|
All Jobs Damage Simulator and Gear Sets
By Veydal1 2024-09-19 01:43:55
Is the 'Initial TP' field working properly? Or am I misunderstanding how it's used? It doesn't seem the simulation for DPS takes this into account. My impression is it would use it to help determine average TP for WS, but that doesn't seem to be the case. Is the value used for anything?
If I set initial TP based on average TP return from WS, and then run Optimize TP set with the TP Metric of 'Time to WS', I would think the TP set generated would be one that is fastest to hit the TP threshold defined, while taking into account the Initial TP value.
Instead, it seems the simulation is only taking the Average TP/Round value and calculating that against the TP thresholds. So if you get an Average TP/Round value of 300 from your TP set, and you have a minimum TP threshold of 1000, it'll calculate your average TP for WS at 1200, since 300 + 300 + 300 + 300 = 1200. On average, 4 swings. However, the results are the same, regardless of what value you put in for Initial TP. If you put a value of, say, 100 TP, the simulation still puts you at an average of 1200 instead of ~1000 (100 + 300 + 300 + 300).
By SimonSes 2024-09-19 02:33:04
My impression was that initial TP field is only used for "Quicklook TP" and maybe for first round on simulation, but every other round initial TP is calculated from TP return on WS that was simulated, but I haven't look that deep into the code.
By Izanami 2024-09-21 16:17:52
For party buffs / additional buffs, I had the following in mind: Thanks for the list. I've added the following additional buffs and debuffs to the "Special Toggles" list. You can mouse-over the text for each and view their bonuses as tooltips in the GUI itself. This list now has a lot of checkboxes for buffs available to all jobs. I've organized it such that the abilities available specifically to the selected main+sub jobs are displayed first. Abilities available to "all jobs" are displayed afterwards.
"Blood Rage (WAR)" provides +20% Crit Rate to all jobs. This also provides +20% Ranged Crit Rate to WAR main from job point gifts. I've also corrected an issue where Blood Rage was previously giving WAR main +40% Crit Rate instead of +20% Crit and +20% Ranged Crit.
"Warcry (WAR)" provides "TP Bonus +700" and "Attack +11.3%" to all jobs. The original "Warcry" toggle is still selectable for WAR subjob and provides attack+% based on your WAR subjob level.
Crimson Howl (Ifrit buff) provides +11.3% attack to all jobs.
Crystal Blessing (Shiva buff) provides +250 TP Bonus to all jobs.
Ifrit's Favor provides +25% Double Attack.
Shiva's Favor provides +39 Magic Attack Bonus.
Ramuh's Favor provides +23% Crit Rate.
"Haste Samba (DNC)" provides +10% JA Haste to all jobs. "Haste Samba" is still selectable for DNC subjob and provides +5% JA Haste.
"Box Step (DNC)" applies "Defense -23%" to the enemy. "Box Step" applies "Defense -13%" to the enemy and is only selectable by DNC subjob. These stack multiplicatively with Dia, Corrosive Ooze, Blade: Kamu, etc
Nature's Meditation provides +20% Attack to all jobs.
Mighty Guard provides +15% Magic Haste and +15 Magic Defense to all jobs.
Corrosive Ooze applies "Defense -33%" to the enemy. This stacks multiplicatively with Dia, Box Step, Blade: Kamu, etc.
Swooping Frenzy (BST ability) applies Magic Defense -25 and Defense -25% to the enemy.
For BST main job: Frenzied Rage provides +25% attack, which I assume stacks with BST's "Rage".
For BST main job: Rage provides +50% attack, which I assume stacks with BST's "Frenzied Rage".
Note that selecting Warcry, Warcry (WAR), Blood Rage (WAR), or Crimson Howl will automatically unselect the other options since these effects do not stack.
I do not know anything about SMN, so I've simply used the "Favor +9" values you've listed.
I found a small error with relic hidden proc. Not all relics has ODT with 13% chance. Some of them have 16% for x2.5 and some of them have ODD with 20% chance. Thanks. I didn't realize that different relics had different hidden damage bonuses. I've pretty much copy/pasted what you provided to update the relic and prime hidden damage bonuses in the code.
