Fair enough question based on what I asked. Let's say I have a 1600 rated 2v2 arena team and I would like to lose as little PVP functionality as possible from what I have in my current build, (I know pvp mut/sub is ideal but I dont want to shell out to respec for TK every week) but still pick up the imp bs 3/3 talent. Would this change your recommendation based on my spec on which talents I should drop? Build again to save you time: WorldofWarcraft.com -> Info -> Classes -> Rogue -> Talent Calculator
I went ahead and specc'd mutilate for a couple of weeks while we were taking a short break from raiding. I took a pve mutilate specc 41/20/0 with 3 points imp poisons and 2 points vile. I put the 3 points in imp backstab because I was hoping the patch would come out while I was playing and I wouldn't have to respecc. I didn't see any problem in my arenas with this specc. That being said, I'm used to doing arenas as combat daggers so the mutilate specc was already a lot more fun and better in 3s and 5s. I don't think imp KS is really that necessary in arenas if you have a good team and good gear and it is pretty much useless in high-end raiding. If you want a pve/pvp mutilate specc, better to take the 3pts out of imp KS for the new puncturing wounds and in the combat tree go 3pts imp gouge and 2 pts SS instead and then your gouge gives you a little longer to regain energy and move behind the target for another mutilate quickly or you can gouge one target and move to the other to dps it down quickly.
This advice is based on my own experience and from what I've been reading in these forums. I'm not a great mathmetician so I can't show you any great formulas for why I think this would work better. But my experience as a mainly pve rogue has been that I can either arena with my pve specc and do ok, or I can pay to respecc on weekends to maximize my pvp abilities and then respecc again for raids. Currently, Blizzard doesn't really offer rogues a truly viable specc that we can raid with and consider also to be really good for pvp. You have to either gimp your pve dps or gimp your pvp abilities.
Fair enough question based on what I asked. Let's say I have a 1600 rated 2v2 arena team and I would like to lose as little PVP functionality as possible from what I have in my current build, (I know pvp mut/sub is ideal but I dont want to shell out to respec for TK every week) but still pick up the imp bs 3/3 talent. Would this change your recommendation based on my spec on which talents I should drop? Build again to save you time: WorldofWarcraft.com -> Info -> Classes -> Rogue -> Talent Calculator
Vile Poisons then. It's usually a minor boost to dps.
Ah, sorry for the confusion; I was looking for the best MH weapon to increase my DPS without having pay for a respec and / or change my ability rotation ( CS, then Hemo until 4 - 5 point Evis, repeat ).
That is to say, I know how to change my spec / ability rot in order to increase my personal DPS, but assuming I'm too cheap / lazy to go that route, which affects my DPS more and why: the better stats of the Ripper or the higher Hemo damage of the Brand? Just trying to get a better general understanding of how weapon stats / speeds affect DPS in this particular scenario.
Because armor pen gets better the more you get, the fact that the rogue I posted already has 679, executioner is by far and away better than mongoose. 100 ap, 2.5% crit, 2% attack speed < ~7.5% dps increase.
A Mongoose proc with Kings gives you 132 AP, 3.3% crit, and 2% haste (which stacks multiplicatively).
Your statement is simply unfounded. Spreadsheets show that even under extreme circumstances (i.e. all the best gear available in game), Executioner is only marginally superior to Mongoose, and usually only against low armor bosses. There is certainly no situation where it is "far and away better".
A Mongoose proc with Kings gives you 132 AP, 3.3% crit, and 2% haste (which stacks multiplicatively).
Your statement is simply unfounded. Spreadsheets show that even under extreme circumstances (i.e. all the best gear available in game), Executioner is only marginally superior to Mongoose, and usually only against low armor bosses. There is certainly no situation where it is "far and away better".
That is to say, I know how to change my spec / ability rot in order to increase my personal DPS, but assuming I'm too cheap / lazy to go that route, which affects my DPS more and why: the better stats of the Ripper or the higher Hemo damage of the Brand? Just trying to get a better general understanding of how weapon stats / speeds affect DPS in this particular scenario.
The extra stats on the Ripper are negligible compared to the effect that weapon type and damage range would have on your Hemo damage.
Basically if you equip a dagger in your mainhand, your special attacks are normalized at 1.7 speed instead of 2.4 speed. What this essentially means is that your Hemo gains roughly 2/3 of the bonus damage from AP with a dagger when compared to what it would gain with a fist weapon, sword, or mace. When you combine this effect with the already lower average damage of the Ripper, your Hemos would hit for somewhere in the ballpark of 70% of what it would hit for with the fist weapon.
Dropping 30% of your Hemo damage would require a huge chunk of DPS stats to recover.
The extra stats on the Ripper are negligible compared to the effect that weapon type and damage range would have on your Hemo damage.
