Is Mr. Robot right? [Prot]

90 Human Paladin
18025
03/18/2013 03:56 PMPosted by Zoopercat
Now, there is some mathematical way to figure out the actual chances to get an item in x number of tries. I have to do some googling to refresh my math to figure that out.


This one is up my alley. Given a probability p that the boss will drop a particular item when killed (p=1/7 in your example), the chance that you will NOT see that item in N kills is

P(!item) = (1-p)^N

The chance you see that item drop is obviously one minus the probability it doesn't drop, thus

P(item) = 1-(1-p)^N

You need to break out binomial statistics if you want more detailed information on specific situations (like seeing it drop 2 times out of N), but that's probably not necessary for the back-of-the-envelope calculation you're performing.

Including coins is more difficult, because while it may be a 15% chance of getting loot from the coin, you're not always guaranteed to get the item you want. The same formula works for coins though, given appropriate values of p and N. It would be multiplicative with the probability of drops.

Example: Let's say there are 2 items on the boss's "coin table" for your class/spec. Then if you coin that boss N times, the chance you get the item at least once is C = 1-(1-c)^N with c=0.15*0.5=0.075. The chance that you fail to see the item drop from either the boss itself or a coin in N tries is

P(crappy luck) = P(!item) C(!coin) = (1-p)^N * (1-c)^N
Reply Quote
90 Tauren Druid
8160
Because I love data points... I'll share what I found. I looked up my new server, Kul'Tiras on wow progress. Our server is ranked 119 out of about 250. So that appears to be almost smack in the middle of the pack. The FIRST guild to kill the end boss of ToeS did it on Dec 18th (a Tuesday). That leaves the top guild 10 weeks of farming T14 content to get all of their 496 gear. That's obviously not enough trials for people to get every piece of 496 gear they want.

If you have a 10 man raid with 10 bosses dropping 2 pieces of T14 gear each week, that's 20 items. Your group needs 160 items to get 16 T14 pieces per person. Over 20 weeks of farming, 200 items would have dropped. Some % would be wasted, and some % extra would be won on bonus rolls.

Here's a list of the average iLevels for players on our middle-of-the-pack server. The top 400 players are 490 or higher. Anyone else know how many people are actively playing on any given server? I'd guess 2,000 to 10,000? If it's 4,000 then 10% of the people on our server are above 490. Everyone else is below that. http://www.wowprogress.com/gearscore/us/kul-tiras

Just another data point worth sharing and thinking about. Because most of us like geeking out over data ;)
Edited by Zoopercat on 3/18/2013 4:53 PM PDT
Reply Quote
90 Blood Elf Paladin
19235
Wowprogress lists raiding (ranked) population, which means any guild that's got at least one normal mode raid kill. For Kul Tiras it's just over 4k, but that excludes people who only pvp, along with people/alts who only quest, only do 5-mans, only do LFR, and may be even guilds that are just too small to register.

I don't know anything that's replaced warcraftrealms.com census addon that tracked real population, but it is most definitely no longer current.
Reply Quote
90 Human Paladin
13195
I don't know anything that's replaced warcraftrealms.com census addon that tracked real population, but it is most definitely no longer current.


http://wow.realmpop.com/

It uses a different method that (should) give better results. It data mines auction information for every realm (which is updated quite regularly with the wow API), and then looks for every guild member of every seller it finds.

So basically, all it takes is for one person in a guild to post an auction with a toon that's in that guild for every character in that guild to get tracked. Tends to work out better than relying on some addon that only goes by characters that you encounter in game.

They don't seem to currently provide any information on gear levels though that I can see.
Edited by Bravehearth on 3/18/2013 6:17 PM PDT
Reply Quote
90 Blood Elf Paladin
19235
Eh for a survey site, accurate population/demographics is all you really need from that aspect of it.

For Dalaran, wowprogress lists about 4k "raiders" and that site shows just over 10k in level 90 population... with 91k total toons. There's still more 85's than there are 90's.