From what I can tell, the ranged relic weapons with Barrage are already handled properly according to BG Wiki. BG Wiki states that the hidden effect on Yoichinoyumi and Annihilator affect the entire Barrage damage. BG Wiki isn't clear on if the hidden damage applies to both hits of Double Shot, but you could argue the wording suggests that only the first hit of Double Shot is affected. I've adjusted the code to only boost the first hit from Double Shot while still boosting all hits from Barrage.
Is the 'Initial TP' field working properly? Or am I misunderstanding how it's used? The "Initial TP" value is used only when finding the best "time to WS" TP set and when using the "Quicklook TP" button to help simulate starting with the TP returned from a previous weapon skill while finding the best sets. I rarely see it make a difference in the automated "best" TP set unless the value you input puts you under some X-hit threshold where a small amount of Store TP would be worth more than multi-attack (or vice versa) while not sacrificing accuracy. I think it's probably best to set "Initial TP" = "TP Returned from WS" if you know how much TP return you expect, but it probably doesn't matter much to leave it at 0.
Other changes:
I was pointed to the DRG thread and realized I was not actually using Conserve TP, despite having the DRG and DNC traits and relevant gear stats in place already. I've implemented Conserve TP procs based on the details on the BG Wiki page (x% chance to conserve 10-200 TP randomly after each weapon skill).
I've added a "Verbose DPS" button in the Settings menu at the top of the GUI. Enabling this will print a summary of each attack round and weapon skill used for DPS simulations, including total damage dealt and total TP returned. This should help debugging and clear up any confusion that was mentioned in the DRG thread.
- Verbose DPS output will print multi-attack, daken, zanshin, and kick attack procs, but not print the damage dealt by each hit individually. I could easily add individual auto-attack damage per hit breakdown (to look like a real in-game chat log for example), and probably will later, but I probably wouldn't be able to break down individual multi-attack weapon skill hits the same way.
- Fotia, Conserve TP, and average Regain over time are also printed when they activate. This makes it easy to track TP gained and used throughout the simulation.
I've added a "Verbose Swaps" button in the Settings menu at the top of the GUI. Disabling this will hide the text spam when automatically finding "best" sets, so the code will no longer show the incremental damage improvements for each equipment swap tested.
- Note that printing outputs is slow in general for all programming languages. Enabling the verbose modes may increase runtime of the code by a marginally noticeable amount due to the number of print statements used.
I've fixed an issue where Fotia procs were returning (base_tp + tp_bonus) for DPS simulations.
I've fixed an issue where using Focus with subjob MNK would not provide the bonus stats.
Known issue:
It seems that the Simulations tab equipment selection box has a list size issue. Equipment slots with more than 41 options can not be scrolled to their end to see all available options. For example: DRK has 49 main-hand weapons available, but the scrollbar in the Simulations tab only shows the first 41. I'm not sure why this is, but the workaround for now is to equip the items in the Inputs tab, then use the "Copy to TP/WS set" buttons to move them to the Simulations tab.
Let me know if there are any other issues.
Edit: Link to Github page: https://github.com/IzaKastra/wsdist_beta
By SimonSes 2024-09-21 16:31:17
Thx a lot Izanami! :)
By Veydal1 2024-09-22 07:53:28
Absolute legend. Thanks for all the effort and work you put into this.
By Izanami 2024-09-22 16:15:28
I've added a "Very Verbose DPS" button in the Settings menu. This button enables printing a breakdown of every individual attack for attack rounds and weapon skills and highlights crits, missed hits, and fotia/conserveTP procs in color. Naturally, this creates a ton of output text, but it provides much more detail for debugging. "Verbose DPS" is still an option if you only want to see summary values for attack rounds and weapon skills as shown in my previous post.
I needed to slightly adjust how "Double Attack Damage+" and "Triple Attack Damage+" are used in the DPS simulation for the Very Verbose outputs. This wasn't a problem, but it made me question if these stats affect weapon skills. The code currently does not apply these bonuses to weapon skill multi-attacks, but let me know if I should change this.