Basically if you equip a dagger in your mainhand, your special attacks are normalized at 1.7 speed instead of 2.4 speed. What this essentially means is that your Hemo gains roughly 2/3 of the bonus damage from AP with a dagger when compared to what it would gain with a fist weapon, sword, or mace. When you combine this effect with the already lower average damage of the Ripper, your Hemos would hit for somewhere in the ballpark of 70% of what it would hit for with the fist weapon.
Dropping 30% of your Hemo damage would require a huge chunk of DPS stats to recover.
Edit: to the rogue asking about t4/t5, make sure you pick up [Bloodsea Brigand's Vest]to compensate for some of the hit loss.
No. You pick up [Bloodsea Brigand's Vest] because it is good, not because it has hit. There is no magical hit number. It is simply another stat that increases your DPS. If the Brigand's Vest had 27 Agi instead of 27 hit, it would still be a strong upgrade.
So in the theorycraft front, I did some testing on the Level 72 Ogres above Shattrath. 2-4 damage daggers, fixed AP (except for an enrage from the Ogres - data during those periods excluded)
75 Glances in 463 attacks or 16.1987%.
Still a rather small sample size for pinning that down. Useful only for very basic range-finding.
Now the really interesting part...
Off-hand hits...
Either 149 or 150 average of 149.6822.
Off-hand glances range from 120-134 with an average of 127.6429
Average glance reduction of 85.2759%
Minimum glance of 120 = 80.0000% of 150 or 80.5369% of 149
Maximum glance of 134 = 89.3333% of 150 or 89.9329% of 149
Main-hand hits...
Either 199 or 200 or 201 with average of 199.6855.
Off-hand glances range from 160-179 with an average of 169.7179
Average glance reduction of 84.9926%
Minimum glance of 160 = 79.6020% of 201 or 80.4020% of 199
Maximum glance of 179 = 89.9497% of 150 or 89.0547% of 149
It really looks like the glances are 80-90% for an average of 85% vs. 72's. The 160 MH was probably a low roll.
To add to DMM's data, but on a slightly different route.. I "finished" calculating my way through several hundred WWS reports.
I took data from Bloodboil (only non-Fel-Rage to help curb some luck streaks), Najentus, Mother Sharahz, Teron, Anetheron and Archimonde. I used data from all 4 rogues that have raided for my guild in the last ~7 months. While I didn't use a precise formula, anyone who "did bad dps" was excluded. None of the rogues in my guild sucked, so if their dps wasn't in the top 7 or 8 overall, its likely due to death or ghosting or excessive air burst or whatever - small sample data sizes. Note, I generally only took the 5-10 SLOWEST kills my guild has recorded. We're tanking 6 minute Terons, 9 minute Bloodboils, 7 minute Najentus', etc. This was to increase the sample size therefore reducing the margin of error.
I recorded Average Normal hit, Average Glance, # of Hits, # of Crits, # of Glances, # of Misses (the WWS one which includes parry / dodge). I then calculated, per fight, the average % dmg which glances did. This produced roughly ~125 "glancing reduction percent" numbers (one per parse per rogue). I averaged all 125 together. Next I calculated the chance to glance via # of Glance / Total # of attacks. This was done by summing hits, crits, etc up first. So, in the end:
72k attacks, 24.08% chance to glance (24% I would assume) with .. roughly 24.5 to 25% dmg reduction. If someone whos name I recognize wants to look at the excel file, I can provide it via PM request. I'll probably toss you my guild WWS link as well - I listed report date, boss and rogue name on each line.
I'll attempt to help out this weekend similar stuff to what DMM is working on - a more systematic approach to finding exact glance reduction %. Obviously getting an entire range for a given AP will provide a much more exact answer as opposed to an average which includes various AP's, mixing of MH and OH data, etc.
On a side note.. what was your AP, DMM? Could figure out what your theoretical dmg range was on those attacks (after getting armor from kicks :S) was to cut back on some of the rounding issues and "oddities" like the 79.6% reduction which likely was just a low roll on something that was normally rounded up a bit. Were you seeming numbers over the whole range of glancing (from 80% to 90%) or just 80 and 90%? From your wording I am assuming the whole range.. wow thats a lot of random rolls being calculated server-side.
With null hypothesis that glancing rate = 25%, we would be 5.7 standard deviations from the mean (aka, no).
With null hypothesis that glancing rate = 24%, we would be within .53 standard deviations from the mean (aka, yes).
Thus, we can say with high confidence that the glancing rate is 24%.
As for the average reduction... I'll have to think a bit to see if there's some statistics we can grind on that. But given DMM's preliminary data, a 24% chance to have a 20-30% reduction seems most likely at the moment, as it's also consistent with your data.