It does report with almost exact accuracy the number of people in my guild, it's missing one toon and I invited a new alt of someone just yesterday or the day before.
Reply Quote
1 Human Rogue
0
As a casual 10N raider, I wanted to weight in. Due to being in a casual raiding guild of average skill level currently 4/6 in HoF and 1/12 in ToT I'm probably the target demographic for AMR.

I have a prot pally tank 491, and a mage 502. I had to make a tank because our guild group needed an off-tank and lack of one stalled our progression. We raid 6 hours a week. We generally have about 10-20 pulls on boss before it goes down, so the slow progression is more a factor of scheduling than anything else.

When I reforge my mage, I generally do it by hand because that is my main toon, I feel like I know what I'm doing, the stat priorities are fairly clear. I check AMR to see what it thinks. I have quite often seen AMR "revise" it's optimal gearing strat when presented with my reforges on the mage and accept my reforges as the new optimum (i.e. 100% gear score moves up and AMR no longer suggests I reforge anything). So obviously, AMR is not flawless, nor do I expect it to be. I realize that it is running under some kind of a time constraint, and the optimal result it returns is the latest optimal result it found during search, not necessarily the global optimum.

However, when I reforge my pally tank, I pretty much select a preset (control-haste) and if the suggestions look good to me reforge into that. I know a little bit about what are the best stats for the tank, I know enough to try to keep my rotation going and to time my SoTR for boss swings, rotate cooldowns, keep up Sacred Shield. But beyond that, I don't have too good of an idea of what I'm doing with this class. I am not a bad tank, but I'm at best average. Therefore, AMR is great for a person like me! I don't have the time to invest into figuring this out, and AMR does the math. Sure, I'm somewhat leery of how much stam it's telling me to stack because I have a feeling I could be doing more dps if I went with more haste but overall I'm happy with the gearing because I seem to be staying alive during boss fights. When I disagree with AMR (like for the Collosus enchant, which I know many on the forums feel is inferior to the STR one) I don't follow AMR's guide on that one slot.

What I'm trying to say here, is that there's a level of skill at which point you "outskill" AMR in the same way that you can outgear a boss in an instance. I've pretty much reached that point on my mage, and I'm starting to take steps towards independence on my pally.

AMR is like the extra wheels on a bike, supporting you until you are ready to be two-wheeling.

What is the point of getting mad at AMR for perhaps not suggesting an absolutely positively 100% optimal gearing strategy? Obviously you can go faster on a 2-wheel bike, without the trainer wheels (i.e. AMR), but that assumes you know what you're doing in the first place. At that point you can configure AMR to suit your needs anyway. AMR works on averages, and on averages certain stats are better than others. On average Stam will save my life more often than Haste because I probably just mistimed that SoTR...

491-495: 12.66 %


^ This is where my pally is now. (Heals OS is high 480s)

In my experience, the users of AMR tend to be not very knowledgeable but curious. We probably know enough to know when we're screwing up, we know the general trends in gear and keep up with the class forums more or less, read the patch notes and so forth. We are not brainless enough to think AMR will magically make us be able to kill a boss. But we are not always aware of the class mechanics enough (or willing to put in the time) it takes to out-math AMR. Some AMR users are more skilled (or hardworking) raiders than others.

This thread was worth reading if only for the theory-crafting bits and pieces that shed some light on my class.

Kudos to the devs of AMR who have kept a level head in this thread despite some pretty brutal put-downs of the tool they clearly spent a lot of time on. I'm a software developer myself and appreciate the effort it must have taken.

Thanks for making my life as an off-tank easier!
Reply Quote
90 Blood Elf Rogue
14960
04/01/2013 03:22 PMPosted by Azea
What is the point of getting mad at AMR for perhaps not suggesting an absolutely positively 100% optimal gearing strategy?
The basis of the argument at the start of this topic wasn't really about whether or not the default stat weights were the 100% best, it was whether or not they were the appropriate weights for Joe Schmoe who just started tanking and wanted a tool to tell him how to gem/enchant/reforge.

