Update for MoP Hit Tables by TheBackstabi · Pull Request #69 · wowsims/mop

TheBackstabi · 2025-05-11T07:45:35Z

I'm kind of stupid so maybe missed something, but this should be right for updating Dodge/Parry to 7.5% and Spell Hit inheriting from Expertise and Hit Rating.

As far as I could tell looking through SimC and old posts, the logic for the rest of the table remained as-is.

TheBackstabi · 2025-05-11T08:04:16Z

https://www.wowhead.com/news/new-ghostcrawler-blog-mists-of-pandaria-stat-changes-201233

I missed one thing - updating Block to be a second roll, after Miss/Dodge/Parry (assuming this applies to Mobs as well and not just players)

I need to sleep now so if someone else gets to that, go for it.

TheBackstabi · 2025-05-12T06:18:03Z

Need to update Enemy->Player block rolls

Crit suppression source: https://www.bluetracker.gg/wow/topic/us-en/5889309137-beta-class-balance-analysis/#97

…ther checks for readability

NerdEgghead · 2025-05-13T10:23:03Z

 func (spell *Spell) OutcomeExpectedMeleeWhite(_ *Simulation, result *SpellResult, attackTable *AttackTable) {
 	missChance := spell.GetPhysicalMissChance(attackTable)
-	dodgeChance := TernaryFloat64(spell.Flags.Matches(SpellFlagCannotBeDodged), 0, max(0, attackTable.BaseDodgeChance-spell.DodgeParrySuppression()-spell.Unit.PseudoStats.DodgeReduction))
-	parryChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, max(0, attackTable.BaseParryChance-spell.DodgeParrySuppression()), 0)
+	dodgeChance := TernaryFloat64(spell.Flags.Matches(SpellFlagCannotBeDodged), 0, max(0, attackTable.BaseDodgeChance-spell.DodgeSuppression()-spell.Unit.PseudoStats.DodgeReduction))
+	parryChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, max(0, attackTable.BaseParryChance-spell.ParrySuppression()), 0)
 	glanceChance := attackTable.BaseGlanceChance
 	blockChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, attackTable.BaseBlockChance, 0)
 	whiteCritCap := 1.0 - missChance - dodgeChance - parryChance - glanceChance - blockChance


The formulas for whiteCritCap and averageMultiplier also need to be updated here.

Crit cap makes sense, but I'm not sure there's anything needed for averageMultiplier...?

NerdEgghead · 2025-05-13T10:25:12Z

 func (spell *Spell) OutcomeExpectedMeleeWeaponSpecialHitAndCrit(_ *Simulation, result *SpellResult, attackTable *AttackTable) {
 	missChance := max(0, attackTable.BaseMissChance-spell.PhysicalHitChance(attackTable))
-	dodgeChance := TernaryFloat64(spell.Flags.Matches(SpellFlagCannotBeDodged), 0, max(0, attackTable.BaseDodgeChance-spell.DodgeParrySuppression()-spell.Unit.PseudoStats.DodgeReduction))
-	parryChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, max(0, attackTable.BaseParryChance-spell.DodgeParrySuppression()), 0)
+	dodgeChance := TernaryFloat64(spell.Flags.Matches(SpellFlagCannotBeDodged), 0, max(0, attackTable.BaseDodgeChance-spell.DodgeSuppression()-spell.Unit.PseudoStats.DodgeReduction))
+	parryChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, max(0, attackTable.BaseParryChance-spell.ParrySuppression()), 0)
 	blockChance := TernaryFloat64(spell.Unit.PseudoStats.InFrontOfTarget, attackTable.BaseBlockChance, 0)
 	critChance := spell.PhysicalCritChance(attackTable)
 	averageMultiplier := (1.0 - missChance - dodgeChance - parryChance) * (1.0 + (spell.CritDamageMultiplier()-1)*critChance)


I think these formulas should still hold though in MoP, right? Worth a double check.

