drumbum

I've temporarily uploaded the latest version (0.10.2) at this location: http://www.filedropper.com/roguegear0102

It's not a permanent link and I'll take it down at some point. This is just a temporary convenience.

You are correct that expertise is rounded down in the game, but as the sheet's primary purpose is to compare different pieces of gear, it makes sense to instead treat expertise as continuous. It's a design decision and not a mistake.

 

Falian

As far as I understood 0 Slack should give you the cycle with highest DPS and adding some slack should always output a cycle with lower DPS than the "best".
Now when I set slack to 0 the sheet recommends a 5s5r cycle with 2069,8 DPS (DamageCalcs sheet shows the Xs5r cycle as 2.9s5r with 2068.99 DPS), changing slack to 1 the sheet suggests a 3.5s5r cycle for 2071.84 DPS which seems quite strange.
I do not only get a Cycle with more DPS it also suggests a tighter cycle which should be a bit harder to maintain although i added more slack.
Is this a bug or did I miss something concerning the new slack mechanic in the sheet?

 

• Aldriana

That... sounds like a bug. If you could upload your copy of the sheet somewhere (so I can see what gear/buffs you're using that is causing this problem) I'll take a look at it.

 

kyd

My question is, what is the optimum sockets for pvp damage for a showstep rogue (10Agility or 20AP)

I did a few experiments on the spreadsheet including removing all raid buffs and changing the talents around (to include the subtlety talents such as deadliness and sinister calling) to see what effect that had on ideal stats/gems. The spreadsheet showed that 10agility was superior to 20AP, which I was kind of surprised at as most rogues socket for AP for pvp.

It then occurred to me that resilience reduces crit chance and damage done by crits. Im not sure if im correct but it seems to be that crit chance reduced has the same effect on agility as it does AP (as long as your crit change is larger than the crit reduction from resilience), therefore the reduction would not chance the preference of socketing for agility. However crit damage reduced does have a big effect on agility, as one of the damage increasing elements of agility is increase crit chance.

Basically I was wondering if it was possible to model in the gear sheet the effects of different levels of resilience on the target to dps done.

Im sorry if this has already been asked, I did look for info on this topic but found little I trust as much as the spreadsheet.

 

• Aldriana

This is a PvE spreadsheet. PvP is such a different ballgame that even if I did put in some manner of "target resilience" calculation, this sheet still would be useless. This sheet is assuming several minutes of uninterrupted cycle spamming, which is just not a good model of PvP. So no, I have no plans to do such a thing - that's not what this sheet is for.

 

drumbum

As Aldriana said, this is a PVP question. However, I think I have exactly what you're looking for, so I'll address your question over on the PVP rogue thread ([Rogue] PVP Stat Standards/advice). Check over there soon.

 

Cyn

One way that could help make it a little more suitable for pvp modelling, and aswell as situational pve, is disrupted combat modelling, (x in, y out), but im assuming you've allready considered this Aldriana.

 

• Aldriana

There are parts of that sort of modeling that are easy to figure out. There are parts that are not. I've been bouncing around some ideas for it, and I won't say it'll never happen... but don't hold your breath. It's a very hard problem to solve in the general sense.

 

Falian

Here's a link to the spreadsheet with my gear and buffs
http://www.filedropper.com/roguegear0102_1

Another rather minor Issue I have:
When switching Items the recommending list for that slot seems somewhat bugged as it shows the Item I equipped in the list as a 0 Point upgrade (which should have been removed in some earlier Version) and for slots like trinkets it also shows some other strange behavior which you will hopefully be able to see with my copy of the sheet. If it has already been posted: /blame having to get up early in the morning.
I'm using Open Office btw dunno if this has something to do with the bug but it might be nice to know.

 

• Aldriana

So, I can tell you what's happening... what I haven't had time to figure out is whether the behavior is correct or not.

So, the idea of this sheet was based on some theorycraft I did while on vacation last summer, where I showed on a scrap of paper that I have *long* since lost that if you ignore proc uptimes and consider a cycle of the form Xs5r, the optimal DPS cycle will either be the minimal cycle that maintains 100% SnD uptime, or 5s5r; at no point will it be anything in between. Now, I'm fairly confident in that assertion (or at least, as confident as I can be in any piece of math I did almost a year ago), but you'll note the rather import condition thrown in there: ignoring proc uptimes. What appears to be happening is that the effects of cycle compression on Ashtongue Talisman influence it's uptime enough that that rule does not appear to hold anymore - the 1-second slack 3.5s5r cycle actually does do more damage than the 2.9s5r cycle.

So, here's the crux of the matter: it's possible that the only problem is with the sheet's abilities to figure out what the best cycle truly is. And if that's really the problem, I'm not that worried about it. It's only going to come up when when value of Xs5r and 5s5r are very very close in value anyway, and you're never going to be recommended a cycle that's inferior by more than a tenth of a percent or so. So if that's all it is, I'm not too worried about fixing it (particularly since it's such a pain to fix).

Now, that said: the other possibility is that my Ashtongue Talisman uptime estimates are a bit off, and/or are breaking from improper interaction with the slack parameter. The difference between the 3s5r uptime and the 3.5s5r uptime is rather larger than I might have guessed. So I'm not prepared to say that there isn't a bug here - and if there is, then AToL is being misvalued.

