Damage per second

[Leyoni]

A lot of discussion on this and a lot of rightful focus. I am trying to find out some critical information.

I have read the DDO wiki on attack chains so I understand that with each attack animation that there are varying numbers of actual attacks that are executed.

I also understand that animation speeds differ so that some attack chains actually execute faster than others.

It seems to me that the speed at which an attack chain executes determines the actual number of attacks over time. It also seems to me that if two different attack chains running at different animation speeds take place in parallel then there will be distinct points at which they will coincide at start, a period during which they are out of sync with one another, and a point at which the start of the animations and attack chains are once again the same.

By measuring the length of time between the points when the chains are in sync and counting the number of attacks that occur in the two different attack chains we should be able to come to a reliable count of attacks over time that will let us make valid comparisons.

I don't think that these animations take place in 1 second bursts or that the point at which they converge is 1 second in length. In fact, I think it highly unlikely that this is the situation. I find it completely more probable that in a 60 second period that there will be some number of convergences and that by looking at the length of time for those convergences it is possible to make a more accurate assessment of attacsk per second (and thus to make a more accurate assessment of damage per second).

It also seems likely that this entire exercise has been done in the past by others. However, searching the forums for the information is painful. So, I am hoping that they will repost information (or links) to this thread.

Thx in advance

[Leyoni]

At L16 a full BAB character gets 5 attacks per animation. TWF increases this to 7, ITWF to 8 and GTWF to 10 (all at reduced BAB). If the attack animations are different speeds then the actual number of attacks in a set length of time may be higher than these numbers suggest.

For example, if the single weapon animation runs 12 times in 10 seconds but the TWF animation runs 9 times then the single weapon actually receives 60 attacks (6 per second) to the TWF's 63 (6.3 per second). This would say that there is less benefit to TWF than is generally presumed based on the 7:5 ratio that we would expect.

OTOH, if the single weapon animation runs 10 times in 10 seconds but the TWF animation runs 11 times then the benefit to TWF is greater than we would presume.

Obviously we can make comparisons based on animation sequences and numbers of attacks in the attack chain. But, that may not actually equate to damage per second. It could be that dps is a poor term and that we should actually talk dpa (damage per animation).

I would actually like to know in order to seriously address some of the dps claims that are being made on the forum.

Again, as always, thx in advance.

[Angelus_dead]

I don't think that these animations take place in 1 second bursts or that the point at which they converge is 1 second in length. In fact, I think it highly unlikely that this is the situation. I find it completely more probable that in a 60 second period that there will be some number of convergences and that by looking at the length of time for those convergences

Yes, convergences exist- but to accurately count them would take more effort than to directly count the number of attacks you make in 30 or 60 seconds. Its harder to watch two separate characters than to pay close attention to just one. To make counting attacks easier, don't try to count each one- just count when your character makes the first attack in the animation, then multiply by 5 or 10 or however many are in his combo. (You'll have to add the attacks from the partial final animation on top of that)

Anyhow, here's a summary of the most interesting numbers, which are swings/minute at BAB 15+.
greatsword: 76
greataxe: 81
staff: 84
1wf: 89
ius: 96
utwf: 150
twf: 166

Remember that two-hand weapons get 2 glancing blows each 5 attacks, which do around 30% the physical damage of a regular attack. IUS stands for Improved Unarmed Strike, which is what monks have. It's also a less-accurately tested number than most of them. UTWF is the new unarmed TWF in module 8, and it's guesswork from developer posts.

[Leyoni]

Thanks A_D.

I've seen most people make similar observations regarding the time/effort issue. And I appreciate your information because it represents sort of the default approach that people seem to take to this.

If you don't mind, I'd like to look at two of your numbers. For 1wf you list 89 swings in 60 seconds and for twf you list 166. I am assuming that twf is just that (the 7 attacks per animation) and not any of the others.

We know that there are 5 attacks per animation for 1wf and 7 for twf. Neither number is divisible by its number of attacks w/o remainders (89/5 = 17 r 4 and 166/7 = 23 r 5). What we should expect is that these two numbers (17+ and 23+) are actually the same -- that is, both are ~17 or both are ~23. At 166 attacks per minute this is actually a closer number to gtwf (166/10=16 r 6) than twf. Do you perhaps mean gtwf?