NerdEgghead · 2025-05-13T10:38:07Z

+	parrySupp := expertiseRating / ExpertisePerQuarterPercentReduction / 400
+	parrySupp -= max(0, spell.DodgeSuppression())


There's a bug here: if the player has 10% Expertise then we should calculate 2.5% Parry suppression against a raid boss, but this formula will return 0 since spell.DodgeSuppression() will return 10%. I think this method will also need to take in an attackTable input to work properly in MoP, since you need access to attackTable.BaseDodgeChance and spell.Unit.PseudoStats.DodgeReduction in order to determine the correct amount of Expertise that was already "spent" on Dodge suppression.

I think we'll need exploration of what would use the DodgeReduction field, and if that actually impacts the Parry outcome too. I don't actually know/see anything interacting with that field anymore.

NerdEgghead · 2025-05-13T10:43:11Z

 	if defender.Type == EnemyUnit {
 		// Assumes attacker (the Player) is level 80.


Delete this comment, since I don't think there is any assumption required on player level for these calculations, just on the relative difference between the attacker and target levels.

NerdEgghead · 2025-05-13T10:43:31Z

 	} else {
 		// Assumes defender (the Player) is level 80.


NerdEgghead · 2025-05-13T10:45:00Z

 	} else {
 		// Assumes defender (the Player) is level 80.
 		table.BaseSpellMissChance = 0.05


Have you checked the NPC-attacking-player values below as well for correctness in MoP?

InDebt · 2025-05-13T13:29:47Z

https://www.engadget.com/2012-03-08-scattered-shots-hunter-expertise-in-mists-of-pandaria.html
https://wowcraftguides.weebly.com/hunter-guide-54-guide.html

Those blogs and guides indicate that Ranged attacks should now be dodgeable. So we might want to add that as well here?

hillerstorm · 2025-05-13T13:40:26Z

https://www.engadget.com/2012-03-08-scattered-shots-hunter-expertise-in-mists-of-pandaria.html https://wowcraftguides.weebly.com/hunter-guide-54-guide.html

Those blogs and guides indicate that Ranged attacks should now be dodgeable. So we might want to add that as well here?

Yep the default ranged outcome should include dodges and perhaps add one for NoDodge? If a spell matching that even exists

TheBackstabi · 2025-05-13T20:32:43Z

Just leaving more notes as I research things in between real work:
Enemy->Player base values

Not relevant to this, but for any tank sims

https://www.engadget.com/2012-03-01-ghostcrawler-explains-stat-changes-in-mists-of-pandaria.html

TheBackstabi · 2025-05-14T00:35:31Z

I've left the negative Dodge/Parry/Block Base values for Enemy->Player as I assume those are needed for accurate mitigation calculations, even though an actual negative value should never happen.

Miss/Spell Miss I've left at 0 for 92/93 since there's no way of mitigating that anymore

I forgot ranged lmao

NerdEgghead · 2025-05-14T06:52:13Z

 	if dot.Spell.Unit.PseudoStats.InFrontOfTarget {
 		if !result.applyAttackTableMissNoDWPenalty(dot.Spell, attackTable, roll, &chance) {


Do you need to handle dodges here as well?

Snapshot initially made me think it was used in DoTs like Explosive Shot (Which can be tested to see how Dodges work once it's fixed), but looking closer I don't see anywhere it gets used

NerdEgghead · 2025-05-14T06:55:43Z

+	// TODO: Ranged Attacks are currently blockable on MoP Beta
+	// However, they shouldn't be! Revisit when fixed


Ranged attacks are also not currently dodge-able on Beta. Both are known issues that will likely be fixed, so we should handle them the same way: either make both Dodge and Block function like 5.4 in the sim under the assumption that both will be fixed, or make both function like Beta with a note to revisit both.

NerdEgghead · 2025-05-14T07:07:28Z

-	*chance += max(0, attackTable.BaseDodgeChance-spell.DodgeParrySuppression()-spell.Unit.PseudoStats.DodgeReduction)
+	*chance += max(0, attackTable.BaseDodgeChance-spell.DodgeSuppression()-spell.Unit.PseudoStats.DodgeReduction)


PseudoStats.DodgeReduction has been unused since WotLK, so I think you can delete it outright from the MoP codebase. Likewise, you should be able to delete the suppress_dodge field from proto.Target.