Other changes:
I've adjusted the Verbose DPS output to print integer damage values, instead of showing 1 decimal place. I think this makes the output easier to read, but it does introduce some "off-by-one" rounding errors in the printed output that you really shouldn't notice anyway on damage dealt. This does not affect any calculations being done.
Endark II and Enlight II now deal magic damage on TP swings based on Enhancing (dark/divine) magic skill. I can't recall if these affect off-hand hits, so I've assumed not for now. Dual-wielding on DRK or PLD is not very common anyway.
I've made it so the one-time bonuses from sneak/trick attack and the various flourishes do not apply during DPS simulations. Previously, the code would apply these bonuses to all weapon skills, which didn't make much sense.
I've updated Oshala's (Prime Staff WS) WSC and FTP values.
Note that the critical hit, fotia/conserveTP and missed hit procs print in color to stand out more. In Windows10 Powershell this is no issue, but if you're not using the .exe file and if you're running Python in a normal command line, then you'll need to "pip install colorama" or the colors will not appear and broken text will appear around the colored words.
Ramuh.Austar
Server: Ramuh
Game: FFXI
Posts: 10481
By Ramuh.Austar 2024-09-22 18:31:52
Double and triple attack damage do not apply to weapon skills.
[+]
By Moonlightagb 2024-09-23 08:28:54
Is Marin Staff +1 taking into account the Wind Elemental Bonus+11? I'm not seeing that stat when I hover over it?
Edit: I was also doing some testing with Quanpur Necklace and it looks like the 5 Earth affinity isn't doing anything even though I see the stat listed on the item's stats
[+]
By Izanami 2024-09-24 16:02:22
Is Marin Staff +1 taking into account the Wind Elemental Bonus+11? I'm not seeing that stat when I hover over it? Marin Staff was definitely not applying any bonus to Wind elemental spells in the code. Do we have confirmation that it's +11% damage or +11 Magic Attack? I vaguely recall testing this myself in game ages ago and finding that it was simply +11 Magic Attack, but I may be confusing it with some Ninjutsu testing that I previously spent a lot of time on. BG Wiki claims Marin is +11% damage, but BG Wiki is occasionally incorrect.
For now, I've assumed BG Wiki is correct and added "Wind Elemental Bonus +11" to the Marin Staff to add +11% damage (x1.11 multiplier) to Wind element magic.
Edit: I was also doing some testing with Quanpur Necklace and it looks like the 5 Earth affinity isn't doing anything even though I see the stat listed on the item's stats Yep. Looks like I forgot to capitalize the element name when I adjusted the nuking part of the code ("Earth Elemental Bonus" vs "earth Elemental Bonus"). This affected all element-specific equipment, but did not affect Orpheus Sash. This issue has been fixed and everything seems to be working now after testing spells/WSs of each element.
By Veydal1 2024-09-24 16:31:16
Is Haste Samba buff from DNC applying properly if DNC is selected as the main job? Seems to be working fine if a non-DNC main job is selected.
[+]
By Izanami 2024-09-24 16:55:47
Is Haste Samba buff from DNC applying properly if DNC is selected as the main job? Seems to be working fine if a non-DNC main job is selected. Nope. Which is crazy because that's one of the things I tested explicitly a few days ago.
Looks like I put an if-statement in to not apply "Haste Samba (DNC)" to DNC main job for some reason. I have no idea why I did that.
This issue has been fixed. I guess I'll have to retest everything else I added recently just in case.
By SimonSes 2024-09-24 19:09:14
I think its the same for warcry(war) on war?
[+]
By Izanami 2024-09-24 19:31:45
I think its the same for warcry(war) on war? Yep, and I see why I did it now. I was trying to hide the subjob buffs from the main job options, so DNC main couldn't use /dnc Haste Samba and WAR main couldn't use /war Warcry, as a way to remove redundant buffs from the list.
I switched how I wanted to handle this situation last minute and forgot to remove the old method.
I think Warcry and Haste Samba were the only two "all jobs" buffs to have potency increases for main job.
[+]
By SimonSes 2024-09-25 02:07:37
I think its the same for warcry(war) on war? Yep, and I see why I did it now. I was trying to hide the subjob buffs from the main job options, so DNC main couldn't use /dnc Haste Samba and WAR main couldn't use /war Warcry, as a way to remove redundant buffs from the list.
