So, about defenses aka. PHB2 defenses feats

[Elric]

You've made a blanket statement, that you -would- argue. However, in truth, you haven't. The method for doing so is easy--calculate the threat value of various brutes' attacks over levels. You've made a statement that can be proven without providing evidence to back this point. I can't respond to dispute or rebut your statement because, in truth, you haven't presented any premises to dispute/rebut.

(Did you mean "can't?")

One could do all sorts of calculations to try to get at an answer here. However, this is a question that it is possible to answer without having done a ton of calculations off of tables in the DMG. To quote James McMurray, who's on your side of "causing FRW to scale more like AC would be bad", but agrees that "attacks against FRW do not become weaker over time relative to attacks against AC in a way that compensates for the better scaling of FRW attacks"

I'd definitely have to agree. The attacks vs. AC don't deal near enough damage to make up for the conditions imposed by most attacks against FRW. And since their secondary effects are usually a little forced movement and/or knocking prone, there's even less reason to worry about them. It takes at least 4 hits from a creature to deal as much damage as your leader will heal with a minor action, so unless you're surrounded or facing some sort of quesinart, attacks versus AC usually don't matter much.

Here's a simple explanation. Those feats are parallels for the equivalent Epic feats from 3.x. People wanted feats to bolster their stuff, Wizards decided to do so. There's no evidence presented that it's a 'fix' when Wizards has shown -multiple times- that they're willing to institute fixes with errata.

Popular demand? Look at all the Char-Op. It's -all- based on maximizing to hit, and maximizing defenses. Char-Op players -are- a signifigant number of players of D&D, and perhaps Wizards wanted to toss them a bone?

That's the easiest explanation.

WotC is going to have an article addressing the Expertise feats. I would be very surprised if they gave this rationale for the feats. That said, if that was truly their rationale when they designed the feats they'd almost surely lie and come up with something more appealing, like "we needed a math fix and didn't want to issue errata" Still, assuming we take that article at its word, this will be settled in due time.

So, the effective threat to hps of a single hit goes down as the monster goes up in level.

So, for the monster's attacks to have equal effectiveness over levels, the chance to hit -must- go up as the monster gains levels.

There's no reason to assume the game is shooting for equal effectiveness for monster attacks in average percentage of PC HP across levels. In the PH it's certainly not the case that PC damage scales as quickly as monster HP, and unlike monsters who gain on to-hit vs. AC and even more so on FRWs, PCs lose on static to-hit against monster defenses as level increases (this isn't to say that the rate at which the number of rounds per combat increases is necessarily fine- see the threads on grind).

I think you also missed the thrust of my argument. Set aside Brutes; they're just one monster type. FRW attacks scale by 5 better over 29 levels; AC attacks by 2. Yet the DMG doesn't tell you to construct monsters any differently at level 28 if they attack FRW instead of AC compared to how you'd do it at level 4. There is nothing like "use higher damage for all attacks against AC than attacks against FRW if you're building a level 20+ monster, since the FRW attacks hit more often compared to AC attacks than they used to."

To hit FRW and to hit AC should according to the same page be within 2 points of each other for a monster. Using light armor as a base (as heavy armor is balanced with light armor in mind) we can see clearly that:

For cloth wearers, the disparity between AC and Reflex (best save for light armor wearers) is 0 at heroic, 1 at paragon, and 2 at epic. For leather, it's 2/3/4 and for hide it's 3/4/5. Everything else is derived from statistics that affect both AC and FRW.

Using a character's strongest FRW lessens the problems that arise. In particular, if you compared a character's weak FRW you'd end up with a very different conclusion.

Additionally, the way you've done this with "Cloth/Leather/Hide" armor obscures the fact that level 1 PCs generally have AC 2 or more higher than average FRW. However, this is a problem relevant from the start, not one intrinsic to the scaling.

...Your 'fix' was already included in the PHB1. It's called Lightning Reflexes, Iron Will, Great Fortitude.