So, basically: there may or may not be a bug there, and I haven't totally had time to figure out which it is.

Regarding recommendations: I'm not seeing the problem you mention - my suspicion is that it's a problem with your spreadsheet program updating oddly, but I can't be sure. Which program are you using?

Re: trinket oddities. Trinkets do tend to do a lot of strange things, so I'm not entirely sure what you're referring to. Can you be more specific?

 

Elhana

Since we are on this question - do you think it might be worth trying to focus keeping AToL buff up with a rotation like 4s4r or shorter igoring the fact that you might loose some energy from time to time?
I was actually going to ask it a while ago, but past experience was telling me energy regen is probably more important than crit, however there is always has to be a borderline.
I've actually tried it few times, but since it becomes random on regen my sample was not enough to proove me anything and math takes too many variables to be evaluated without a spreadsheet, so I just slack with 5s5r when I use it nowadays

Edit: Thanks for the answer, I guess I missed it at some point due to reading recent pages mostly
Roguecraft 101 link: Roguecraft 101 - post #2769 till the end of the page

 

• Aldriana

This was discussed fairly recently in the Roguecraft 101 thread - I'd suggest looking there for the lengthier discussion. The short version is that shortening your cycle to, for instance, 4s4r, doesn't tend to increase your uptime very much - as while it's true that a shorter cycle makes for a higher uptime fraction when it procs, it's equally true that the trinket no longer procs every time. When you work everything out, it tends not to increase the uptime by more than a trivial amount, and the advantage is easily dwarfed by the other inefficiencies of the cycle.

 

Falian

When I change my chest to Bladed Chaos Tunic the recommendation list doesn's change the listed Items, the only difference is the Value (Bladed as 0 point upgrade where it should show Carapace of Sun and Shadow as an upgrade and remove Bladed Chaos Tunic from the list)
Forgot to mention that the recommendations are fine when I close and reopen the sheet (again blaming having to wake up that early )
So it's probably a problem caused by Open Office 2.4.

 

Grimrage

I've spent over and hour now trying to search through 112 pages of this thread for information on this. I know it has to be there but the search function is very limited. My sincerest apologies for repeating this but I just can't find it.

[Talbuk Hide Spaulders] and the twin [Blackened Leather Spaulders] aren't in the sheet. I would like to add them. I chose to unhide the Armor sheet and added a new row, copying from a previous row. The problem is that column H, the ranking column, comes back with an #N/A result. Digging in a bit I find it's checking the ShoulderTot for a range that doesn't include the new row I created. So clearly I screwed up. Can anyone assist in adding in an item? Thanks and sorry again for repeating something that's surely been asked.

 

• Latito

http://elitistjerks.com/search.php?searchid=1253660
2nd post in that search: Rogue Gear Spreadsheet
3rd post in that search: Rogue Gear Spreadsheet

Hope that helps.

Disbanded Recruiting!

 

Grimrage

Thank you very much! Seems "unhide" was the search term I failed to include that would have made it easy.

 

roq

Ald, you accidentally classified the Blackend Naaru Sliver as a Gruul's Lair Trinket, and not a Sunwell trinket. Outside of that keep up the good work.

 

kwinto

If I can suggest you something, Aldriana:
In each one of your spreadsheet versions I set Buffs_Menu!A1 cell to "=Talents_Equipment!B4" (Rough DPS = ...)
It helps me in finding the best consumables and/or party buffs (ie answering to shammy question "Agi or WF?").

Maybe you'll find this tip useful too...

 

• Freddie

Strange bug with current sheet: using default gear and just changing neck to Vindicator's shows 1983.89dps, with Hard Khorium Choker being 1977.61 :o

 

• Aldriana

This would be the same bug that was discussed in posts 2772 and 2773 of this thread. Simply rename the first instance of Vindicator's Pendant to Guardian's Pendant and things will work.

 

• Freddie

Dur :| Have you made much progress on your Python 'version' of the sheet? I meant to PM you about it at some point and completely forgot.

 

• Aldriana

Not recently. I was working on it for a while, but then with the 2.4 items I started working on this again, and now I've been busy with Sunwell. In a month or two once everything is on farm and I get back to 2-3 night a week raiding (instead of the current 5-6) I'll get back into it.

In terms of actual progress: I'm pretty close to a working damage calculator for Xs5r type cycles. At that point, with a little bit of UI to make gear entry easier and a little logic to compute the cycles for each stat, and it'd be basically a working gear estimator for combat swords. We're probably talking a week or two of work once I get back working on it.

Note that at that point it'll be about as functional as version 0.1 of this sheet (that is to say, not particularly)... but it'll be a start.

Honestly, I think the expected timeline for the python version being actually useful (barring a sudden influx of assistance) is "in the expansion". It's still something I'd like to do, but there's just enough work to be done that I wouldn't expect it in the immediate future.

 

kwinto

I don't undestand, why changing faction between scryers and aldor (but staying with one inscription) changes "rough DPS"?

 

Carnivori

kwinto do you have the SSO exalted neck equipped? The Aldor version of the neck is better than the Scryers one.

 

kwinto

Thank you, thank you, thank you! It was driving me mad... Now it's obvoius

Aldor inscription is better, Aldor neck is better... Geez, I've choosen Scryers because I though it will be better for me... I wish I have patience to change faction now...