DPS Spreadsheets

[Motenten]

Have done a rewrite and update on the sam spreadsheet. Added R2/R15 Mikinaak gear, and allowed splitting subjobs/food/etc between set 1 and set 2.

I set up the two gear sets to check out the conditions again, leaving them mostly identical aside from the body/hands/feet. Target is Delve Fodder, since that was one of the samples used for ccl's comparison.

Changing to Usu feet+Usu body capped haste for that set, so both sets are at 90.1 delay, negating that difference.

As I was going through and reapplying buffs to make sure they were the same, there was a pretty constant 17 DPS difference in Portus's favor until I added Minuets. Checking, it seems the Portus build's higher attack (53 att lead for Porthos/Phorcys/Whirlpool compared to Usu/Kari/Usu, after accounting for Berserk/Chaos) gave it a bit of an edge up until attack is capped. Once attack is capped, the difference narrows to just 3 DPS (still in Porthos's favor, though).

Next was a change in food. Double minuets puts you far enough past attack cap that RCBs do very little for you. With all the buffs, you'd still be capped even without any food at all; the only gain is from the str. So I tried swapping to the suggested Carbonara.

Porthos: 1170.935 => 1170.978
Usukane: 1167.604 => 1165.291

For the porthos build, there's no change. For the Usukane build, total damage actually drops. This is where things get tricky -- how much is gained and lost with changes like these depends on subtle variances from exact multi-attack rate (and thus average number of rounds per weaponskill), Save TP, and Store TP values. Finding the exact point where things might tilt in favor of one build or the other (assuming there are actual tipping points) can be difficult.

One such difference was the realization that I had Hunter's Roll on, not Samurai Roll. Switching those over, and suddenly there's a precipitous shift:

Porthos: 1184.798
Usukane: 1261.238

Why this shift? One reason is that by removing Hunter's Roll, the Porthos build falls below capped accuracy for the exact mob target value. It's only 5 points though, so I'll add an arbitrary 5 acc to keep things capped for the comparison. That only increases Porthos to 1199.585, however. The remainder is due to the Usukane build dropping to 2.13 rounds per weaponskill, compared to 2.72 rounds for the Porthos build. Without Samurai Roll, Porthos is at 2.74 rounds per weaponskill (ie: almost no change), while the Usukane build is at 2.78 rounds (higher than the Porthos build).

The minimum Sam Roll needed to make this shift is +25 Store TP. That puts normal melee hits at exactly 25 TP per hit, giving you a 4-hit build with the 25 Save TP. The Porthos build only manages 24.3 TP per hit with +35 Store TP from Sam Roll. The Porthos build would need an additional 6 Store TP to tip over into a 4-hit build, or 41 Store TP from Samurai Roll. That can be reached on a Lucky or 11 roll, giving 32+10 or 40+10, for 42 or 50 total +Store TP.

In the event that you get a lucky roll, things shift back to Porthos being slightly ahead:

Porthos: 1271.481
Usukane: 1265.948

So the main edge the Usukane build has is being nearly guaranteed to reach that tipping point, given that it will work with pretty much any safe Samurai Roll value, whereas the Porthos build needs Lucky or 11.

Note that the above numbers took a conservative value of 10% for Fighter's Roll. Higher roll values there increase total DPS, but the relative standings of the two sets remain the same.

Shifting the Usukane set back to RCBs does increase total DPS by a very small amount (1268.693), but removes a couple safe rolls from what will tip it over into a 4-hit build.

This optimization always depends on getting Lucky or 11 on Miser's Roll. The Porthos build needs Lucky or 11 on Samurai Roll. Poor values from Chaos Roll (dropping you under capped attack) favor Porthos (though not enough to make up for an insufficient Samurai Roll). Fighter's Roll generally favors both equally.


It's certainly easier to reach this DPS level with the Usukane build, though technically the Porthos build still has a higher max damage potential, unless accuracy comes into play; the Usukane build has a substantial 41 acc advantage.