There are feats that boost AC in the PH as well. In general, you can get +1 to AC with one feat (+2 if you take Leather Armor Proficiency). AC is attacked much more than any given FRW. Is +1 to AC significantly worse than +2 to one FRW in a way that would lead you to conclude that Lightning Reflexes is a "fix", while Armor Specialization is not, or are the FRW bonuses +2 simply as result of AC being more attacked than FRW, so feats that boost FRW should give a larger bonus? I think it's the latter.

If these boosts aren't valuable because of poor scaling, then they aren't exactly fixes for the problem, are they?

You can't have a case where they are logically fixes for the problem and yet, are not valuable -because- of the problem. This is a logic fail.

I actually edited the part you responded to out of my post before your response was posted because I didn't want to sidetrack the thread. That said, the PH-II feats are not valuable because of particularly poor scaling; they're valuable because the bonuses are so large. There was even an entire discussion earlier in this thread about why taking Epic [your weak FRW] alone isn't as valuable because you already get hit on a 2+ and now you might get hit on a 4+ or so (the first two points of the bonus being wasted).

In PH going from +5 heavy armor to +6 is worth an astounding +4 AC, while +5 to +6 neck slot is worth +1 FRW. That's vastly better for Armor than for a Neck-slot. Even in AV, you're getting +3 AC compared to +2 Fort/+1 Ref/+1 Will for Scale (only +1 FRW for Plate). This goes back to my point above: if you think that Lightning Reflexes is a "fix" because +2 to a particular FRW is worth more than +1 AC, then you might weight +4 across FRWs higher than +3 to AC.

Similarly, in PH to-hit bonuses are even harder to get from feats than boosts to defenses. This reflects the fact that +1 to-hit (not even counting the extra damage) is worth more than +1 to each FRW. For example, you could take a higher level weapon, and one Great Fort/LR/IW feat and end up close to even on FRW defenses, but there's no feat the fighter could take with a higher neck-slot item that would leave him with a comparable to-hit bonus.

 

[DracoSuave]

So, you're claiming that the feats are a fix, and -simultaneously- that the feats fix nothing?

Also, the reason I did brutes was because finding two different monsters with the same role and level at epic tier means you pick brutes. But putting that aside, every monster I tested had AC and FRW two points away from each other at the same level. You're claiming that this is not true--can you provide an example where two monsters of the same level and same role do -not- have AC attacks two points above FRW?

I've sat down, did the math. AC and FRW -are- two apart in my Monster Manual. You can dispute this only by providing a counter example. This argument is going -no where- with blanket declarations. Try some evidence. Try presenting a case using -facts-.

My theory, in truth, is simple. The FRW feats are designed to do two things: 1) Catch weaker FRW defenses up with stronger FRW defenses. 2) There's more feats boosting AC than FRW, but the AC boosting feats are of less magnitude. This is so that you -can- boost your non-AC defenses, should you choose. The same people who think AC-boosting feats are mandatory would find these feats valuable for -exactly the same reason-.

Now, regarding your 'FRW is 5 better, AC is 2 better' statement. As we've already established, the -actual- number is that FRW is 1-3 points off AC. 1-3 averages out to 2. Which means -any fix- would be in the +1-+3 range, wouldn't it? I mean, that's the -obvious- answer, fix it by the amount of disparity!

And what feats are in the +1-+3 range? The paragon boosts. The Epic feats are more to cover weaknesses you've borne up to that time.

I mean, you're assuming that the math is all based around all characters running around in Hide and/or Plate armor.... perhaps the game wasn't balanced around maximizing your AC all the time? Perhaps the disparity between FRW and AC is because when you minmax your AC, your AC is going to be higher than normal!

 

[Elric]

So, you're claiming that the feats are a fix, and -simultaneously- that the feats fix nothing?

No, my position on this has been quite consistent. Let me try to summarize my views: The preponderance of the evidence suggests that the PH-II feats are intended to fix the scaling of FRW being so poor in the PH. Using overpowered epic feats in PH-II to do this is bad.