If so then 16 r 6 animations to 17 r 4 animations is much closer to what should be happening (ideal would be for both numbers to be exactly the same).

If the number is for twf (or even itwf and not gtwf) then it indicates a problem in the game because it distorts twf and makes it more powerful than it ought to be. OTOH, if it is for gtwf then the numbers are very close and differences might be attributed to observation errors or to basic programing expedients that make a closer sync more trouble than it is worth.

[Angelus_dead]

We know that there are 5 attacks per animation for 1wf and 7 for twf. Neither number is divisible by its number of attacks w/o remainders (89/5 = 17 r 4 and 166/7 = 23 r 5)

Why would it be evenly divisible? There's no guarantee that the animations will perfectly line up on the 60 second mark.

At 166 attacks per minute this is actually a closer number to gtwf (166/10=16 r 6) than twf. Do you perhaps mean gtwf?

Naturally I mean the TWF style, which at BAB 15 requires you to have the GTWF feat.

A BAB15 character with only TWF feat (or no feats) would have 116 attacks per minute. ITWF would mean 133 apm.

[Leyoni]

A_D, thanks again.

No, I wouldn't expect the animations to line up exactly in 60 seconds, that wasn't what I was getting at. What I would expect is that the number of animations in 60 seconds should be roughly the same regardless of weapon style.

GTWF gives 10 attacks per attack chain while 1wf gives 5 (at BAB 15+). So what I would expect is say 90 attacks in 60 seconds for 1wf and 180 in 60 seconds for gtwf -- that is, a 1:2 relationship mirroring the 1:2 ratio in the attack chain.

What you show is actually a slightly lower number of attacks for gtwf than we should expect. This means that gtwf is slightly underpowered in DDO. Again, the difference is not especially significant and may be attributed to practical reasons related to programing.

But, it is important in making a dps argument because it indicates the actual relationship between attacks using the different weapon styles.

How is this relevant?

A lot of recent discussion has been on the dps ability of DEX based twf builds. These builds typically opt for high DEX and WIS while setting STR at relatively low numbers (8-12 being typical, 13-15 being somewhat more uncommon). Obviously this is a necessity because some stats need to be dump areas if a build is really going to maximize DEX/WIS (generally for the AC advantage).

In contrast, armor wearing builds tend to have low WIS and only enough DEX to get full DEX bonus from MFP. Instead they have significantly more STR.

In making a dps argument for the DEX/WIS build it seems to me that people often confuse the number of attacks with damage. Again, going to your research, 166 attacks compared to 89 seems like a huge advantage. But, I'm not certain that it is.

Assume every attack hits for the entire time -- no misses in either case. Typical DEX builds using finesse weapons do d6 damage per hit. It is not atypical for STR builds to use non-finesse weapons with typical damage being d8 or more. Let's assume a d8.

Remember that we are saying every swing hits. So, the 166 swings give 166 hits with an average damage of 3.5 points per hit for 581 points of damage. Likewise, the 89 swings give 89 hits with an average damage of 4.5 for 400.5 points of damage.

Advantage to the gtwf. But, STR is added to damage. If the STR difference is +4 for the 1wf that is 2 points more damage per attack -- 178 points. All of a sudden the advantage has all but disappeared. If the STR difference is +6 then the 1wf is doing more damage.

As I look at final STR numbers I typically see twf builds with final STR in the 20-24 range. In contrast, STR based characters typically end in the 30-32 range. This represents a 3-10 point advantage during damage calculation for the STR based characters and shows that their actual damage output is much higher.

Naturally, this does not calculate critical hits and with more attacks and more hits the gtwf character should get more criticals. But, critical damage multiplies STR bonuses so a similar effect is taking place with critical hits to that with regular hits -- the STR based character makes up most or all of the perceived lost ground because they are working with larger base numbers to start with. It is also important to look at critical multipliers and threat ranges. Rapiers (favored finesse weapon) are x2 with a threat range of 18-20 while kopesh (favored 1hw) is x3 with a threat range of 19-20 and dwarven axe (another populare 1hw) is x3 with a threat range of 20.