NerdEgghead · 2025-05-14T07:27:54Z

 	critChance := min(spell.PhysicalCritChance(attackTable), whiteCritCap)
 	averageMultiplier := 1.0 - missChance - dodgeChance - parryChance + (spell.CritDamageMultiplier()-1)*critChance - glanceChance*(1.0-attackTable.GlanceMultiplier) - blockChance*result.Target.BlockDamageReduction()


I mathed it out, the correct formula for the MoP 2-roll white table should be:

averageMultiplier := (1.0 - missChance - dodgeChance - parryChance + (spell.CritDamageMultiplier()-1)*critChance - glanceChance*(1.0-attackTable.GlanceMultiplier)) * (1.0 - blockChance*result.Target.BlockDamageReduction())

Essentially, the average Block damage reduction from the second roll now gets multiplied across all the terms due to the possibility of Crit-Block and Glance-Block outcomes in MoP.

NerdEgghead · 2025-05-14T08:02:30Z

 	critChance := spell.PhysicalCritChance(attackTable)
 	averageMultiplier := (1.0 - missChance - dodgeChance - parryChance) * (1.0 + (spell.CritDamageMultiplier()-1)*critChance)


Did the math for this one as well, in MoP the formula becomes:

critFactor := (spell.CritDamageMultiplier() - 1) * critChance averageMultiplier := (1.0 - missChance - dodgeChance - parryChance) * (1.0 + critFactor - blockChance * (critFactor + result.Target.BlockDamageReduction()))

where you'd omit the -= term on the next line from the original formula.

The intuitive explanation is that it's the same formula from Cata, except that the "effective Block chance" now gets suppressed by the total miss chance for the ability, since misses now prevent Block from even being rolled unlike before.

NerdEgghead · 2025-05-14T08:11:59Z

-		table.BaseParryChance = UnitLevelFloat64(attacker.Level, 0, -0.002, -0.004, -0.006)
+		table.BaseSpellMissChance = UnitLevelFloat64(defender.Level, 0.06, 0.03, 0, 0)


It should be attacker.Level here similar to the expressions below.

NerdEgghead · 2025-05-14T08:19:57Z

+		table.BaseMissChance = UnitLevelFloat64(attacker.Level, 0.03, 0.015, 0, 0)
+		table.BaseBlockChance = UnitLevelFloat64(attacker.Level, 0.03, 0.015, 0, -0.015)
+		table.BaseDodgeChance = UnitLevelFloat64(attacker.Level, 0.03, 0.015, 0, -0.015)
+		table.BaseParryChance = UnitLevelFloat64(attacker.Level, 0.03, 0.015, 0, -0.015)


Are you sure that +2 and +3 level enemies have a 0% chance to Miss with their autos? This seems odd, given that simc has specific logic coded to make auto-attack Miss outcomes not grant extra Vengeance and only refresh the existing Vengeance buff, unlike auto-attack Dodge and Parry outcomes which grant Vengeance as if they had landed. If boss auto Miss outcomes aren't even possible in the attack table, then they shouldn't have needed this branching logic.

This was also from the GC posts. I can do more log digging to double check, but so far everything that was said in that thread has held true when verified.

The mechanics of Vengeance were also mentioned - I think the case of a Miss is just mechanically present, but will never occur on a 92+.

It does hold relevance for trash mobs in Challenge Modes, if SimC had any form of their DungeonSlice model then

Some bosses appear to be flagged as Dual Wielding, and receive the very same +19% modifier to their miss chance. If we want to support that in an Encounter design, I will need to add in the negative values so the final miss rate is correct.

https://www.warcraftlogs.com/reports/aqYQNvtbV3L8CgmT?fight=3&type=damage-taken&source=22&ability=1&target=102
https://www.wowhead.com/mop-classic/npc=71479/he-softfoot

I've been otherwise unable to find any Miss hits from non-DW bosses (the rest of SoO, basically)