I recorded Average Normal hit, Average Glance, # of Hits, # of Crits, # of Glances, # of Misses (the WWS one which includes parry / dodge). I then calculated, per fight, the average % dmg which glances did. This produced roughly ~125 "glancing reduction percent" numbers (one per parse per rogue). I averaged all 125 together. Next I calculated the chance to glance via # of Glance / Total # of attacks. This was done by summing hits, crits, etc up first. So, in the end:
72k attacks, 24.08% chance to glance (24% I would assume) with .. roughly 24.5 to 25% dmg reduction. If someone whos name I recognize wants to look at the excel file, I can provide it via PM request. I'll probably toss you my guild WWS link as well - I listed report date, boss and rogue name on each line.
One-proportion z-test for your data:
According to Latito's sample, there is sufficient evidence to reject the null hypothesis that the glancing blow rate is 25%, as is currently thought, at an value of 0.05 (there is a 5% chance that the extreme value observed is due to chance). It would be nice to check this from others and to make sure that the logs for these are correctly parsed such that there are no glances on trash mobs in there. It's also past 3am here and I'm tired so please check my math.
One (stupid?) question about hit cap : I managed to find the +hit i need against raid bosses, other players, but what about 10man raid trashes and heroics dungeons ? I'm nearing the cap actually I suppose I have to reduce hit for ap/crit against theses mobs in order to optimize my dps
No. You pick up [Bloodsea Brigand's Vest] because it is good, not because it has hit. There is no magical hit number. It is simply another stat that increases your DPS. If the Brigand's Vest had 27 Agi instead of 27 hit, it would still be a strong upgrade.
What should be the minimal hit a combat rogue must have then? I'm currently 2*Sword (after Talon of Azshara finally dropped) and I finally switched from 2*T4 to almost full 2.3 badge/ZA outfit (I love the tribal look of it :) ). I found it better according to spreadshets, but I don't feel too comfortable with really sorry 225 hit rating. I'm aware that linear spreadsheets calculations might not be accurate at extremely low or high stat values, so my question is: When hit rating is absolutely too low even if spreadsheets say otherwise?
[Yes, I know, [Shifting Nightseye] in belt is a mistake, it's a souvenir after meta requirements and I plan to get 2.4 badge belt anyway :> ]
What should be the minimal hit a combat rogue must have then?
There is no such thing. You gear for the most DPS, and the optimum -- and minimum -- hit for that build and gearset is... whatever you get. Don't worry about your hit falling too low; if you're still putting up the numbers, it doesn't matter how they're getting there, be it lots of little hits from gearing for hit or fewer, bigger ones because you have more AP and crit.
What should be the minimal hit a combat rogue must have then? I'm currently 2*Sword (after Talon of Azshara finally dropped) and I finally switched from 2*T4 to almost full 2.3 badge/ZA outfit (I love the tribal look of it ). I found it better according to spreadshets, but I don't feel too comfortable with really sorry 225 hit rating. I'm aware that linear spreadsheets calculations might not be accurate at extremely low or high stat values, so my question is: When hit rating is absolutely too low even if spreadsheets say otherwise?
...
Originally Posted by AtADeadRun
There is no such thing. You gear for the most DPS, and the optimum -- and minimum -- hit for that build and gearset is... whatever you get. Don't worry about your hit falling too low; if you're still putting up the numbers, it doesn't matter how they're getting there, be it lots of little hits from gearing for hit or fewer, bigger ones because you have more AP and crit.
Well, almost. In general, the "minimum" hit that a rogue should have is an amount of hit rating that makes your special attacks unmissable. This is +9% (142 hit rating) for boss mobs, or +4% (63 hit rating) if you have 5/5 precision. Below this hit cap for specials, hit becomes much more valuable. You should aim to have at least enough +hit to be hit capped on specials, as it makes cycles much smoother and more reliable to maintain, as well as preventing your finishers from missing.
That said, it is true that the value of hit rating is quantifiable even below the hit cap on specials. If you had, for example, 130 hit rating and no precision and were offered the choice between gaining 1 hit rating or 100 agility, then clearly the agility would be superior. However, due to itemization, when below the hit cap on specials you are most likely to find that gear with +hit will increase your DPS the most, due solely to the increased value of +hit at such low levels of hit.
In general, the "minimum" hit that a rogue should have is an amount of hit rating that makes your special attacks unmissable.
Granted. But the gent in question was worried about being too low at 225 with precision, rather than 20 without it. Perhaps I took that as being too obvious to really need saying.
I'm aware that linear spreadsheets calculations might not be accurate at extremely low or high stat values, so my question is: When hit rating is absolutely too low even if spreadsheets say otherwise?
I'm not sure why you believe the spreadsheets would become inaccurate at high or low stat values but that doesn't make sense to me at all. Neither of them makes the assumption that stats behave linearly. Their models should be just as accurate at 0 hit rating as at 300 hit rating, barring any possible bugs.