I switched how I wanted to handle this situation last minute and forgot to remove the old method.
I think Warcry and Haste Samba were the only two "all jobs" buffs to have potency increases for main job.
Tbh I noticed warcry instantly after your upload, but I just fixed it by adding regular warcry back, when main job is war and completely forgot to pass the info to you (it was very late), so good that Veydal1 wrote about Haste Samba, because that reminded me about warcry.
Fenrir.Ahlen
Server: Fenrir
Game: FFXI
Posts: 261
By Fenrir.Ahlen 2024-09-26 14:37:15
"Blood Rage (WAR)" provides +20% Crit Rate to all jobs. This also provides +20% Ranged Crit Rate to WAR main from job point gifts. I've also corrected an issue where Blood Rage was previously giving WAR main +40% Crit Rate instead of +20% Crit and +20% Ranged Crit.
I'm pretty sure this is 40% crit not ranged crit. The only source I saw that its ranged crit is BG wiki's War job page. In game it just says attacks and on the blood rage bg wiki / jp wiki page both say the job points give 20% crit.
By Shiva.Myamoto 2024-10-03 10:05:17
Not sure how the numbers work out, but I use Sroda earring instead of thrud for decimation on BST. 7% double attack just makes more sense to me given its ftp replicating and the boost from nukumi body applies to any damage/all hits (at least in my experience anyhow).
I don't have python installed so if anyone has any feedback, that would appreciated. Probably minor nonetheless, but I am just curious as to Thrud actually "winning" on that posted set.
By Izanami 2024-10-13 17:43:15
I'm pretty sure this is 40% crit not ranged crit. The only source I saw that its ranged crit is BG wiki's War job page. In game it just says attacks and on the blood rage bg wiki / jp wiki page both say the job points give 20% crit. That does sound more likely than WAR JP gifts giving Ranged Crit Rate. I've unfixed this issue, so Blood Rage is +20% Crit Rate, with an additional +20% for WAR main.
Not sure how the numbers work out, but I use Sroda earring instead of thrud for decimation on BST. 7% double attack just makes more sense to me given its ftp replicating and the boost from nukumi body applies to any damage/all hits (at least in my experience anyhow).
I don't have python installed so if anyone has any feedback, that would appreciated. Probably minor nonetheless, but I am just curious as to Thrud actually "winning" on that posted set.
The sets on BGWiki are from an older version of the code (but are probably still very good sets). If I run BST/NIN Decimation sets on the most up to date version of the code, I find that:
If I allow all gear to vary, then the Decimation set shows Thrud+Lugra=31322 average damage
If I force one of the earrings to be Sroda, then the Decimation set shows Sroda+Sherida=31198 average damage, essentially zero difference, especially if you consider that the code is probably only accurate to within a few percent.
This is also surprising to me after considering that Decimation is an FTP transferring WS, where multi-attack is usually better than WSD. Decimation is already a 3 (+1 off-hand) hit weapon skill, so an additional hit from a DA, while still a large increase in damage, is not worth as much percentage-wise as a DA on a 1-hit FTP transfer WS. I don't see an issue with Decimation in the code, so I suppose the STR+10 and WSD+3% on Thrud are roughly equivalent to DA+7% on Sroda for Decimation.
I don't have python installed You actually do not need Python installed to run the code, but you will need Windows10 or Windows11 or things might break. Downloading the files and running the executable (gui_wsdist.exe) will allow you to run simulations without Python installed.
Downloading and running a potentially sketchy executable is generally not recommended, so I've included a GitHub "Action/Workflow" to generate the executable directly on GitHub in a completely transparent way, so you can check that nothing potentially dangerous has gone into it. The executable generated by GitHub is found on the "Actions" page ( https://github.com/IzaKastra/wsdist_beta/actions), and the exact code used to generate the executable is also provided for each file version.
[+]
Necro Bump Detected!