The arguments from the people complaining about AMR was that they feel that the default stat weights in AMR should be tailored towards the "average" player, the player that doesn't know enough about the class to dig through all the settings and customize the stat weights. Those are advanced options and should exist for advanced players.

And on that point, the people against AMR are 100% correct. That's a fundamental concept of product design: make it so the defaults work for the majority of the time and leave the advanced settings to the people smart enough to know they need to use them. The question remained though: are the default weights in AMR appropriate for the novice player or are they just for the best of the best?

The people defending AMR seemed to miss this aspect of the argument and just kept hanging onto the idea that it doesn't matter what the default settings are because they can be customized. Luckily for them, Theck showed up and managed to defend AMR all by himself by providing compelling arguments that the default stat weights may in fact be appropriate for the "average" tank. (ie. he actually address the issue that was being raised)

Although I have to say that personally I'm not entirely satisfied with the idea that stam-stacking is the best choice for the default weights. Not because I doubt or disagree with Theck's reasoning, but because the value of stam depends so much on what content you're doing and how geared you are. Also because I think the concepts of "average" is far too binary. I think you'll find "novice" players (in terms of knowing enough to be able to customize their stat weights appropriately) at nearly all ilevel ranges. So I don't think it's necessarily a good thing to only cater to the "average" novice that is undergeared and needs to gear one way and not the "good" novice who isn't undergeared and needs to gear another way.

I really think this is where some additional settings (preferably up-front, in-your-face, obvious ones) would be appropriate to differentiate between different raid difficulties, combined with the suggested "stam-cap" idea. Based on Theck's post it seems to me that what's going on at a basic level is still that you need X EF when doing Y content, since when you're undergeared stam is valuable/necessary to increase your EF, but when you're overgeared you have enough EF just from what's naturally on your gear.

So if someone was to calculate reasonable EF/stam amounts for different content settings they could be used as a "stam-softcap". That way novice tanks only need to set the content that they're doing and will be optimized around an appropriate amount of stamina for that content. The problems though would be that A) someone would need to calculate these stam-caps for all available contents and for every content in the future (and the stam-caps would likely be controversial due to subjectivity), and this seems like more of a tank-specific need/feature for AMR to include, so I don't know how they would go about adding it to a tool designed to work for everyone and not just tanks (would the content setting just not do anything for non-tanks? would you implement the stam-caps as a series of separate predefined stat-weights and just have dozens to choose from?)
Reply Quote
90 Human Paladin
13195
The basis of the argument at the start of this topic wasn't really about whether or not the default stat weights were the 100% best, it was whether or not they were the appropriate weights for Joe Schmoe who just started tanking and wanted a tool to tell him how to gem/enchant/reforge.The arguments from the people complaining about AMR was that they feel that the default stat weights in AMR should be tailored towards the "average" player, the player that doesn't know enough about the class to dig through all the settings and customize the stat weights. Those are advanced options and should exist for advanced players.


There's a fatal flaw with this argument. It's flawed because a person that is new to tanking, and undergeared, would typically be better off gearing for a lot of EH via stamina, since they will likely not be as good at keeping up their active mitigation, etc.

Theck pointed out quite well how even doing easy stuff, stamina works out pretty good. If tanks were truly worried about how much damage they take, they'd be gearing for flat out avoidance, ignoring hit, expertise, mastery, haste, and stamina.

The thing is, there IS NO SUCH THING as a default best gearing strategy for tanking. That's been nearly universal since, well, forever. There's too many factors involved. Different fights, different player preferences (both the tank and the tank's healers). Mr. Robot has to default to something, and lots of stamina are not a bad thing. It works just as well when someone is undergeared for the content as it does for someone that's pushing heroic 25m.

