minusfactor value. What do the other two numbers do?

I don't remember off the top of my head.

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

Is the calculation you provided documented anywhere on the wiki?

It is similar to the existing one on the wiki if I recall, but this one is more detailed. I probably should update the wiki at some point.

[bcbrown]

I'd be happy to collaborate with you on the wiki update, or serve as a reviewer to double-check the math, because https://wiki.project1999.com/Game_Me...bat_and_Damage is the only wiki reference I see, and it's very much incomplete and outdated.

[DeathsSilkyMist]

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

I'd be happy to collaborate with you on the wiki update, or serve as a reviewer to double-check the math, because https://wiki.project1999.com/Game_Me...bat_and_Damage is the only wiki reference I see, and it's very much incomplete and outdated.

Yeah I wouldn't mind doing that. Only outstanding thing is testing attack bonus. I need to test with avatar at some point to see if extra attack is actually added as extra strength. That's what it seems like, but only having +21 attack is not a huge difference in the calculation. so that could be incorrect. I am pretty confident extra attack does increase your max damage, so it has to go somewhere.

Oh and minimum damage calulations.

[Jimjam]

On the subject of rounding dmg, how does this work woth the 20 damage intervals? Is it round up to 1 otherwise round to closest?

for i=1,20

if((i/10*dmg)<1,1,ROUND(i/10*dmg))

What would be the best way to check? A 10 dmg vs 3 dmg weapon on a low level caster?