Quick ret question

[RodneyStanger]

As the topic says, I'll keep it short: http://www.wowhead.com/item=47729 versus http://www.wowhead.com/item=50803? All the hit is used from the bloodshed band.

[Copain]

If you need hit to be at the hit cap, I'd grab that personally.

[Peewee34]

Well as a dps you need to be hit capped, although this could be affected by your class and faction (draenei hit buff/talents that increase your chance to hit).

[Cardolan]

Well as a dps you need to be hit capped, although this could be affected by your class and faction (draenei hit buff/talents that increase your chance to hit).

That's not entirely true. It's possible to have a combination of gear that leaves you under the hit cap provide more DPS than a combination of gear that keeps you at (or above) the hit cap.

To take my Retribution gear as an example, switching out the heroic Taiga Bindings for the Polar Bear Claw Bracers and the heroic Citadel Enforcer's Claymore for Shadow's Edge is going to be a DPS increase, despite no longer being hit capped.

[Copain]

That's not entirely true. It's possible to have a combination of gear that leaves you under the hit cap provide more DPS than a combination of gear that keeps you at (or above) the hit cap.

To take my Retribution gear as an example, switching out the heroic Taiga Bindings for the Polar Bear Claw Bracers and the heroic Citadel Enforcer's Claymore for Shadow's Edge is going to be a DPS increase, despite no longer being hit capped.

You're basing yourself off luck then.

For Example:
You hit for: 1500 on Average
Paladin A hits for: 1000 On average.

You both hit something twice.

You miss one? Paladin A has done 500 more damage then you.

You miss both? He's done 1000 more damage than you.

If you miss, you lose a good chunk of DPS, and you're leaving yourself open to miss more and more meaning more and more of a DPS loss.


~By the way, if you're basing these numbers off something like Rawr, don't. Because Rawr can't (nothing can) factor in random things like hit (While not at cap) or Crit (Unless you have 100% or 0% Crit chance).

[Cardolan]

You're basing yourself off luck then.

For Example:
You hit for: 1500 on Average
Paladin A hits for: 1000 On average.

You both hit something twice.

You miss one? Paladin A has done 500 more damage then you.

You miss both? He's done 1000 more damage than you.

I don't think that an example fight that lasts for two hits is useful or relevant, to be honest. I also don't think that the numbers you used for the damage figures are useful either. You also made a typo, if I don't hit either time then Paladin A will deal 2000 more damage than me.

There are a lot of people that act as if hit rating operates between 0% and 100%, where you'll never hit anything if you don't have any hit rating at all. Of course every sensible and knowledgable WoW player knows that melee hit rating operates between 92% and 100%, where you'll hit 92% of the time (on average) if you don't have any hit rating at all.

Now, let's say I'm 1% (about 33 hit rating) off of the hit cap, and therefore expect to hit 99% of the time, or 99 out of 100 attacks. On average, I'll expect to miss one out of every 100 attacks. Given a sample size of 100 attacks, I may get lucky and not miss at all, or I may get unlucky and miss a couple of hits (let's say 3).

I'm aware that being below the hit cap I'm placing myself under the effects of luck. However, if the difference in damage between my attacks and Paladin A's attacks is high enough, it will outweigh the attacks I miss and result in an increase in damage dealt (and therefore DPS). Of course, in my example, it's not two different people - it's the same person using two different sets of gear.

If you miss, you lose a good chunk of DPS, and you're leaving yourself open to miss more and more meaning more and more of a DPS loss.

This is either badly worded or wrong, I'm not sure which. If we assume that it wasn't badly worded, then you seem to be suggesting that missing has a snowball effect, where every attack that misses increases the chance for subsequent attacks to miss. That's obviously total crap, so we'll assume it was badly worded and move on.

~By the way, if you're basing these numbers off something like Rawr, don't. Because Rawr can't (nothing can) factor in random things like hit (While not at cap) or Crit (Unless you have 100% or 0% Crit chance).

This is a pretty popular opinion. And despite the amount of times I've seen people say that Rawr can't factor in random chance, none of them have gone on to explain why it can't. Perhaps it's because they don't know why it can't and they're just repeating what they've read from other people.

So, for anybody who cares to answer it, a question: If the game engine can factor in random things like hit and crit, why can't Rawr?