[42 days between previous and next post]
By Izanami 2024-11-24 10:19:34
I've finally updated the sets presented on BG Wiki (see the main page) based on the most recent version of the code. The sets previously posted on BG Wiki were based on a version of the code from over a year ago, which has since seen many updates and corrections. The individual gear sets on each job page now include information on the Master Level, subjob choice, TP used, and the date the set was last updated. This information is visible after clicking the "Expand" button at the bottom of each set. There are still some sets missing from the BG Wiki guide. I'll slowly add these some other time.
Additionally, the main All Jobs Gear Sets BG Wiki page has been updated to include basic information on how each set was calculated, such as enemy stats.
I haven't heard about any issues recently, so there have been no major updates to the code itself. The small updates for the code are:
Added a "hidden" function to print gearsets in BG Wiki format to make the previously-mentioned updates faster. This feature is only available if you manually change a "False" to "True" in the code.
Added a checkbox to the optimize tab which, when enabled, automatically selects Nyame R25B instead of R30B when using Odyssey Rank = 30. This simply saves users from having to manually deselect R30 and then select R25 for Nyame if they haven't completed the final Bumba challenge, but have R30 everything else.
By Kaffy 2024-11-30 15:39:54
Hi Izanami, I noticed on the bg wiki page for scholar, the helix sets use obi. I believe Skrymir Cord +1 should be there instead since helices always get weather bonuses.
[+]
By Izanami 2024-11-30 16:22:08
Hi Izanami, I noticed on the bg wiki page for scholar, the helix sets use obi. I believe Skrymir Cord +1 should be there instead since helices always get weather bonuses.
Thanks for catching this. I've re-updated the code to once again always use weather bonuses when casting a helix spell if the correct weather is enabled. I've also rerun the nuke/burst Helix sets for SCH and updated the BG Wiki sets.
[+]
By Minaras84 2024-12-19 16:52:55
Just wondering if the Bst one is accurate?
It differs alot from Falkirk guide.
There are also pieces of gear in both guides I don't understand, like Ankusa helm +3 instead of nyame at rank 25 or 30
[+]
By Nariont 2024-12-19 17:01:42
1% more wsd, 3 more atk, 6 less str, 11 less dex(for those few dex based axe skills) on nyame r30 vs ankusa +3
By Manque 2024-12-19 17:59:41
For the nuking sets for mages, the waist slot has O.Sash. Given the distance requirements for this item I don't feel it makes much sense to include. Can we run things again without?
[+]
Server: Asura
Game: FFXI
Posts: 464
By Asura.Dexprozius 2024-12-19 19:01:39
Corsair Evisceration (High End) looks like copy paste of Leaden Salute by mistake
[+]
By Minaras84 2024-12-19 19:59:45
1% more wsd, 3 more atk, 6 less str, 11 less dex(for those few dex based axe skills) on nyame r30 vs ankusa +3
I understand that purely for ws purpose it is slightly better, but is ignoring the SC bonus entirely a thing?
Usually, one doesn't wear just one piece of Nyame, which means more SC bonus.
I don't know, just a thought
Asura.Bippin
Server: Asura
Game: FFXI
Posts: 1090
By Asura.Bippin 2024-12-19 20:19:51
***is situational. The math is not looking at SC bonus at all.
If you are asking if you should have a different WS set for when you are closing skillchains to do more damage then the answer is yes.
[+]
Server: Asura
Game: FFXI
Posts: 146
By Asura.Illuminate 2024-12-19 20:50:08
It's amazing after all these years how far it's come where we can simulate our dps and optimize gearsets from code written by the player base. Just a little nostalgic here.
... Imagine if we combined our powers for good instead of SE
By Izanami 2024-12-19 21:08:45
Just wondering if the Bst one is accurate?
It differs alot from Falkirk guide.
There are also pieces of gear in both guides I don't understand, like Ankusa helm +3 instead of nyame at rank 25 or 30 You are correct in that there was an issue with the BST sets.
I'm not sure how Falkirk determined those the sets in the BST guide, but I do not see a mention of a spreadsheet or simulation used so it may simply be a by-eye estimate from in-game situations with some intuition based on stats and WSC. The BST guide is one of the few well organized and maintained guides, so those sets are probably at least close to best in slot for high buff situations.
