It's garbage because it posts unexpected results that do not fall in line with widely known and accepted knowledge (such as what we know from Landsouls sheet) that has existed for 10+ years, I've mentioned these enough times now I'm not repeating myself. I'm also not digging into every aspect of the sheet to try to win a petty argument that has become a personal issue that you're persuing for your own goals after the thread has already moved on. Obviously you didn't read the thread so we have to go over this for the 140th time I've not got the ability to fix it myself, my skill/knowledge of spreadsheet functionality is limited so I'm not going to persue a waste of my time.
The sheet was not well respected or used at the time for Fury and that was for good reason, you seem to think you have an argument here which is "unless you can prove it's wrong, it's not wrong" without yourself even being able to identify why it posts unexpected results in many areas. Like I've already said I could not care less if you want to use the sheet for gearing/gemming/otherwise, but I won't be joining you. But by all means sit there in your chair and tell yourself that you're right and everyone else was wrong to trust the Landsoul sheet, which gave different results.
You seem a lot less interested in a working accurate sheet than the rest of us here, and more interested in what only seems to be a matter of pride surrounding this other sheet you're using, which is a complete disconnect from what everyone else here is trying to achieve, and also why I'm bored of arguing with you about it.
-
Member
Edited: August 6, 2017
-
Banned
2+2 != 2*2 because i say so
!
-
Member
Mostly the issue I had with the sheet was the statweights. Haste is being valued at more than str 1:1 at rather low stats and so is crit iirc.
Edit: I can't actually recall if I got the haste result when I ****ed up with the data input, but crit is being valued extremely highly.
-
Member
I can't reproduce either of those actually. One thing you need to keep in mind though is that when you're changing stats you have to change the attack power as well. If you only change Strength and not Attack Power all the sheet will calculate is the Strength bonus from Kings, the rest of the attack power (2) must be manually added to the Attack Power part. When I do this Haste and Crit are slightly less valuable than Strength.
A short summary of the bull**** you've just written:
"I suck at maths, I can't be bothered to actually look into the sheet to get a firm basis for my arguments, you suck, please stop talking to me because I can't admit I'm wrong but I can't give any decent arguments anymore."
Just admit your loss mate and walk away with your dignity while you still can.
-
Member
Edit : I will list some of the small annoyances, but you need to identify the holes in the sheet. I'm not using up my own free time (and a considerable amount of it, it would take hours and hours) just to satisfy your desires in an argument I don't care about. There is a good reason I'm speaking more about Landsouls sheet and the SEP Pawn list than this sheet, because this sheet doesn't interest me, I'm not invested in it.
1. Dragonstrike Proc.. I can't find the calculations for it (because there aren't any), it's giving the same value on offhand and main hand, this is wrong. Dragonstrike is a % chance to proc and cannot stack, nor does dual wielding double the uptime and nor would it be worth double the dps value if it did. The procrate on Dragonstrike is affected by almost everything, the uptime can vary massively depending on factors, and this isn't remotely modelled here, it's actually less accurate than the gear ranking sheet here.
2. Expertise - 6.25% is the known cap, sheet has it 5.6%, easy to change but the sheet can't decide if the player is doing 33 dps or healing the boss for 200 HPS when boss dodge is set to 100%. Some wonky things happening with dodges.
3. Armor / Ignore Armor has no cap, it continues to scale exponentially forever in the sheet, base armor set for bosses is 7685 while most bosses have 7700 armor unless I'm mistaken, small difference. I can't find the calculations for armor/dps gains, so I'm moving on.
4. Bloodthirst damage - min/maximum on hits and crits is the same number, no variance.. This is obviously due to the way everything is calculated as purely static at all times, but it's oversimplifying things too much.
5. Heroic Strike gameplay modelling?
6. Madness trinket drops the boss armor in the sheet 120 passively, so it's calculating a 40% uptime and blanketting it as an average 120 ARP... That's an extremely primitive way of calculating the trinket, granted the pawn sheet is using the same method, but that sheet isn't calculating dps or stat weight. Trinket modelling without accounting for any variance/synergy factors in proc mechanics, just averages.. It's actually more primitive than the calculations used for the pawn sheet ranking.