It all comes down to it being the user's fault for not using a tool properly. Mr. Robot is a tool for helping a person reforge/gem/enchant their armor based on stat weights. It's not trying to teach people all the details of their class. It's not there to teach rotations, etc. You don't blame a calculator when a person tries to divide by zero (no jokes about imaginary numbers, please).
Edited by Bravehearth on 4/1/2013 10:50 PM PDT
Reply Quote
90 Blood Elf Rogue
14960
There's a fatal flaw with this argument. The flaw is that a person that is new to tanking, and undergeared, would typically be better off gearing for a lot of EH via stamina, since they will likely not be as good at keeping up their active mitigation, etc.
Again, you are COMPLETELY missing the point of what you quoted. In fact you say that there's a flaw with what I said when in fact you are actually agreeing with what you quoted.

What you quoted had absolutely NOTHING to do with whether or not stamina is correct for the "average" player or not. What you quoted simply said that default settings (for any and all tools/products) should be tailored to be appropriate for the "average" user, and that additional settings and features exist for the advanced users. NOWHERE in that quote did I pick any side as to which stats should be suggested to anyone.

You saw me point out an aspect of the anti-stam camp's argument (an aspect which has NOTHING to do with stamina, btw), labeled me as one of them and preceded to ignore what I actually said in order to shoot down statements I never even made.

I NEVER said that stam stacking was bad.
I NEVER said anything about total damage taken or total damage reduction.
I NEVER said that there was a single best gearing strategy for tanking.
In fact, the BULK of my post was directed at the problem of how to try to accommodate the stamina needs of everyone regardless of ilevel or what content they were doing.

Read full posts before you start jumping to conclusions about what people are saying.

edit: and again, you even edited your post to throw in another "it's their fault for not using the tool properly" argument, which isn't a good one at all. When discussing default settings it does not matter in the slightest what you can make a tool do when you manually change all of its settings. The only thing that matters when it comes to defaults is how appropriate those default settings are. If I make a cellphone who's default setting blocks all incoming calls and instantly deletes all messages then those are BAD default settings, it doesn't matter that they can be changed by the user it is still a BAD design decision to make those the default settings.

Likewise it doesn't matter what you can do when you alter AMR's settings, it doesn't change the fact that IF the default settings are not appropriate for the average user then they are bad settings. Again, I want to stress the IF in that sentence because I am NOT saying that AMR's default settings are correct or not, I'm simply describing the design concept that should be used when picking default values.
Edited by Pancakê on 4/1/2013 11:17 PM PDT
Reply Quote
90 Dwarf Paladin
12680
Why is this thread not dead yet?

You don't blame a calculator when a person tries to divide by zero (no jokes about imaginary numbers, please).
You want to use a calculator as an example? Fine I'll play ball.

If a calculator is set to use radians as a default I will be pissed and you can be damn sure I will blame the calculator. Your average calculator user wants/needs degrees and probably doesn't even know what makes them different. So what if you can change the settings between radians and degrees. If the default was wrong and I just wanted a quick answer I'm going to take the "wrong" answer and use it not knowing any better. The argument in the thread is/was what should the calculator default to; not which is better because both work in different situations.

I may have radians and degrees backwards. High school math was a long time ago but you get my point.
Edited by Ussil on 4/2/2013 9:26 AM PDT
Reply Quote
90 Blood Elf Paladin
7570
The only thing I rely on Mr Robot to do correctly is to help me reshuffle my reforges and gemming to get as close to 7.5% hit and 15% expertise.

I have to drop the stamina weight equal to dodge/parry or AMR will start reshuffling my gems/enchants in favor of piling on stamina. Even then I will go for stam/mastery gems to meet socket bonuses as I personally believe that all of the bonuses added up surpass a straight up mastery gemming strategy.