When comparing it is necessary to take these types of differences into consideration. In the case of gtwf this means that 3 in 20 attacks is likely a critical for the typical rapier wielding gtwf build. Assume each potential actually is a crit. That means in 166 attacks there are ~25 criticals each doing an additional 3.5 points of damage -- total bonus damage is 87.5 points. (Note that this gives the benefit of the doubt that all of the criticals occur on the rapier and it is unlikely that the character is actually dual wielding rapiers. It means the result is skewed slightly high.)

Meanwhile a kopesh using 1wf is getting 2 criticals every 20 attacks for an additional 9 points of damage. In their 89 attacks this is ~5 criticals. The bonus damage is 45 points. Again, consider a +4 STR advantage to the STR build. This +2 advantage is multiplied for crits as well and results in 10 points more damage. Advantage still to the gtwf build.

Make the STR advantage +6 and the gtwf build's damage difference disappears.

GTWF build with criticals does (from earlier in the post) 581+87.5=668.5 points of damage during its 1 minute.

1WF build with criticals and +6 STR advantage does (again, from erlier in the post) 667.5+60=727.5 points of damage during its 1 minute. [667.5 is 400.5 plus 89*3, the bonus of +6 STR to damage. The 60 is 45 + 15, the bonus to crit damage of +6 STR.]

This shows the gtwf build does ~11.1 points of damage per second compared to the 1wf ~12.1.

Obviously weapons will make a difference as will other add ons -- things that I have elsewhere called ancilliary damage. However, there is nothing in the damage arena that is unique to one build over the other when we start talking about these add ons. They will be exactly the same for either build (or are potentially in any case).

The only factor I have not yet calculated is for when the threat range is extended due to keen weapons or improved critical feats. To do that means taking into account the gtwf's actual off-hand weapon. I think that the advantage should shift slightly to the gtwf build -- but I'm not positive.

This modeling also assumes every attack hits and every critical is confirmed. Changing those assumptions will make a difference in the results. In this case I think the advantage should shift to the 1wf fighter although, again, I'm not positive.

What it does show is that the difference in dps between the two is almost negligible as long as the STR difference remains in the +6 area. It moves significantly in favor of 1wf as the STR difference goes higher.

[Angelus_dead]

What you show is actually a slightly lower number of attacks for gtwf than we should expect. This means that gtwf is slightly underpowered in DDO. Again, the difference is not especially significant and may be attributed to practical reasons related to programing.

That doesn't mean it's "underpowered". It's not underpowered in practice, and it's also not underpowered compared to some promise made about what TWF should be able to do- no game documentation told you it would allow 2x attack rate. It's especially not underpowered compared to TWF in the D&D rulebooks, which has many serious drawbacks. (And which, in fact, is underpowered in that game until you bring in cheesy exploits)

A lot of recent discussion has been on the dps ability of DEX based twf builds.

There hasn't been a lot of discussion on that. TWF characters are often analyzed, but they're not biased towards dex as opposed to strength. In fact, the dex-dump Strong Ranger is a major archetype, as is the level 20 Tempest Barb.

In contrast, armor wearing builds tend to have low WIS and only enough DEX to get full DEX bonus from MFP.

Actually, the dex to qualify for GTWF is higher than to use MFP, so that's what their stats are limited by.

In making a dps argument for the DEX/WIS build it seems to me that people often confuse the number of attacks with damage. Again, going to your research, 166 attacks compared to 89 seems like a huge advantage. But, I'm not certain that it is.

It is. Even if you ignore that some monsters are killed by vorpalling or wounding and not hp damage at all, it still is.

As I look at final STR numbers I typically see twf builds with final STR in the 20-24 range. In contrast, STR based characters typically end in the 30-32 range. This represents a 3-10 point advantage during damage calculation for the STR based characters and shows that their actual damage output is much higher.