7. Mongoose proc adds a passive 1.29% crit and 0.16% haste, I haven't looked further into it but I don't see how 2% attack speed = 0.16% haste averaged (please correct me if I'm missing something obvious here), but then the calculation is again primitive in just assuming an uptime and then adding averaged stats without modelling the proc in any way. At least MH/OH are showing different values and kings scaling is working.
And actually I'm done, the sheet is clearly unfinished in many aspects. Procs are basically the guy deciding an uptime and then adding the average to the character sheet stats. Everything is just adding/subtracting in a very simple manner, rage modelling is extremely primitive... Like I could sit here all night for hours trying to find holes in the sheet just for this argument but it's pointless, it really is a waste of time.
I don't understand why you're so defensive of the sheet and would rather lash out at me and try to push me into arguing with you than simply talking about the game. With how simplistic the sheet is, it's quite possible I could fix a few of these things myself but why.. This is so far behind Landouls sheet in complexity and in user friendlyness too, and the results I get from it just don't add up, they aren't the same and spending some time looking at the sheet starts to indicate why. But you won't have it unless you have me sit here for days to point out to you what you're quite capable of seeing yourself.
You need to grow a pair, I've never heard such a sad attempt at saving face. Like I said you're sitting there unable to tell why the sheet is wrong yourself and asking others to do it for you, It's on you to do it, not on me. I stated quite clearly from the start I have no interest in someones 10 year old sheet based on a different version of the game posting results that redefine everything we know about the game. Edit - Realise I sound as triggered as you did here, but understand the only reason I'm talking about this sheet at all is because you keep on about it, when I'm talking about Landsouls or other stuff.
It's a good sign that someone has lost an argument when they reply with malice and insults, you just began that trend. But go on keep going, you won't come out of this the better man by being stubborn and throwing around baseless and empty insults. Get over your pride, get over yourself.Edited: August 7, 2017
-
Member
Had another quick look this morning.
Haste : Not only is it overvalued from 0, but it gives continiously growing forever increasing gains the more you add, which is actually the opposite of how haste scales/interacts with the spec. Haste should scale around other stats and gain in value as other stats are increased but lose value while stacked with itself. While it does very slightly reduce flurry uptime in the sheet (by the order of roughly1% per 1000 rating) it's massively overvalued in dps contribution, presumably because of the poor heroic strike modelling. In this sheet haste is almost equal with Expertise from 0, adding 67 of each rating and comparing them, but if you bump the base haste to 300 and then compare adding 67 of either stat haste is completely outclassing Expertise by a large margin and the difference continues to increase, what a mess.
Crit - Slightly overvalued and effectively does not move in value whether you have 47 crit rating or 547crit rating it gives identical dps value right up until you reach the crit soft cap where it starts to dip. It's one thing for a pawn sheet to show completely static values, it's another for a gameplay model sheet which is meant to generate accurate weights to run a constant, this is not accurate crit modelling.
Expertise - The sheet gives expertise a dps value for every 1 rating (not in chunks of 0.25% as per game). I'm also getting a 3.8% damage increase from 4.25% dodge reduction (2% baseline from talents, with the boss dodge changed from 5.6% to 6.25%) and that damage increase percentage is diminishing with higher gear levels, the same is occuring for hit rating below the cap.
To summarise it's oversimplifying calculations across the board to such a degree that it's far outside of the realms of being reasonably reliable, some calculations do not exist at all and number fudging or placeholder number placement is taking place instead. The more time I spend looking at this sheet the more holes I will find and the more answers will unravel to explain why these unexpected things happen, but I've wasted quite enough time on your behalf at this point, so this will do.Edited: August 8, 2017
-
Banned
im pretty sure expertise is supposed to work like that (reduce dodge according to each point of rating's value) unless it's a 3.x change
it just doesnt show up on the tooltip which only displays integer expertise values
-
Member
Expertise works that every 3.9xxx rating converts into 1 expertise, which then converts into 0.25% dodge reduction. Any rating that does not make 1 expertise is completely wasted because expertise does not have decimals. This is well known TBC theorycraft and widely available information, i can't speak for WOTLK as I haven't looked into it for a few years.
