Jobs Where Non-REMA Is BiS.

I actually tend to think there may be a similar issue for 1-hander modelling, but it just isn't quite as prominent.

Melee damage is being extremely over represented in the current spreadsheets. We're at 80-90% of our damage from WS on the high end. Also the spreedsheets assume a situation that doesn't exist, Undispellable perfect buffs with infinite duration on a target with infinite HP that doesn't fight back. That's how you get "over 9000DPS" on spreedsheets but 3-5K DPS during actual fights with it dipping to 1K during certain times.

Those sheets were originally modeled for fighting greater colibri for merits or similar situations at level 75. All that's been done is updating variables,buffs and mechanics but underlying assumptions were the same. They are a good tool but people that don't know game mechanics take them as gospel FAR too often.

Honestly I dont use a spreedsheet to often anymore, they do seem to calculate magic dmg like leaden well, but thats about it. What I use is apex crabs with the same trusts buffs for me. I find parsing my dmg VS my dmg gives me the best results for real dps increases.

We start fights with 3000 TP. Fights last for 30 seconds. SAM has Meditate which heavily skews towards WS, especially in 30 second fights. For zergs, SAM has SP1/2, again, massive advantage to WS over melee damage since it isn't accounted for in the spreadsheets either. It doesn't mean melee damage is necessarily wrong, but that these unaccounted for factors just add a ton more WS damage to shift the melee:WS ratio. Not saying there might be something off(especially how going from 4hit to 3hit doesn't show much of a decrease in melee damage on the sheets), but these are certainly things to consider.

The overall DPS is the least useful part of the spreadsheet. The data tab is far more useful, people can see the effects of various thinks like fSTR, average hits per round, cRatio and min/max for pDiff.

We didn't begin any of those Neak fights with 3000 TP or use any SPs. The white damage for our Doji SAM was 13.5% of his total damage around around 10% on WoC. So those factors do play a role yeah but they are certainly not enough to explain the gap we see in modeling vs measurement. I'm skeptical that Empyrean weapons (Masamune included) are nearly as good as we have modeled them to be.

In regards to how useful the spreadsheets are, I still consider them more accurate than napkin math although I usually only use them for selecting WS gear.

The most common thing to do with the spreadsheets is assume capped accuracy and attack then see what variations of multi-attack and STP gear/buffs will win under those situations. Obviously if you're uncapped accuracy or attack adding more through gear will be better than the "capped" sets, but assuming capped accuracy/attack this day in age is absolutely a realistic thing to do.

To actually disprove your personal set you feel is better or use napkin math and only factor 3 things(accuracy/hits per round/xhit), you would have to run 1000s of parses using almost unrealistic control factors to eliminate player error. This is essentially what the spreadsheet does for us, which we obviously can't emulate perfectly without all sorts of outside factors influencing the real results we see from scoreboard.

I find it difficult to believe you actually have sufficient data to prove one specific set the spreadsheet predicted would do poorly actually did measurably better in practice that isn't attributed to player error.

What does this have to do with the topic at hand? Sounds too much like a personal agenda and trying to drag another argument into this thread.


The spreedsheets just use basic modeling against a target with infinite HP that doesn't fight back using static buffs that last an infinite duration, there isn't a single fight in the game like that. Those sheets were made for fighting Colibri and were just kinda updated to be generic community tools for people who don't have deep knowledge of the damage mechanics in FFXI and how each component is interrelated. They were never authoritative, nor meant to be absolute.

Example, many NM's have moves that inflict accuracy / attack down, or blind, stat down, amnesia, stun or various other status ailments that effect your total damage yet aren't modeled. Using JA's creates a 2s pause in the attack round and you need to use those JA's, but that's not calculated or accounted for. Then we have strategic considerations like not immediately going hog wild on 18-man WoC otherwise you'll have hate when it does call wyvern and you die a horrific death.

Thus my advice to anyone using those sheets is to be spend more time analyzing the data tab then then final DPS tab to better understand the context of the results.

JA delay is actually accounted for in the spreadsheet afaik although I don't know the details.

Just checked the WAR spreedsheet and all I can see is a percentage, which isn't particularly useful because your not fighting a NM with infinite HP. Plus you don't use all your JA's the moment their up, some you need to time a bit to get max benefit out of. That's really hard to model accurately so it's just something people gotta learn to judge for themselves.

You might be thinking of Kparse, which was out during the Colibri era. I don't recall spreadsheets making an appearance from Mote until Abyssea.

And yes, they always came with the exact same warning you're trying to spin as your original thought: they aren't a perfect representation of actual DPS, but a very good tool to compare sets under similar buff conditions/theorized mob stats.


What's your argument here? For players to fulltime a super high accuracy set that is overkill 95% of the time? As far as I know, most people have 3-5 levels of accuracy to account for any situation their accuracy dips from debuffs/dispels while maintaining ideal DPS. Are you arguing TPing in resist amnesia gear, because not sure why you'd add that since it affects all TP sets equally.

I'm curious who you think is saying the DPS sheets are exact and precise indicators of real DPS. They are used to compare sets. Just because the DPS number shown isn't a carbon copy of what we see on scoreboard doesn't invalidate what set the sheets predict will perform better overall. If you have a better algorithm/tool/program/data to provide us that is better than napkin math and feels(your 'game knowledge') to discredit the spreadsheets, please share.

Any problem that can be solved via swapping your gear in this game does not actually constitute a problem that should be considered anything more than a joke to anyone hearing it used as an excuse.

Exception: Summoner needs the Mythic for it Avatar level +2!

I'm still a tad confused where you are getting this problem from...

AM3 is pretty crazy on multihits too. And well all physical bps. But really nowadays smn isn't really classified as a much but a pet job

Since the easier AG update, REMA weapons have been the way to go for all Melee

The problem was that you were presenting aymur's tpbonus going above 3000tp when using tpbonus gear during ready as a legitimate con despite the fact that you can easily just remove tpbonus gear (for better equipment) when needed in other slots. Then when this was pointed out as not being a problem you doubled down and/or decided to say that you can't see why anyone has a problem with your logic.
You were presenting a thing that is not a problem (due to solutions existing since the game was created,designed, and released) as one and now you're possibly feigning ignorance when it's being pointed out.

That "problem" was a small math error only in the first post and error I might add that was in Aymurs favor (specifically I had switching to Aymur as giving an extra 500 tp forgetting to put back in the tp bonus from the alluvion axe) and wasn't actually the problem I was focusing on in that post.

If you had actually read the posts you'd see I actually mathed out non capped tps showing how small the gains are at those higher effective tp values and never at all presented an argument of it being bad purely because the tp was going to waste. Hence why I don't see a problem with what I actually posted minus the small error in 1 post that was actually was worse for my argument. Good job at cherry picking though