NerdEgghead · 2025-05-15T09:36:52Z

 	if spell.Unit.PseudoStats.InFrontOfTarget {
-		if !result.applyAttackTableMissNoDWPenalty(spell, attackTable, roll, &chance) {
-			if result.applyAttackTableCritSeparateRoll(sim, spell, attackTable, countHits) {
-				result.applyAttackTableBlock(spell, attackTable, roll, &chance)
-			} else {
-				if !result.applyAttackTableBlock(spell, attackTable, roll, &chance) {
-					result.applyAttackTableHit(spell, countHits)
-				}
-			}
+		if !result.applyAttackTableMissNoDWPenalty(spell, attackTable, roll, &chance) &&
+			!result.applyAttackTableDodge(spell, attackTable, roll, &chance) &&
+			!result.applyAttackTableCritSeparateRoll(sim, spell, attackTable, countHits) {
+			result.applyAttackTableHit(spell, countHits)
 		}
 	} else {
 		if !result.applyAttackTableMissNoDWPenalty(spell, attackTable, roll, &chance) &&


You also need to handle the Dodge outcome in the case where the Hunter is attacking from behind. In fact, now that Blocks are no longer possible for ranged attacks, I don't think you even need to branch off InFrontOfTarget anymore - it should be the same attack table in both cases.

NerdEgghead · 2025-05-15T09:41:05Z

 	if dot.Spell.Unit.PseudoStats.InFrontOfTarget {
-		if !result.applyAttackTableMissNoDWPenalty(dot.Spell, attackTable, roll, &chance) {
-			if result.applyAttackTableCritSeparateRollSnapshot(sim, dot) {
-				result.applyAttackTableBlock(dot.Spell, attackTable, roll, &chance)
-			} else {
-				if !result.applyAttackTableBlock(dot.Spell, attackTable, roll, &chance) {
-					result.applyAttackTableHit(dot.Spell, countHits)
-				}
-			}
+		if !result.applyAttackTableMissNoDWPenalty(dot.Spell, attackTable, roll, &chance) &&
+			!result.applyAttackTableDodge(dot.Spell, attackTable, roll, &chance) &&
+			!result.applyAttackTableCritSeparateRollSnapshot(sim, dot) {
+			result.applyAttackTableHit(dot.Spell, countHits)
 		}
 	} else {
 		if !result.applyAttackTableMissNoDWPenalty(dot.Spell, attackTable, roll, &chance) &&


Same here, you're missing Dodge outcomes when attacking from behind, and shouldn't need to branch off InFrontOfTarget anymore in MoP if I'm understanding the changes correctly.

NerdEgghead · 2025-05-15T09:45:47Z

 	if spell.Unit.PseudoStats.InFrontOfTarget {
-		roll := sim.RandomFloat("White Hit Table")
-		chance := 0.0
-
 		if result.applyAttackTableCritSeparateRoll(sim, spell, attackTable, countHits) {
-			result.applyAttackTableBlock(spell, attackTable, roll, &chance)
+			result.applyAttackTableBlock(sim, spell, attackTable)
 		} else {
-			if !result.applyAttackTableBlock(spell, attackTable, roll, &chance) {
+			if !result.applyAttackTableBlock(sim, spell, attackTable) {
 				result.applyAttackTableHit(spell, countHits)
 			}
 		}


Remove ranged Block outcomes here for consistency if you're removing them everywhere else, which also means that InFrontOfTarget should no longer matter.

NerdEgghead · 2025-05-15T09:53:19Z

+	if sim.RandomFloat("Player Block") < chance {
 		result.Outcome |= OutcomeBlock
 		spell.SpellMetrics[result.Target.UnitIndex].Blocks++


In MoP this also needs a branch for if result.DidCrit() like the player-vs-npc case, since CritBlock outcomes are now possible for enemy autos. It won't be relevant in practice since all tanks are Crit immune, but we might as well handle it correctly for completeness.

I believe this should already be handled by the way outcomeEnemyMeleeWhite flows, in the exact same way Player->Target does. It checks Miss/Dodge/Parry, if none of those are true it then checks Crit. If it does Crit, then it rolls Block. If Block is true, it takes the Crit outcome and applies |= Block to the Outcome (as opposed to just setting it equal to, as the Crit/Miss/Dodge/Parry/Hit functions do)