Taking Epic [your weakest FRW] before Robust Defenses isn't necessarily that good compared to taking Epic [stronger FRWs] because you're so far behind that +4 won't help as much as you think. For many characters/opponents you won't get the full benefit of the feat because the monster would have hit you on a 1 if not for the auto-miss rule, pre-feat, so some of the benefit of the +4 from the feat will be lost. This is even more true of smaller bonuses, like a marginal +1 to your weakest FRW from a higher level neck-slot item at epic, if going by PH-alone.

Robust Defenses improves your weakest FRW without having this problem (since it also provides +4 to your stronger FRWs), and once you take it, chances are that taking Epic [your weakest FRW] is nearly as good as the others Epic FRWs, and could even be better depending on the degree of selective targeting of your weak FRW. +4 AC is going to be much better than +1 to each FRW, whether your FRWs scale reasonably well or not. AC is attacked far more than each FRW, so despite the generally worse status conditions that FRW attacks inflict, +1 to one FRW isn't as good as +1 AC.

Also, the reason I did brutes was because finding two different monsters with the same role and level at epic tier means you pick brutes. But putting that aside, every monster I tested had AC and FRW two points away from each other at the same level. You're claiming that this is not true--can you provide an example where two monsters of the same level and same role do -not- have AC attacks two points above FRW?

I've sat down, did the math. AC and FRW -are- two apart in my Monster Manual. You can dispute this only by providing a counter example. This argument is going -no where- with blanket declarations. Try some evidence. Try presenting a case using -facts-.

I never claimed this. I did claim that PC defenses tend to have AC more than 2 above FRW at level 1. For example, looking at WotC's KotS pregens gives an average of 3.7 higher AC than FRW at level. Excluding the two outliers, the Dragonborn Paladin and Human Wizard, leaves this essentially unchanged at a 3.8 average. This average is made a little higher by only having one striker among these PCs, but including, say, an Elf Archery Ranger and Tiefling Feylock wouldn't substantially alter this general conclusion.

.Now, regarding your 'FRW is 5 better, AC is 2 better' statement. As we've already established, the -actual- number is that FRW is 1-3 points off AC. 1-3 averages out to 2. Which means -any fix- would be in the +1-+3 range, wouldn't it? I mean, that's the -obvious- answer, fix it by the amount of disparity!

AC goes up by 27 from level 1 to 30. FRW goes up by an average of 24 from level 1 to 30 (22, 25, 25 in particular if you spread stat bonuses to two stats that affect different FRWs). Monster to-hit goes up by 29 from level 1 to 30. That is where I am getting the 'FRW is 5 better and AC is 2 better' statement from. Monsters gain an average of 5 more to-hit against FRW and 2 more to-hit against AC than the players gain in defensive bonuses.

It looks like you thought that my scaling argument was based on monster to-hit going up faster for FRW attacks than AC attacks, rather than monster to-hit going up relatively faster for FRW attacks relative to PC defenses than for AC attacks relative to PC defenses. I spelled this out on page 1 in my first post in the thread. Your first post was on page 1 of this thread. I'm not sure why you didn't have this in mind; this is the essential analytical framework that people who say that FRW scales poorly are working with.

Looking at your previous post makes it pretty clear you had this misconception in mind:

EDIT:

I'm going through the Monster Manual, and doing some quick comparisons between equivalent level monsters' FRW attacks and AC attacks. So far, just looking at Terrasque and Red Dragon, the pattern of 2 off seems to mostly fit. Comparing Crownwing and Storm Gorgon (both brutes of the same level) also shows a 2 point difference between an attack vs AC and vs FRW.

It appears that 'these feats are an obvious fix' isn't obvious when looked at with any level of rigor.