While checking this, I noticed that the gear.py file used BST +2 Empyrean body twice, instead of including both the +2 and +3 versions (I overwrote the +3 version when I added the +2 version). This caused the code to never select the BST Empyrean+3 body for testing, despite how good it clearly is based on its stats. I've rerun all of the BST sets after correcting the issue. The sets I'm finding are still a bit different than Falkirk's, but that does not worry me.
Since I was rerunning things anyway, I've also added Barbarity +1 to the code since it was recently mentioned in some other job thread. I'm not sure how well it performs for TPing, though.
For the nuking sets for mages, the waist slot has O.Sash. Given the distance requirements for this item I don't feel it makes much sense to include. Can we run things again without? Oops. I've rerun the BLM and GEO nuke/burst sets without Orpheus or Hachirin. Looks like Skrymir, Acuity, and Sacro are all decently close to each other, depending on the element and tier of nuke you're using. Enabling weather for each set shows that only the waist changes to Hachirin-no-Obi, while all other slots remain the same.
I think most content these days would heavily favor Acuity +1 R15 for its high Magic Accuracy and INT+.
Corsair Evisceration (High End) looks like copy paste of Leaden Salute by mistake Thanks. That was certainly the case. I've updated the COR Evisceration sets too.
The math is not looking at SC bonus at all. This is correct. The code's automatic set finder ignores skillchain damage. I could probably add a checkbox to consider WS+SC damage together since the rest of the code already includes Skillchain Bonus stats and traits. An unpublished test version of the code does use 2-step skillchains in the DPS simulations, but I haven't touched this in a while since I wanted to include up to 4-step skillchains.
By Minaras84 2024-12-19 21:27:34
Just wondering if the Bst one is accurate?
It differs alot from Falkirk guide.
There are also pieces of gear in both guides I don't understand, like Ankusa helm +3 instead of nyame at rank 25 or 30 You are correct in that there was an issue with the BST sets.
I'm not sure how Falkirk determined those the sets in the BST guide, but I do not see a mention of a spreadsheet or simulation used so it may simply be a by-eye estimate from in-game situations with some intuition based on stats and WSC. The BST guide is one of the few well organized and maintained guides, so those sets are probably at least close to best in slot for high buff situations.
While checking this, I noticed that the gear.py file used BST +2 Empyrean body twice, instead of including both the +2 and +3 versions (I overwrote the +3 version when I added the +2 version). This caused the code to never select the BST Empyrean+3 body for testing, despite how good it clearly is based on its stats. I've rerun all of the BST sets after correcting the issue. The sets I'm finding are still a bit different than Falkirk's, but that does not worry me.
Since I was rerunning things anyway, I've also added Barbarity +1 to the code since it was recently mentioned in some other job thread. I'm not sure how well it performs for TPing, though.
For the nuking sets for mages, the waist slot has O.Sash. Given the distance requirements for this item I don't feel it makes much sense to include. Can we run things again without? Oops. I've rerun the BLM and GEO nuke/burst sets without Orpheus or Hachirin. Looks like Skrymir, Acuity, and Sacro are all decently close to each other, depending on the element and tier of nuke you're using. Enabling weather for each set shows that only the waist changes to Hachirin-no-Obi, while all other slots remain the same.
I think most content these days would heavily favor Acuity +1 R15 for its high Magic Accuracy and INT+.
Corsair Evisceration (High End) looks like copy paste of Leaden Salute by mistake Thanks. That was certainly the case. I've updated the COR Evisceration sets too.
The math is not looking at SC bonus at all. This is correct. The code's automatic set finder ignores skillchain damage. I could probably add a checkbox to consider WS+SC damage together since the rest of the code already includes Skillchain Bonus stats and traits. An unpublished test version of the code does use 2-step skillchains in the DPS simulations, but I haven't touched this in a while since I wanted to include up to 4-step skillchains.
Fantastic, thanks!
Server: Asura
Game: FFXI
Posts: 464
By Asura.Dexprozius 2024-12-20 09:11:13
Is there a reason why the "All jobs gear sets" cant include TP sets for the jobs? It'd be nice to see a one stop shop for starting points for new jobs you pick up along the way. Maybe Max TP / Hybrid DT as 2 options for people to reference. Or is that not easy to plug into a calculator?
|
|