Take the extremes of the range: dex guys with 20 str, str guys with 32.
High str: +11 damage main hand, +5 offhand
Low str: +5 damage main hand, +2 offhand

The high-str 1wf has 11*89 = 979 damage from str.
The high-str twf has 11*83 + 5*83 = 1328 damage from str.
The low-str twf has 5*83 + 2*83 = 581 damage from str.

Naturally, this does not calculate critical hits and with more attacks and more hits the gtwf character should get more criticals.

Critical hits apply the same factor to 1wf and twf characters, so they don't change the results between those two fighting styles. Including crits would approximately raise both sides of the comparison by 25%.

However, crits do change the results between str twf and dex twf, because the strong guy has a khopesh while the dex guy does not. This is a minor flaw in DDO's adaptation of the D&D rules, because in the source rules there is no non-finesse weapon with superior crit power to the best finesse weapons (rapier and kukri). There are a variety of ways that problem could be addressed, but str vs dex TWF is not your topic of interest.

In the case of gtwf this means that 3 in 20 attacks is likely a critical for the typical rapier wielding gtwf build. Assume each potential actually is a crit.

That line of thinking only matters for a weapon like Radiance, which gives you a damage bonus on the attacks following the crit. That's too rare and complex for your interest.

Obviously weapons will make a difference as will other add ons -- things that I have elsewhere called ancilliary damage. However, there is nothing in the damage arena that is unique to one build over the other when we start talking about these add ons. They will be exactly the same for either build (or are potentially in any case).

No, that is completely wrong.

So-called "ancillary" damage additions are absolutely and positively biased towards the character who gets more attacks per minute, which is TWF (and Tempest). Holy gives you +7 damage per hit, so it provides a 1wf guy with 89*7 = 623 dpm, the twf guy gets 166*7 = 1162 dpm.

Basically, any buff or enhancement which adds +X to a hit will add 89X dpm to 1wf and 166X dpm to twf. That means TWF guys gain 86% more benefit from bard songs, sneak attack goggles, seeker, greater bane, and nearly anything else you could name (except Rage)

[Leyoni]

No, that is completely wrong.

So-called "ancillary" damage additions are absolutely and positively biased towards the character who gets more attacks per minute, which is TWF (and Tempest). Holy gives you +7 damage per hit, so it provides a 1wf guy with 89*7 = 623 dpm, the twf guy gets 166*7 = 1162 dpm.

Basically, any buff or enhancement which adds +X to a hit will add 89X dpm to 1wf and 166X dpm to twf. That means TWF guys gain 86% more benefit from bard songs, sneak attack goggles, seeker, greater bane, and nearly anything else you could name (except Rage)

Thanks again A_D. This is becoming a two-way conversation but an enlightening one.

[Desteria]

Where TWF gets it's HUGE DPS bost is from added effect on every hit Stuff like:
Holy +2d6 or 7 average
Sneak attack (tharns OR Haflings) +8 A rouges is a LOT MORE as a haflign with tharns
Bard song +6
Prayer +1
Force Ritual +1
Slicing 1d4
Thast some basic numbers most melle shoudl have most of the time in DPS fights

Lets do a little math I'll use 2 of my chars for examples my Dwarf THF fighter 33str[stupid +3 tomes don't exist I swear], and my hafling 8 starting str dex wis monk, I'll use harry for the example takign in to account his 50% fortification
I'll ignore tempest haste adn fighter haste boost over a typical haryr fight they average fairly similar... I'll pretend my fighter has recrafter his sword in to an AXE beacuse it's better that way.

Fighter 81 swings all hit.
Average Damage per swing 3d6 base +5weapon +2d6 holy +1d4 Slicing +1 force +8/2=4 sneak tharns +6 bard +1 prayer +10 power attack +17 str +2 rage str +1 madstoen rage str, +2 second madestone rage str +3 favored enimy == 72 average damage
5832 Damage
add in crits 2 per 20 swings /2= 4 crits damage per crit
Average Damage per CRIT (3d6 base +5weapon +6 bard +1 prayer +10 power attack +17 str +2 rage str +1 madstoen rage str, +2 second madestone rage str +3 favored)x2 +18 bloodstone average additional damage = 133x4= 532
Total Damage Str based THF 6364