Edit :
Elitist Jerks explanation in Warrior compendium sums it up nicely.Expertise can be treated similarly to Hit when under 4.25% Dodge reduction (2% free from talents) most players however don't see it as mandatory as Hit under 9% is seeing as it's harder to obtain. Humans using Swords/Maces and Orcs using Axes gain an additional 1.25% expertise, so the "cap" is 3%. Also note that Expertise Rating rounds down, partial amounts of Expertise do nothing. The "cap" therefore would be 25 Expertise.
Expertise Rating "requirements":
No racial, No talent: 99 Expertise Rating
Human/Orc racial, No talent: 79 Expertise Rating
No racial, 1/2 talent: 83 Expertise Rating
No racial, 2/2 talent: 68 Expertise Rating
Human/Orc racial, 1/2 talent: 64 Expertise Rating
Human/Orc racial, 2/2 talent: 48 Expertise Rating
* Expertise still doesn't have a definitive cap per se, but due to extensive discussion on the topic it's been accepted that shooting for 6.25% dodge reduction is optimal. If the cap is slightly over 6.25%, it's not worth itemizing to get that tiny bit less dodge when Expertise only comes in chunks of 0.25%. The other hypothesis is that Expertise is impossible to cap completely and functions similarly to the spell hit cap, where there will always be a small percent chance that the boss will dodge, regardless of your Expertise. In either case, shooting for 6.25% dodge reduction brings you as close to the cap without overshooting, if there is a cap.Edited: August 8, 2017
-
Member
It bloody well worked because you've finally invested the time to (finally) actually go through the spreadsheet and come up with some decent points.
1) I added that manually and only in page 3 because I wanted to quickly compare Fury with 2x Dragonstrike to Arms. The other pages only allow you to select DS for the MH because at the time of release you could only carry one.
2) Fair point, dodge is too low.
3) That's actually how ARP works. It was only capped to 100% in 3.1 (Patch 3.1.2 (2009-05-19): Capped to 100% (or 1232 armor penetration rating))
4) Show me any TBC damage model that does not average out AP like this because I don't think there are any.
5) It's in there, it calculates your rage per second and you get the point.
6) Every sheet averages trinket procs, the only difference is uptime calculation. 40% is reflected in my Rogue spreadsheet as well (theoretically 40%, calculated uptime of 35.8% in that sheet).
7) Calculated as 2% haste with 1 PPM with 15 sec uptime (so 0.02 / (60 / 15) * uptime), I don't really see what's off here?
I'm not being defensive about this spreadsheet, I've said before that it's not perfect and can be improved, but I am being defensive about you talking shyte about something you don't want to properly investigate. Don't enter a discussion if you're unwilling to do the actual work. Your points here are valid (I had personally overlooked the dodge!) and that's all I was asking for. A good summary of exactly why this should not be used so that we can have an actual discussion.
-
Member
1. On Mongoose the averaged ratio on attack speed isn't consistent with the averaged Crit contribution, I only glanced at it so I haven't looked into why or if it's just an oversight on my part. Also it's not complex to run averages and still have proc value effected by other factors, which then in turn increase the contribution.
2. On trinkets, same as above, if I increase crit then the AP uptime value of Hourglass for example should increase even if only fractionally. Case in point is the ability to have as much as 16% uptime fights like Morogrim even at low gear, due to more chances to proc. In addition to that the uptime contribution also doesn't change with fight length, which it very much should, same for all trinkets. Example if you have a 2min fight vs a 2min 20s fight Bloodlust Brooch contribution should vary a lot, it doesn't in the sheet.
3. Pretty sure variance on Bloodthirst damage is included in Landsouls sheet, I don't have it on hand to check. Either way it's still bad no matter the answer to that question.
4. Rage/Heroic strike modelling are off, like I said If I spend enough time I will be able to identify but I don't have the time to spend, it might answer the crazy haste values, or maybe that has something to do with the haste scaling calculation itself.