 

[Bayuer]

I just saw this previev:
Monster Manual 2 Excerpts: Adamantine Dragon

21 lvl solo soldier
+26 vs. REF
Highest - 7 on die
Middle - 4 on die
Lowest - 2 on die

Now the defender -2 to attack doesn't mean anything, becouse dragon makes 5 attacks and one of them will be targeted into fighter. Now this is nice monster attacking NADs.
4x 1d12 +8 dmg on hit plus one 2d8 + 8 damage, and ongoing 10 damage (save ends) in turn

 

[keterys]

It's got notable burst damage, too - I mean in true silliness you can have start of combat bomb of frightful presence to stun and get another +2 to attacks, AP for breath, next round AP again and do 8 claws... so 8d12+64+3d12+6+15 or 157.5 damage at 90% hit... or enough to drop most any non-defender, which is pretty decent.

More damage if you just claw, but the fright + breath are cool to use

 

[Jhaelen]

Alright. So I took a step back and tried to look at these issues with a fresh mind.

What is complicating this discussion is that you cannot isolate a single aspect (PC defenses) and make any conclusions about the underlying math of the system.

Some aspects that are relevant to this discussion:
1) The gap between AC and the other defenses
2) The gap between a character's highest and lowest defense
3) The gap between the PC defense progression and the monsters' attack bonus progression

To tackle these issues, I wanted to take a slightly different approach and look at some concrete examples. So I created example characters for all the PHB1 classes using PHB1 races using the standard array (16, 14, 13, 12, 11, 10). I then examined their stats at levels 4, 14, and 24 and compared them against sample encounters straight out of the MM1.

For level 4 I used the following encounter:
- 1 human mage
- 2 human bandits
- 2 gravehound zombies
- 3 zombie rotters

For level 14 I used the following encounter:
- 1 githzerai mage
- 1 githzerai zerth
- 4 githzerai cenobites

For level 24 I adapted the encounter template for the flameskull:
- 1 greater flameskull
- 1 eladrin lich
- 2 rakshasa dread knights
(I did this because the greater flameskull has been mentioned as an example before)

I assumed a party of five covering all roles plus a second striker, because I think that's the most common setup. I didn't make any attempt to min-max them, I just did some of the more obvious optimizations.

I've scribbled some 6 pages full of notes but I don't want to bore you with the minor details (and I'm too lazy to type it all up) and just tell you about my findings.

At this point, I'd like to thank DracoSuave for his post #179. What he wrote is similar to my calculations.

Getting back to the different aspects I mentioned above:
1) The average gap between AC and other defenses remains fairly constant across levels. Defenses are on average 3-5 lower than AC.

2) The average gap between highest (non-AC) defense and lowest increases across levels. Depending on how well the selected race fits to the important abilities for a given build the gap starts with a difference of 2-4. This difference increases by 1 with each tier, so at epic level the gap is 4-6.
As DracoSuave already mentioned, this is easily compensated by taking one of the PHB1 paragon feats granting +2 to a defense.

As an aside: Though obvious in hindsight, I found it interesting that the gap will be larger if the race fits well to the build, assuming you use your stat increases to only increase the two most important abilities for your build.
I think for builds with a slight MAD it might be a reasonable variant to use your second stat increase to alternate between two abilities.

3) The average chance for the monsters to hit characters increases across tiers:
At heroic tier: 45% (30-35% to hit AC, 55-60% to hit other defenses)
At paragon tier: 60% (50-55% to hit AC, 65% to hit other defenses)
At epic tier: 65% (45-50% to hit AC, 80% to hit other defenses)

For characters it's the opposite: their average chance to hit drops across tiers:
At heroic tier: 80%
At paragon tier: 60%
At epic tier: 45%

Then I used those numbers to calculate the (normalized) average damage output:

For a monster:
- 4.5 at heroic levels,
- 9 at paragon level, and
- 14 at epic level

For a pc:
- 9 at heroic levels,
- 12 at paragon levels,
- 15 at epic levels

One thing that's worth noting: I didn't include anything but static effects in the calculations for the pcs. I also only used the characters' at-will powers for this.
I did assume, though, that the party's strikers would be able to add their bonus damage on every attack.

Next I compared the monsters' total hit points to the party's total hit points. I added 50% to the party's hit points, the equivalent of two healing surges (without applying any other bonuses).