TEmpest monk 166 Swings all Hit (averaging str to 75% beacuse of full on main hand and half on off hand)
Average damage per swing 1d8 base +5 weapon +2d6 holy +1d4 slicing +1 force +16/2 =8 Thans+haflign sneak +6 bard +1 paryer +0 combat expertise +1 Rams +5 favored +4.5 str (8+2tome+6item+2rams+2rage+2madestone=22+6*0.75 not addign secodn madestoen becasue ac builds dont get hit enough) == 55 average damage
9130 Damage
Add in crits 2 per 20 swings/2= 8 crits damge per crit
Average damage per CRIT 1d8 base +5 weapon +6 bard +1 paryer +0 combat expertise +1 Rams +5 favored +4.5 str (8+2tome+6item+2rams+2rage+2madestone=22+6*0.75 not addign secodn madestoen becasue ac builds dont get hit enough) == 27 additional average damage = 27x8= 216
Total Damage DEX/WIS based TWF with KAMAS!!!! 9346

Sooo TWF is ruffly 3000 damage ahead or an extra 50% more damage when comparign a LOW DPS TWF build to a 'high' DPS THF fighter Now just thing if that twf was a STR ranger with FULL favored enimy using ohh Kopeshes insted of Kamas and havign power attack!!!!!!!!!!!!!!!!!!!


This IS why peopel say TWF is much better DPS it's not even close IF it really did attack at full twice the speed OMG my mind explodes thinging of how much further of fit would be.

It's alll abotu the effects that add damage to every attack and there are a fair muber of thouse in D&D/DDO.


Hope this helps you in your investigation of DPS or as the case me be DPM

[Desteria]

Thanks again A_D. This is becoming a two-way conversation but an enlightening one.

DAngit AD types faster then me well i said the same thign basicaly though with a specfiic example of my built for DPS THF fighter my build fro AC monk who out damages him by 50% due to TWF.....

[Angelus_dead]

Where TWF gets it's HUGE DPS bost is from added effect on every hit Stuff like

That is why I have previously suggested the devs add some magic properties which give +X damage on a 1h weapon but +1.5X or even +2X on 2h weapons.

[oogly54]

That is why I have previously suggested the devs add some magic properties which give +X damage on a 1h weapon but +1.5X or even +2X on 2h weapons.

They already do, with power attack, 1.5X the STR damage from power attack.

What all this argument is assuming is that every attack is hitting. A high STR fighter will have a much higher to hit bonus that a TWF from higher STR and feats and BAB from monk splash. On my fighter, which has a +37 to hit unbuffed, he still DOES miss on more than a roll of 2. If I have ~+6 better attack bonus and still miss on a roll of a 2, then the TWF will miss on a roll of an 8 or 40% of the time with that attack sequence swing.


http://forums.ddo.com/showthread.php?t=139604 Iron Monk Link
Even Illuminattis's Iron Monk has a 20 STR, 34 DEX, and 20 WIS. For a 60 UNBUFFED AC.

His attack bonus is:
Attack: (Rapier/Short Sword)
+15 BaB
+12 Dex
+5 Weapon
+4 Greater Heroism
-2 TWF
-5 Power Attack/CE
+1 Haste
+1 Halfling
------------------
+31 (w/ Greater Hero / Power Attack/CE)

OR only a +26 unbuffed attack bonus, with a 34 DEX. Compared to a STR based fighter with a +32 with CE/PA. The STR based RANGER will have an even lower attack bonus from having either a lower DEX or lower STR (finesse or not).

What also is seldom discussed is the fact everyone believes that a monk/ranger/whatever will have the highest DPS AND the highest AC at the same time. In order for this character to get to the highest AC it must have high dex and wis, sacrficing STR. When you sacrific STR you lose DPS, when you increase DPS you lose AC. You can't have all three stats, STR, WIS, DEX, all really high at the same time. IF you try, you gimp all three stats and moderating them all, but what about CON and HP?