Well, almost. In general, the "minimum" hit that a rogue should have is an amount of hit rating that makes your special attacks unmissable. This is +9% (142 hit rating) for boss mobs, or +4% (63 hit rating) if you have 5/5 precision. Below this hit cap for specials, hit becomes much more valuable. You should aim to have at least enough +hit to be hit capped on specials, as it makes cycles much smoother and more reliable to maintain, as well as preventing your finishers from missing.
That said, it is true that the value of hit rating is quantifiable even below the hit cap on specials. If you had, for example, 130 hit rating and no precision and were offered the choice between gaining 1 hit rating or 100 agility, then clearly the agility would be superior. However, due to itemization, when below the hit cap on specials you are most likely to find that gear with +hit will increase your DPS the most, due solely to the increased value of +hit at such low levels of hit.
I think this point has been discussed before, but: even the yellow hit cap isn't a hard-and-fast minimum from a DPS perspective. Even below this point, you want to pick the gear that gives you the highest damage contribution - whatever that happens to be. Now, hit does become somewhat more valuable, so hit items will become correspondingly better... but if it so happens that the very highest DPS item you can get for every slot totals to 0 hit rating... well, then, you have 0 hit rating.
It also might be noted that hit below the yellow hit cap doesn't increase your damage as much as you might think; hit below the yellow hit cap has basically the same benefit as Expertise, which is only about 10% more valuable than hit. Now, if your hit is that low, your white damage is probably a slightly lower fraction of your damage, meaning that the yellow DPS portion will be *somewhat* more important - but we're still talking an increase of at most 20% to the DPS contribution of hit.
That said: there is a *bit* of an argument that can be made for stacking hit to get to the yellow hit cap, even if it costs you DPS, namely that making your specials unmissable is reasonably important on fights where you need to kick, which there are a few of at the T4 level (Aran, Mag). So from that perspective, getting up to the yellow hit cap might make sense even at the loss of DPS - but it isn't the case that getting to the yellow hit cap is necessarily what's going to maximize DPS.
Of course, this is all extremely theoretical, as it's almost impossible to be wearing halfway decent gear and have hit that low. But the point remains: even at hit as low as that, you still want to be picking what's best, independent of how much hit it may or may not have.
Thanks for your comments, they made me less worrying about my current gear selection.
So I'd like to make last thing clear: Even having about 250 hit rating, and having to choose between one item with 23 AP and the other with 10 Hit (both 23 EP), it's better to get the first one (AP), as it may serve better on trashes and in heroics and give me more freedom with gem and consumable selections before hit cap?
You might check the spreadsheet(s) to make sure that the equivalence is true for your exact gear/buff setup; but as a general rule, I would say yes. Because hit (along with haste and expertise) are what I call "white damage stats" - stats that boost only your white and not your yellow damage - they don't tend to scale as well for interrupted fights as they do for the theoretical sustained fight modeled in the spreadsheets; thus, for interrupted fights, their value can be up to 5% lower, which doesn't really change the overall value of the stats that much but does mean that in the event of a tie, I believe it's better to go with yellow damage stats (crit, AP, arPen) than white damage stats assuming that the DPS output is otherwise identical/comparable.
I had heard that 2.4 was making bleed immune mobs no longer immune. Granted I don't know if this is true or not, but if it were true, it would make the 2pt5 set bonus effectively useless, correct? Have any of you heard of this, and if it's true would you happen to know whether or not there are plans to revamp our 2pt5 set bonus into something viable?
Frankly, the T5 2/5 has been more or less useless as it is; there aren't that many bleed-immune bosses in 25-man content, and even on the ones that do exist the 2/5 bonus is hardly earthshaking. In practice, it was almost useless before, so I don't really see them changing it now, nor have I heard anything about such a plan.
Specifically, these are the changes to mob immunities according to the current PTR patch notes:
Non-corporeal Undead and Mechanical creatures are now susceptible to bleed effects.
Elemental creatures are no longer explicitly immune to poison and disease effects. Elementals with nature school immunities will still be immune to nature-based poisons, however.
If I'm reading that correctly, Hydross would thus still be rupture-immune? The only 25-man bosses that appear to be explicitly covered by that are Void Reaver and Doomwalker, though it might effect Supremus as well. Either way, it doesn't seem to me to be a significant enough change in the value of 2/5 T5 that I would expect a change to it.
03/19/08, 3:22 PM
). I found it better according to spreadshets, but I don't feel too comfortable with really sorry 225 hit rating. I'm aware that linear spreadsheets calculations might not be accurate at extremely low or high stat values, so my question is: When hit rating is absolutely too low even if spreadsheets say otherwise?
Similar Threads