Ah I see, to add to SpellMetrics for -specifically- CritBlock. I'll add this

Spell Power/Damage updates

(Attempted) Update for MoP Hit Tables

ea23361

TheBackstabi added 2 commits May 11, 2025 23:01

Merge https://github.com/wowsims/mop into fix/mop-attack-table

80da31c

Update Crit Suppression and Player->Enemy Block rolls

f8d4550

TheBackstabi added 3 commits May 12, 2025 12:15

Merge https://github.com/wowsims/mop into fix/mop-attack-table

db17a90

Fix auto attack checks to allow Crit/Block and Glance/Block, refine o…

1536105

…ther checks for readability

More readability changes

f3e40b4

NerdEgghead requested changes May 13, 2025

View reviewed changes

Address feedback

01a3825

Ranged can be Dodged

31afaff

TheBackstabi requested a review from NerdEgghead May 14, 2025 00:53

Slide in the attack speed buff fix

19975ae

NerdEgghead requested changes May 14, 2025

View reviewed changes

TheBackstabi added 2 commits May 14, 2025 11:05

PR Feedback

9a25473

Merge https://github.com/wowsims/mop into fix/mop-attack-table

7e3cce2

NerdEgghead requested changes May 15, 2025

View reviewed changes

TheBackstabi added 2 commits May 16, 2025 13:19

PR Ranged Dodge feedback

4d6b2c9

CritBlock metric for Enemy Attack Table

f3fb06e

NerdEgghead approved these changes May 17, 2025

View reviewed changes

1337LutZ merged commit 63cd7f0 into master May 19, 2025
2 checks passed

espionn pushed a commit to espionn/mop that referenced this pull request Mar 8, 2026

Merge pull request wowsims#69 from Jazz405/rhonnie/school-damage

b3ce9f6

Spell Power/Damage updates

		parrySupp := expertiseRating / ExpertisePerQuarterPercentReduction / 400
		parrySupp -= max(0, spell.DodgeSuppression())

		if defender.Type == EnemyUnit {
		// Assumes attacker (the Player) is level 80.

		if dot.Spell.Unit.PseudoStats.InFrontOfTarget {
		if !result.applyAttackTableMissNoDWPenalty(dot.Spell, attackTable, roll, &chance) {

		// TODO: Ranged Attacks are currently blockable on MoP Beta
		// However, they shouldn't be! Revisit when fixed

		*chance += max(0, attackTable.BaseDodgeChance-spell.DodgeParrySuppression()-spell.Unit.PseudoStats.DodgeReduction)
		*chance += max(0, attackTable.BaseDodgeChance-spell.DodgeSuppression()-spell.Unit.PseudoStats.DodgeReduction)

		critChance := min(spell.PhysicalCritChance(attackTable), whiteCritCap)
		averageMultiplier := 1.0 - missChance - dodgeChance - parryChance + (spell.CritDamageMultiplier()-1)critChance - glanceChance(1.0-attackTable.GlanceMultiplier) - blockChance*result.Target.BlockDamageReduction()

		critChance := spell.PhysicalCritChance(attackTable)
		averageMultiplier := (1.0 - missChance - dodgeChance - parryChance) * (1.0 + (spell.CritDamageMultiplier()-1)*critChance)

		table.BaseParryChance = UnitLevelFloat64(attacker.Level, 0, -0.002, -0.004, -0.006)
		table.BaseSpellMissChance = UnitLevelFloat64(defender.Level, 0.06, 0.03, 0, 0)

Conversation

TheBackstabi commented May 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

TheBackstabi commented May 11, 2025

Uh oh!

TheBackstabi commented May 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

InDebt commented May 13, 2025

Uh oh!

hillerstorm commented May 13, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

TheBackstabi commented May 13, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

TheBackstabi commented May 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

TheBackstabi May 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

TheBackstabi May 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

TheBackstabi May 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

TheBackstabi May 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

TheBackstabi May 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

TheBackstabi commented May 11, 2025 •

edited

Loading

TheBackstabi commented May 12, 2025 •

edited

Loading

hillerstorm commented May 13, 2025 •

edited

Loading

TheBackstabi commented May 13, 2025 •

edited

Loading

TheBackstabi commented May 14, 2025 •

edited

Loading

TheBackstabi May 14, 2025 •

edited

Loading

TheBackstabi May 14, 2025 •

edited

Loading

TheBackstabi May 14, 2025 •

edited

Loading

TheBackstabi May 16, 2025 •

edited

Loading

TheBackstabi May 16, 2025 •

edited

Loading