[shepardG]

Very good work mo1010 ^^

I'm working on my usukane +1 body now,. But it's nice to know that porthos isn't a waste of an inventory slot. I'll have both for up/down roll/acc situations.

[Motenten]

Working my way through the spreadsheets to update the stat conversion changes. Will update as they're completed. Note that this does not include any gear updates.

Done so far:
War
Blu (includes club for offhand)
Bst
Mnk
Cor
Dnc
Thf (includes additional bug fixes and sword offhand)
Drk
Pup
Nin
Run
Pld
Rdm
Whm
Rng
Brd

~ Complete

Also, Magic Damage has been added to the magic damage spreadsheet.

[Fake]

Not sure MNK sheet has been uploaded, every other one you listed as done has been updated in the last 20hrs. Mnk is at 5 days since last.

[Motenten]

Uploaded a fresh copy, just to be sure.

[Byrthnoth]

The one I downloaded last night was also missing Miodio gloves, though I'm not sure they really have a use for monk.

[Vinnyv11]

Any update on the drg schedule pending?

[Motenten]

Drg and sam shouldn't need to be updated at all. They aren't affected by the 1-handed changes. (Gear updates will be an entirely separate pass.)

Miodio added to mnk.

[lolwut]

Can you make the Cor sheet calculate dps for QD? It doesn't seem to do so currently, and doesn't appear to have a way to track the dps increase from (or even add) -QD delay or the double tap from af3+2 feet.

This is less a request, and more something I was just wondering about for DPS calculations in general: would it be possible to calculate the average expected value from rolls, modified by gear/party members? I know people have done individual calculations in the past, but something that would be able to account for hat proc, toggles for jobs in/out of the party, stopping after 6, etc. would be really handy, and give a more realistic baseline to build DPS sets around. It seems like it would be a lot of hassle, but I wanted to get your take on the pros and cons of it.

[Vinnyv11]

Drg and sam shouldn't need to be updated at all. They aren't affected by the 1-handed changes. (Gear updates will be an entirely separate pass.)

Miodio added to mnk.

Thanks for the update didn't know if the Wyvern changes are factored but realized the spreadsheet is absent there impact. Thanks for the great update on thf schedule.

[futan]

I get a #NAME? error when trying to change the weapon on the DRK spreadsheet. How exactly can I fix that? D:

Thanks! :D

[Motenten]

New version of drk spreadsheet uploaded that should fix that.

Note that I'm still getting an odd error when I save the drk spreadsheet as .xls, however it gives me know clue as to where the issue may be occurring in the spreadsheet, so I haven't been able to fix it. If you come across anything else odd, let me know.

@lolwut: on the cor stuff, I'll consider it, but it's not a high priority at the moment. Particularly the rolls are difficult to predict an 'average' for, due to different user behavior. I may put together a sheet just for that at some point, but those kinds of calculations are a mess, so not something I want to work on right now.

[lolwut]

Totally understandable. Thanks for taking it under consideration though, and for all the awesome work you put into these sheets already. They're a huge boost to the xi community.

[Motenten]

All spreadsheets have finished the conversions.

When I start on the gear updates, I'll probably start to remove all the non-99 magian weapons to reduce clutter a bit.

[Motenten]

New gear added to all the spreadsheets. New 'combat skill' aspect of some weapons implemented.

[Zirael]

Hi Motenten, thank you for keeping it going!
One small thing I've noticed so far: in Magic spreadsheet Augur's Gaiters give 0 Fast Cast. BGwiki lists them as 3%.

[Motenten]

Updated Augur's Gaiters.

[fattrav]

Thanks for your hard work!

[Fireon]

The Mondaha Cudgel doesn't do the STR/DEX paths, it goes MND/INT. You might want to fix that on the BLU spreadsheet. Also missing Letalis Mantle and Halakaala.

[Motenten]

Blu updated.