Total monster hp:
- 223 for the level 4 encounter,
- 785 for the level 14 encounter, and
- 1066 for the level 24 encounter.

The level 24 encounter is worth mentioning because both the flameskull and the lich can regenerate hit points, the lich can heal itself once, and the rakshasas have an aura that makes healing surges less effective.

Now I looked at the expected number of rounds these combats would take and decided that at least for the paragon and epic encounters critical hits might be relevant for the calculation (the expected number of rounds is 13 - 14, i.e. there'd be about 6 - 8 crits).

I also recalculated the DPR taking some of the monster abilities into account:
- I assumed area effects would on average hit half of the party
- I added the lich's aura damage
- I took the rakshasa's increased accuracy into account
- I included the effect of the githzerais' trace chance ability

The new DPR numbers:

For a monster:
- 5 at heroic levels,
- 9.5 at paragon level, and
- 15 at epic level

For a pc:
- 9.5 at heroic levels,
- 13 at paragon levels,
- 16.5 at epic levels

The net result, though remains the same: comparing the dpr values the party will win every encounter even if they're using only at-will powers!

This leaves one question open: what about non-numeric affects, like the daze powers that are involved in some of these encounters?

I actually think they don't matter a lot. I'd rather say the odds are stacked in the pcs' favour when you're looking at paragon or even epic tiers:
Pcs start out with fewer or the same number of powers than monsters and quickly overtake them:
- at L4 a pc has at least 2 encounter, 1 daily, 1 utility powers and can use 2-3 item dailys*
- at L14 a pc has at least 4 encounter, 3 daily, 4 utility powers and can use 3-4 item dailys*
- at L24 a pc has at least 4 encounter, 4 daily, 6 utility powers and can use 4-5 item dailys*
*: depending on the number of milestones reached. I think 3-5 combat encounters should be feasible for a party. They'll typically also have at least one skill challenge per day.

This means, the party as a whole can use an average of 3, 6, or even 8 daily powers in every encounter.
Considering some of the very powerful effects that can last for the entire encounter, I have a hard time believing epic parties will have any trouble with level equivalent encounters.
Add the paragon and epic abilities and things look better still! They'll have ways to deny enemy actions and ignore attacks and damage or negate it with healing effects.

As I mentioned in my previous posts, I also expect experienced players to be prepared for all kinds of encounters. E.g. they _will_ have ways to deal radiant damage to the lich every round, even if they don't have a divine character in the party (which for a pure PHB1 party is unlikely).

This is something I've seen in my 3E campaign a lot:
When encountering a certain monster type or specific combat tactic the first time, my players may have a hard time. But they immediately realize this, analyze the reasons for their problems and take steps to prevent having these problems ever again:
They'll retrain feats & skills, learn new spells and powers, buy items, and/or take levels in other classes.

In 4E they'll do exactly the same thing and probably be doing fine.

I'd only expect problems if
- an encounter is of a significantly higher level than the party, and/or
- the number of enemies with action-denial abilities is higher than the number of pcs.

Anyway, my conclusion remains the same: the math works.

It just works in different ways than people expect. The guidelines as presented in the DMG1 are somewhat misleading, because they don't give you the whole picture. But wasn't that to be expected, considering the goal of these guidelines? I.e. to present a simple rule-of-thumb which can be used to modify existing monsters and create new ones.

So what's the reason for the PHB2 feats if they aren't a math fix?
Imho, they are a fix for bad or disadvantaged character builds.

Choosing a MAD build with two abilities that add bonuses to the same defense and min-maxing the character can lead to problems in the epic tier that a player may not have been aware of when creating the character. So the feats are a way to fix those specific problems particular characters may have without the player having to create a new character from scratch.

But they're neither required nor must-have for balanced characters created with a bit of foresight. Players worried about their defenses may of course take them to be hit (slightly) less often, but this is only a perceived problem.

Likewise Weapon Expertise is not required by game math but it _will_ help to take some of the grind factor away that will become noticeable at high levels. Shortening combats will probably mean they're more enjoyable for players. But it will also make epic combats a lot easier. Then again, they'll allow a DM to use more difficult encounters.