5. On armor ignore, there is a cap in TBC otherwise we wouldn't be dropping Executioner enchants in favour of Mongoose in certain gear setups on Sunwell, we would just stack it infinitely to abuse increasing gains, but that isn't the case. Armor Penetration in WOTLK is a different stat to Armor Ignore in TBC, even if it uses a similar concept it works differently, one is a %, the other is static.
6. Dragonstrike - Fair enough.
Like I said though overall the sheet is posting unexpected results and is overly simplified on a lot of calculations or just uses fudged numbers, stat scaling is mostly static (such as Crit only changing in value after the soft cap is reached), Expertise is a little wonky... It's a very incomplete model overall, the biggest issues being Crit and and Haste. Landsouls sheet was no Simcraft but it was a lot more comprehensive than this.Edited: August 8, 2017
-
Member
Thanks for the interesting discussion guys, it's been an interesting read for sure. I'm totally shyte with math so sadly I can't contribute but just some input. Apart from the apparent lack of math in the OP's sheet, even if we make it work and re-calc it all, what'd keep me away is the inconvenience of using it. Having to add all the numbers. So... what options do we got? There's some Landsoul sheets for WotLK, a Landsoul sheet for 2.4 that no one seem to have. There's Rawr which I don't trust at all and then there's SimCraft, does it exist for 2.4? Sadly I have no time to look into it right now but I did find this.
http://simc.heroicstrike.org/BC_Downloads/
Could be worth a check. Iirc SimC loads from an armory profile though so it'd probably be kind of inconvenient as well.
-
Member
Yeah it's mighty inconvenient...
That simcraft is quite bad, I've checked it out before in hopes of finding a great working Warrior model but it doesn't support Warriors yet (same for Warlocks and Paladins). I think it wouldn't work anyway because it takes your talent build directly from the WoW website, not sure if it would work with iradei for example. Haven't tested it though.
I think the best option would be to take that early Landsoul spreadsheet version and go through the entire spreadsheet, update and change the maths where needed. I personally don't have the time to do it though... Maybe another option would be to take the EJ Rogue spreadsheet, it's actually very good in calculating proc uptime etc, and create a Warrior spreadsheet from that. The basic maths for melee hits, WF, buffs, trinket/weapon procs, ... everything is in there already. It would still no doubt be a lot of work to do.
-
Member
Suppose I'll stick to common sense and that google doc Pawn list SegaRally linked earlier. I've gotten a quite nice list so far. After all things haven't gotten too complex yet (opposed to later expansions).
-
Member
The problem with the pawn list is that it's showing the best pieces under a microscope from an individual item perspective, but doesn't necessarily show you how to make the best set, nor does it answer gearing questions specific to your character. Example would be things like set bonuses, hitting the ignore armor cap for specific bosses, staying above the hit cap or combining the right pieces to reduce expertise waste (such as in T5 with shoulders/belt/gloves). More specifically in making the best dps set under those conditions.
Where a gameplay sheet really helps is in experiments like figuring out Talon of the Phoenix (MH) + Dragonstrike (OH) vs Dragonstrike (MH) + Talon of Azshara (OH). Of course Talon of the Phoenix is a better weapon than Azshara, but it forces you to put Dragonstrike in the offhand which will in turn at least marginally lower the uptime (less successful hits due to no heroic strike, no Bloodthirst), so you're weighing off a better weapon vs better Dragonstrike uptime. If you're a Human it's an easy answer, but different for Horde.
On Simcraft, the program as it is now is incredible and the Warrior model is very well supported (though even simcraft these days can be unreliable, and specifically was really off with Ravager talent setups in this latest raid tier). but back in those days it wasn't really supported. The combat sheet I guess is the one I have also? "Combat.1.4.4"? I'm too slow working with sheets and I don't have the talent to convert that, it would take months.
-
Member
Exactly, that's why common sense is needed too. ^^ Right now I've figured out which 4 T5 parts to use and with which expertise gear and so on. It is pretty straight forward to be fair. It's those weapon comparison things that is a bit harder (when procs are involved and what not).
One thing that hit me is Dragonstrike, Dragonstrike, Rod of the Sun King, Talon of Aszhara. Which 2 to use? Would love to just put it into SimC.
Quote