So, to make a long story short, ranger/monks do NOT hit on all 10 attacks and miss at a much higher rate than a fighter/barb, especially at the high end elite quests, which most of us build our characters for. Also the ultimate AC AND DPS is not feasible and if you try for high of both you actually hurt your to-hit bonus even more causing your DPS to drop even more. A ranger can have a really high DPS and a decent AC. A ranger/monk can have a really high AC, but only a moderate amount of DPS.

Thank you,
Oog

[krud]

I never trusted that 89 attacks/min for 1 handed fighting so I had to test it myself. In test after test my bab 14 rog mutt gets 94 swings per minute (23.5 animations*4 attacks/animation) with a rapier, longsword or scimitar. I thought surely i can do even better with a higher bab, so I grabbed a divine power clickie and checked again. I only get 90 attacks/min (12animation/40sec =--> 18 animations/min*5attacks/animation). My DPS is worse at a higher bab? maybe I shouldn't level him up when the cap goes up, lol!

btw - oog, what's the AC of these elite raid bosses?

[Angelus_dead]

I only get 90 attacks/min (12animation/40sec =--> 18 animations/min*5attacks/animation). My DPS is worse at a higher bab?

Yes, that's right. Many combat styles experience DPS decreases as they advance in level. I thought this was well known, but I guess players haven't been talking about it in a while.

I won't bother typing out the litany of errors that must underlie a mistake like that, you can figure it out yourself I'm sure.

[vtecfiend99]

Yes, that's right. Many combat styles experience DPS decreases as they advance in level. I thought this was well known, but I guess players haven't been talking about it in a while.

I won't bother typing out the litany of errors that must underlie a mistake like that, you can figure it out yourself I'm sure.

I would rather you did, in a whole new thread dedicated to how crazy that is.

[Angelus_dead]

They already do, with power attack, 1.5X the STR damage from power attack.

Feats are not "magic properties".

What also is seldom discussed is the fact everyone believes that a monk/ranger/whatever will have the highest DPS AND the highest AC at the same time.

Hardly anyone believes that.

What is believed are two separate claims:
TWF guys can have more AC and more DPS than an S&B guy.
TWF guys have more DPS than a THF or S&B guy.

So, to make a long story short, ranger/monks do NOT hit on all 10 attacks and miss at a much higher rate than a fighter/barb, especially at the high end elite quests, which most of us build our characters for

Hello, I'd like to introduce you to my friends Warchanter, Recitation, and Breastplate of Destruction. Oh, and look, it's Ooze Sunder, Unbalancing Strike, and Eagle Claw Attack. Hi everybody, get ready to apply a total effective 9+2+4+4+2+4 = +25 to my attack rolls.

[Borror0]

The high-str 1wf has 11*89 = 979 damage from str.
The high-str twf has 11*83 + 5*83 = 1328 damage from str.
The low-str twf has 5*83 + 2*83 = 581 damage from str.

A slight problem in this comparison is that it doesn't take into account all the bonus that will favor TWF over 1WF.

[Angelus_dead]

btw - oog, what's the AC of these elite raid bosses?

The old raid bosses were AC 35 and went up +2/+4 on hard/elite. Newer raid bosses (module 5+) are more like AC 43, and may go up more than +2 on hard.

A common player character might have BAB 16 + 10 str + 5 weap + 9 bard + 4 hero + 2 recit + 2 axe + 2 flank + 5 tharne = 55, meaning he could roll a -12 and still hit the boss. Of course, maybe he doesn't have Tharne Goggles, so -5, but maybe his strength is above 30, so plus whatever. And as already mentioned, the other players are likely to do something to decrease the enemy AC. In a raid of 12 guys, it's unlikely that nobody is wearing Breastplate of Destruction.

[Angelus_dead]

A slight problem in this comparison is that it doesn't take into account all the bonus that will favor TWF over 1WF.

That's outside the scope. Strength bonuses ONLY were under consideration.

[Borror0]

They already do, with power attack, 1.5X the STR damage from power attack.

Right, but that is not enough.

Weapon procs like Holy, Flaming, Bursts, Blasts and the like favors TWF over THF. Adjusting the bonus they give to two-handed weapons would give two-handed fighters a little help.