So, all in all, they may have a positive net effect.

 

[Elric]

Getting back to the different aspects I mentioned above:
1) The average gap between AC and other defenses remains fairly constant across levels. Defenses are on average 3-5 lower than AC.

As has been stated many times, by default over 29 levels PCs lose 2 on AC relative to monster to-hit and lose 4 on their two strong FRWs relative to monster to-hit and 7 on their weak FRW relative to monster to-hit (for an average FRW loss of 5). So AC should scale better.

3) For characters it's the opposite: their average chance to hit drops across tiers:
At heroic tier: 80%
At paragon tier: 60%
At epic tier: 45%

These numbers are much larger than the general trends. Players don't lose 7 points of to-hit bonus by level 24 compared to level 4. They only lose 4 to-hit over 29 levels as to-hit goes up 25 while monster defenses go up 29. You're comparing monsters with different roles and elite/non-elite status, which alters this quite a bit. AC is essentially constant at 14+level for the average monster in the MM at all three tiers; see http://www.enworld.org/forum/d-d-4th-edition-rules/229092-lots-statistics-monster-manual.html (indeed, the DMG, in the creating monsters section, has Skirmishers, the baseline combatant in terms of numbers, at AC=level+14).

Then I used those numbers to calculate the (normalized) average damage output:

For a monster:
- 4.5 at heroic levels,
- 9 at paragon level, and
- 14 at epic level

I also recalculated the DPR taking some of the monster abilities into account:
- I assumed area effects would on average hit half of the party
- I added the lich's aura damage
- I took the rakshasa's increased accuracy into account
- I included the effect of the githzerais' trace chance ability

The new DPR numbers:

For a monster:
- 5 at heroic levels,
- 9.5 at paragon level, and
- 15 at epic level

The epic level numbers are clearly wrong when compared to each other. If the lich's aura hits one PC, that's +5 damage per round, across 5 monsters= +1 DPR/monster. The Rakshasa ability is worth +25 percent points hits if base to hit is 50% (an essentially identical +24.75% at 45% to hit); at 6 attacks across the 2 Rakshasas, this averages 1.5 extra hits, for 11.5 each, which is over 15 extra damage. Without the lich's aura, we're already at +16.75 damage over 5 monsters, which is at least +3 DPR per monster per round, not the +1 you have listed. So given all of these easy to spot errors, I tend to doubt your numbers in general.

You also seem to have missed that the Githzerai Cenobite has an at-will attack that stuns. This is a rather important ability. In short, this type of analysis with encounters at different tiers is tough to adequately control for other variables with in the best of circumstances, and this isn't a good example of the genre.

 

[KarinsDad]

Here's a simple explanation. Those feats are parallels for the equivalent Epic feats from 3.x. People wanted feats to bolster their stuff, Wizards decided to do so. There's no evidence presented that it's a 'fix' when Wizards has shown -multiple times- that they're willing to institute fixes with errata.

Err, wrong.

There is one piece of designer evidence that has been presented, but has been totally ignored by people who do not believe that WotC is creating a feat math fix.


WotC stated on multiple occasions before the 4E release that they wanted the sweet spot to work at all levels.

Not just heroic levels, all levels.

There are certain band of levels where sort of the math and complexity of DND is just right. Characters have the right number of options, monsters have the right complexity, combats last the right number of rounds, rounds of combat last the right amount of time.

So what we want to do is find the realities of the game that make those places the sweet spot and replicate that across the band of levels.

YouTube - D&D 4th Edition: Behind the Scenes, Part 1

We're making all levels of play equally fun and playable.

YouTube - Dungeons & Dragons 4th Edition: Part 1

PCs do not get hit on a 2 in the heroic tier. PCs do not need a 17 to hit in the heroic tier. It's extremely unlikely that the designers think that getting hit on a 2 is fun. It's extremely unlikely that the designers think that 20+ round encounters are fun.

The sweet spot claims by the designers is solid evidence of designer intent and goes against the claim that there "is no math problem".