[duraz0rz]

Hit and crit are easy to model. I would imagine they assume ideal conditions where you will only miss 0.05% of your attacks if you are 0.05% under hit cap. Of course, ingame it can be more or less because of the RNG nature, but Rawr and other spreadsheets is intended to show your theoretical max DPS.

[Copain]

I don't think that an example fight that lasts for two hits is useful or relevant, to be honest. I also don't think that the numbers you used for the damage figures are useful either. You also made a typo, if I don't hit either time then Paladin A will deal 2000 more damage than me.

My bad there, I was in a hurry xD.

There are a lot of people that act as if hit rating operates between 0% and 100%, where you'll never hit anything if you don't have any hit rating at all. Of course every sensible and knowledgable WoW player knows that melee hit rating operates between 92% and 100%, where you'll hit 92% of the time (on average) if you don't have any hit rating at all.

It's either a hit or a miss is my point. If you need to kill something and you miss, that can cost you.

Now, let's say I'm 1% (about 33 hit rating) off of the hit cap, and therefore expect to hit 99% of the time, or 99 out of 100 attacks. On average, I'll expect to miss one out of every 100 attacks. Given a sample size of 100 attacks, I may get lucky and not miss at all, or I may get unlucky and miss a couple of hits (let's say 3).

For all you know you can miss almost every attack. It's based off luck, the less hit you have, the more you tease it.

I'm aware that being below the hit cap I'm placing myself under the effects of luck. However, if the difference in damage between my attacks and Paladin A's attacks is high enough, it will outweigh the attacks I miss and result in an increase in damage dealt (and therefore DPS). Of course, in my example, it's not two different people - it's the same person using two different sets of gear.

Depends on how much you miss. The numbers could be closer than in my example.

This is either badly worded or wrong, I'm not sure which. If we assume that it wasn't badly worded, then you seem to be suggesting that missing has a snowball effect, where every attack that misses increases the chance for subsequent attacks to miss. That's obviously total crap, so we'll assume it was badly worded and move on.

The "And" part is where the point is changed. I'm saying that missing hurts your DPS, and by being under the hit cap, you risk more.

This is a pretty popular opinion. And despite the amount of times I've seen people say that Rawr can't factor in random chance, none of them have gone on to explain why it can't. Perhaps it's because they don't know why it can't and they're just repeating what they've read from other people.

So, for anybody who cares to answer it, a question: If the game engine can factor in random things like hit and crit, why can't Rawr?

Sorry but how do you not get this?

It's Random. Random is Random and you can't factor Random.

50% Crit chance does not mean you will crit Every two hits (Not critting once doesn't mean your next chance has your previous chance's added (for 100%)).

The in game engine does not go "Oh, he has 50% crit chance so he'll crit this much each fight!" It's basically like rolling a dice each time to see if you crit or not.

Rawr can roll random numbers to check, but you can't factor in random because random numbers change each time you go to get a random chance (A die can land up a 2 when you roll it one time, a 5 the next).

Same goes for hit. Rawr can say you lose the hit and get more DPS, but because hit chance is random if it's not 100%, you could miss and lose DPS.

[Wonderboy]

Sorry but how do you not get this?

It's Random. Random is Random and you can't factor Random.

50% Crit chance does not mean you will crit Every two hits (Not critting once doesn't mean your next chance has your previous chance's added (for 100%)).

The in game engine does not go "Oh, he has 50% crit chance so he'll crit this much each fight!" It's basically like rolling a dice each time to see if you crit or not.

Rawr can roll random numbers to check, but you can't factor in random because random numbers change each time you go to get a random chance (A die can land up a 2 when you roll it one time, a 5 the next).

Same goes for hit. Rawr can say you lose the hit and get more DPS, but because hit chance is random if it's not 100%, you could miss and lose DPS.

What? If you have a 1% chance to miss then assuming infinite swings you would miss 1% of them. This can be factored into any mathmatical model. Just because something is random doesn't mean it can't be factored in. Sure its not going to be 100% accurate, but then again the Rawr model assumes perfect rotation and consistant latency. Its a fine model to use but it will be subject to variation.

Hypothetical Ex. If on a melee hit for 5,000 and crit for 10,000 with 50% crit and 0% chance to miss. My average melee hit would be 7,500, i would never hit 7,500 but it could still be modelled even based on chance. The model is just less accurate the fewer hits you make.