It took me quite awhile to realize why AMR was stacking stamina so much. I still don't know if I agree with the theorycrafting. I would be inclined to believe that only some gearing strategies want as much stamina as AMR would recommend "out of the box".

Perhaps if AMR had a short survey before loading the first set of weights that would help tweak them?

Some questions like this:
"Are you bleeding edge progression?"
"Are your healers telling you that you die quickly often?"
"Do you/your healers prefer to take less damage more often, or more damage less often?"
"How is your computer/latency?" (this would help steer 56.6k modem 25 man raiders away from trying to stack haste)
"Are you only interested in reforging/regemming to get closest possible to hit/expertise caps?" (this would be just for me lol)
Edited by Kaskarum on 4/2/2013 10:15 AM PDT
Reply Quote
I actually just used it today, though I made some minor changes to the setup - I bumped Mastery up to 0.9 (With Haste at 1.0) and I dropped the Stamina to 1.3. Also, I decided to match the shoulder socket despite it telling me to forego it (+90 Stamina is still Stamina). Obviously forced the Hit / Expertise caps.

Thus far, between the two this seems to be the most optimal setup as shown by Theck's blogs, at least in my opinion. It still has the quick aspect of having Holy Power on faster demand while giving extra Block % and, of course, an increased physical reduction in SotR. So yeah, I think that AMR, with Theck's weights, has Prot Paladins down nicely. Adjusting Stamina per request of the tank him / herself or the healers, of course, should always be considered. I sort of wondered if dropping my 476 trinket for the 502 Mastery one would be worth it; however, the loss of 28K is a little eery (it puts me just above 500K HP before buffs I think).

But that's also what others mean - we don't necessarily follow a perfect world. It's up to the player, the healers, and ultimately - the raid group what should probably be followed.
Reply Quote
90 Night Elf Druid
12985
Late to the party, but so what? With the exception of Theck, this thread is, as they say in academia, long on assertion and short on demonstration. I wanted to just weigh in on one thing and then leave the same way I came (through Theck's excellent, evidence-based blogging)...

03/14/2013 12:52 PMPosted by Ussil
2. The best way to reduce stress on healers is by lowering incoming damage. Stamina does not lower incoming damage at all. You take more damage if you forgo tanking stats for stam. The term manasponge came about for a reason. It is not a term of endearment.


Absolutely untrue. I have to wonder if you've ever actually healed cutting-edge progression content. As an ex-shammy who once healed heroics in a high-ranked guild, the best way to reduce my stress was to simply give me a fraction of a GCD more to get a heal off. Sometimes tanks accomplished that with mitigation, other times by soaking damage. Most often it was a bit of both. Your black-and-white approach to stamina is utterly ridiculous. You want to set a cap for gemming, fine. 0 is not the cap. I'm glad your guild is beginning HToT progression, but if this is your attitude toward gearing, I have to wonder if your healers don't secretly hate you.

tl;dr But please don't go around saying that you're lowered healer stress by mitigating damage instead of soaking more/bigger hits. You couldn't be more wrong.

Sorry for the pseudo-necro, but I felt this little piece of errata had to be addressed before some hapless tank came along and start repeating it as if it were true. Flame all you want. I won't be reading the responses. This post is purely for other tanks who may one day find this.

PS I think AMR is great. As a tool that 80% of the community can use so they can access content, it's phenomenal. Sure, they may trust it a bit too much, but who cares? If it's the difference between them actually reforging something and just going in with 11% hit, I'm all for it. The bravado here is just silly. If you're raiding heroics, don't use ARM. If you're raiding LFRs or normals, do. In the end, ARM doesn't prevent the average player from downing content, and instead helps them. That's what matters. If the raid got purplez, it was a good night.

EDIT: spelling
Edited by Pandalishis on 5/2/2013 11:10 AM PDT
Reply Quote
90 Draenei Paladin
0
I like the help AMR has given me with my Holy Spec. Thank you AMR.
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)
Submit Cancel

Reported!

[Close]