They did not want 8 round encounters at low level and 20 round encounters at high level. They said that problems like these caused the lack of sweet spot below level 7 and above level 14 in the 3.5 game.

Listen to the GenCon announcements and interviews to find out designer intent.

This information is straight from the designers.

 

[Bayuer]

@Jhaelen
You made some math. That's nice. But:
1)Compare AC to DEF of PCs and damage that is dealt with both attacks.
Flameskull (24 lvl Artillery):
+28 vs. REF; 2d8+10 dmg and the target is dazed until end of next monster turn.
Highest (36) - 8 on die
Medium (33) - 5 on die
Lowest (29) - 2 on die

Fell Wyvern (24 lvl Skirmisher):
+29 vs. AC; 2d8+10 dmg or 2d6+10 dmg and knock target prone
Plate (without shield) (40) - 11 on die
Hide (without shield) (37) - 8 on die

EDIT: There is also Raksasha Dread Knight on 24 lvl (Soldier)
+29 vs. AC; 1d8 + 7 damage (Rakshasa makes two attack rolls) and she can make 3 attacks if two hit the same target he is dazed (save ends)
Plate (without shield) (40) - 11 on die
Hide (without shield) (37) - 8 on die

When you campare Flameskull attack to Wyvern attack you can see that they are the same but Flameskull gives also an effect! My question is:

Why attacks that give the same amount of damage + effects are fine to you? You can easily see that the gap is too huge to simply say "this what it suppose to be!". Monsters that hit AC are so... useless, becouse they hit chance is lower and they do the same amount of damage. If hitting on 2-5 was what 4E designers have in they minds, why they did AC targeting of monsters attacks on Epic tier? They should just do all attacks target NADs or scall AC the same as NADs. The fact they didn't... Bingo. They just forget about NADs and then PHB2 comes out and we have defense feats to fix this issue.

Even if you will still be saying that this is what it suppose to be (you must just don't see the obvious facts in this thread if so):
2)Why we have PHB2 feats? they are too good! +6 to NADs on Epic? Man. It's huge! So:
a) Feats are broken
b) They fix the math in very bad way

If I haven't seen the Adventurer's Vault masterwork armors fix I can say maybe a) is true. But when they fixed something that was already good (compared to NADs attacks) the obvious thing is that only b) can be correct.

 

[Jhaelen]

The epic level numbers are clearly wrong when compared to each other. If the lich's aura hits one PC, that's +5 damage per round, across 5 monsters= +1 DPR/monster. The Rakshasa ability is worth +25 percent points hits if base to hit is 50% (an essentially identical +24.75% at 45% to hit); at 6 attacks across the 2 Rakshasas, this averages 1.5 extra hits, for 11.5 each, which is over 15 extra damage. Without the lich's aura, we're already at +16.75 damage over 5 monsters, which is at least +3 DPR per monster per round, not the +1 you have listed. So given all of these easy to spot errors, I tend to doubt your numbers in general.

You also seem to have missed that the Githzerai Cenobite has an at-will attack that stuns. This is a rather important ability. In short, this type of analysis with encounters at different tiers is tough to adequately control for other variables with in the best of circumstances, and this isn't a good example of the genre.

Thanks for looking over the numbers!

You're right, I did make a couple of mistakes. One of them being that I first added up all damage and then multiplied the whole lot by the average to hit chance. This is obviously wrong if some of the damage doesn't have a miss chance, like the lich's aura.

Regarding the rakshasas, it seems I forgot to include the additional damage I had calculated for the increased accuracy and more frequent crits. When I posted, I didn't use the totals I had calculated before; instead I added up the various damage sources again which gave me smaller numbers.

In my original notes I had a dpr increase of 6.5 but the dpr increase for pcs was higher, as well. I didn't these because I thought I had made an error in the original calculation

But as mentioned, these numbers won't be completely correct either...

Regarding the stun at wills of the githzerai cenobites:
Yep, I did notice them, but how do you include these in a dpr calculation?