[Copain]

What? If you have a 1% chance to miss then assuming infinite swings you would miss 1% of them. This can be factored into any mathmatical model. Just because something is random doesn't mean it can't be factored in. Sure its not going to be 100% accurate, but then again the Rawr model assumes perfect rotation and consistant latency. Its a fine model to use but it will be subject to variation.

Hypothetical Ex. If on a melee hit for 5,000 and crit for 10,000 with 50% crit and 0% chance to miss. My average melee hit would be 7,500, i would never hit 7,500 but it could still be modelled even based on chance. The model is just less accurate the fewer hits you make.

No. You can't factor in random. You can come up with theories, but nothing can be for sure with random chance.

Flip a coin. You have a 50% chance to get heads or tails. But you can flip it 5 times and land all heads or all tails. That's just how it is.

[Cardolan]

Flip a coin. You have a 50% chance to get heads or tails. But you can flip it 5 times and land all heads or all tails. That's just how it is.

You can flip it five times and get all heads or all tails. You can't flip it five million times and get all heads or all tails, though.

If you have a 50% crit chance, you will expect that half of your total hits will be critical strikes, on average. If the sample size is large enough the results WILL reflect this, because the outcome of individual rolls (the random element) is washed out by the sheer number of rolls being made. Random is only an issue with small sample sizes, which isn't (or at least shouldn't be) how Rawr works.

[underdogba]

You can flip it five times and get all heads or all tails. You can't flip it five million times and get all heads or all tails, though.

Yes you can. It's just extremely unlikely.

[Nillo]

Yes you can. It's just extremely unlikely.

(1/2)^5000000 to be precise

[Eurytos]

Im going to blow your mind right here.

Lets say Person A is in all ilvl 200 gear, and hit capped.
And Person B is in all ilvl277 gear, and has no hit on gear.

Spec and rotation and latency are all exactly the same. The only difference between these two players, is their gear and hit cap. Who will do more dps?

For the sake of convience, I will just tell you that Person B will do more dps. While Person A will never miss, his attacks will just not hit as hard. Person B will have a few misses in his dmg meter, but will hit much harder and therefore do more dmg.

The point being, that sometimes it is ok to take a better piece of gear, even if you happen to drop below hit cap. Now, if it is a choice between items that are more comparable to each other(close or same ilvls), then the piece that hit caps you over the piece with 1 or 2 slightly extra stats will likely be better for your dps.

[Copain]

You can flip it five times and get all heads or all tails. You can't flip it five million times and get all heads or all tails, though.

If you have a 50% crit chance, you will expect that half of your total hits will be critical strikes, on average. If the sample size is large enough the results WILL reflect this, because the outcome of individual rolls (the random element) is washed out by the sheer number of rolls being made. Random is only an issue with small sample sizes, which isn't (or at least shouldn't be) how Rawr works.

It can do that, and as said it's just unlikely. It's all based of a random number generator, and that's part of the point. Having no chance is better than having almost no chance most of the time.

Assuming you have 50% Crit chance, and pretending it's long term and you don't have bad luck or good luck, you can average about 10% either way from the base 50%. Which can have a decent affect on your DPS (40 to 60 percent average).


I'm not saying you'll get a crap ton off with Rawr, but things that are left up to a RNG cannot be fully computed. Because you will not always deal the amount of crits that your crit rate is (50% being 50% of the time in a long battle). It could be close, but you would still be off at times.

Im going to blow your mind right here.

Lets say Person A is in all ilvl 200 gear, and hit capped.
And Person B is in all ilvl277 gear, and has no hit on gear.

Spec and rotation and latency are all exactly the same. The only difference between these two players, is their gear and hit cap. Who will do more dps?

For the sake of convience, I will just tell you that Person B will do more dps. While Person A will never miss, his attacks will just not hit as hard. Person B will have a few misses in his dmg meter, but will hit much harder and therefore do more dmg.

The point being, that sometimes it is ok to take a better piece of gear, even if you happen to drop below hit cap. Now, if it is a choice between items that are more comparable to each other(close or same ilvls), then the piece that hit caps you over the piece with 1 or 2 slightly extra stats will likely be better for your dps.

I would say it depends on the gear though, if they had similar item levels and similar skills and so on, the no miss might make up for it.

But I do agree to a point on that extreme, yes, the gear would be better than the hit.

[Cardolan]

Assuming you have 50% Crit chance, and pretending it's long term and you don't have bad luck or good luck, you can average about 10% either way from the base 50%. Which can have a decent affect on your DPS (40 to 60 percent average).

I'm not saying you'll get a crap ton off with Rawr, but things that are left up to a RNG cannot be fully computed. Because you will not always deal the amount of crits that your crit rate is (50% being 50% of the time in a long battle). It could be close, but you would still be off at times.

The thing about probability is that the percentage of times that something actually happens will converge towards its probability of happening as you continue to run the test. If you flip a coin once, and it lands heads up, then the percentage of times that it landed heads up is 100%. Flip it a few million times and the percentage of times that it landed heads up will be 50%. If you were modelling those flips using a computer system, and you were 10% off after a few million runs, you wouldn't say "Oh, that's just the effects of luck.", you'd say that the system was flawed.

What Rawr gives you is an average, with a large enough sample size that the percentage of times that things occurred is equal to their probability of occurring. If you have a higher average with one set of gear than with a second set of gear, you say that the first set of gear is worth more DPS.

In a real world setting, you may see slightly lower DPS in one specific fight because you missed slightly more times than probability says you should, but on average you'll see higher DPS. It really comes down to whether or not you're willing to settle for slightly lower DPS on average to eliminate some of the random element in individual fights, or if you're willing to take the chance and have a higher average DPS.

I, personally, will go with the second option. And considering the Retribution BiS lists for Alliance aren't hit capped, I don't think I'm the only person who feels that way.

[Broxigar the Red]

Well as a dps you need to be hit capped, although this could be affected by your class and faction (draenei hit buff/talents that increase your chance to hit).

http://www.wowarmory.com/character-s...os&cn=Bluedeep says hi

[Deleted]

You're basing yourself off luck then.

For Example:
You hit for: 1500 on Average
Paladin A hits for: 1000 On average.

You both hit something twice.

You miss one? Paladin A has done 500 more damage then you.

You miss both? He's done 1000 more damage than you.

If you miss, you lose a good chunk of DPS, and you're leaving yourself open to miss more and more meaning more and more of a DPS loss.


~By the way, if you're basing these numbers off something like Rawr, don't. Because Rawr can't (nothing can) factor in random things like hit (While not at cap) or Crit (Unless you have 100% or 0% Crit chance).

this argument is abit flawed in that it could also be used to argue that any crit chance less than 100% isnh't worth it bercause your leaving yourself up to chance.
going by average higher dps is better since most fights alst long enough for you to actually get a high enough number of attaks to make it fairly acurate while the shorter ones can both net you higher or lower numbers.
heck we should all just buy emblem gear and not even kill bosses that drops loot cause were leaving it up to chance.
and rawr models those random elements pretty good by averaging them out

[Valhuur]

http://www.wowarmory.com/character-s...os&cn=Bluedeep says hi

What makes me wonder about a gearsetup like that is wether the much higher raidbuffed crit will be worth losing the massive amount of base strength. I've got only 10 man stuff to choose from and I do prefer to be around the hit cap and expertise cap (with or without glyph, just due to gear that has dropped and hasn't dropped yet I'm currently at 26 from gear) and I've got a good 400 more strength on him (only 150 AP before buffs though due to all his leather with static AP) Raidbuffed he must have a good 10% crit on me. I should sim it someday with 10 man gear equivelents I guess...

[Malthanis]

I have to admit, I'm kind of confused by all this talk about the RNG. To be entirely honest, it's not random at all. An RNG in any computing situation is just a complex algorithm that takes in a seed number and manipulates it to an output between a minimum and maximum value (fun fact: if the seed number remains the same, you'll always get the same 'random' output). We don't know the exact formula that the WoW engine uses for it's RNG, but we can come up with an equivalent thanks to basic mathematic principles. We have the initial variables in the form of your stats, and we have the output in the form of your DPS (hits, crits, glances, misses, etc.). It'd be like solving for x in algebra, but far more complex, and you'd need what my programming professor would call a 'metric f***ton' of data to solve for the formula.

All of my math nerdiness aside, time to be on topic.

Can you do more DPS without being hit-capped? Sure. However, it's not always guaranteed. If you're going with pure numbers in a statistically clean situation (latency, not having to toss utility spells around, etc.), you can equate it to a scientist deciding that they're okay with leaving a variable unaccounted for because it'll only happen 1% of the time.

At the end of the day, being hit-capped means you do consistent DPS, which is something I would much rather prefer seeing as opposed to someone hitting 10k on one fight and 6k on another.