E5una
commented
2 years ago Update: Arcanekings was nice enough to donate his time to do some retail collection on Overdrive with me. He used a level 40 animator, no other gear on his master. The automaton was a Valoredge Head + Body to match the previous data we have, and the only attachment he had on was Reactive Shield (to test something completely unrelated, since all it does it give Blaze Spikes).
Defensive breakdown of the screenshots that I'll post at the end of this:
The first column is the Boss's attacks against his automaton
The second column is the Boss's attacks against his automaton during Overdrive
The third column is the enemy Sharpshot Automaton's ranged attacks against his auto (it was parked at a distance)
The fourth column is the enemy Sharpshot Automaton's ranged attacks against his auto during Overdrive (it was parked at a distance)
As you can see, the Boss's melee damage is nearly identical with and without overdrive. Admittedly I would like to see a bigger sample size of ranged attacks here, because there is some questionable variation, but its hard to conclude that its not just the small sample size and randomness coming into play. I will say that it is weirdly suspicious in this tiny sample, but its going to be hard to get more data vs a ranged automaton since you can only do this every 5 days and its random which auto you fight.
Note: In the melee side of things, I omitted a crit of 86 and 93, as I wasn't sure how to include crits in the data. These were BOTH DURING OVERDRIVE, so if anything, the damage taken during overdrive would have looked worse.
Offensive Breakdown:
Very hard to draw any conclusions, its a really small sample size and I don't think we can do much with it.
In my opinion the two 25 dmg hits must be Guard procs by Fantoccine, but the sample size is too small to draw any real conclusions. If you remove the two 25 procs (assuming these are guard procs) the sample size is abysmal but goes back to making it seem like there is no extra benefit of overdrive besides 3x maneuver and 25% haste.
Note: I omitted a crit of 99 that was done BEFORE the overdrive, since these procs wreak havoc on small sample sizes.
Checklist
Details
Overdrive seems to give the right categories of bonuses (maybe?), but the numbers appear to be off offensively/defensively. When comparing to retail videos from varying older years, the damage bonuses from being in overdrive and the damage reduction seem to not align with Eden.
I'll start by saying the benefits of overdrive are exceedingly vague everywhere I've looked, including the JPwiki, and the only things that seem concrete are:
1) Automaton gains 25% haste 2) You can't overload during it 3) Having 1 maneuver of an element up gives the benefit as if you had 3 maneuvers of that element up. 4) Automaton is extremely resistant to any status ailment (paralyze etc.)
There is a 5th attribute of overdrive but it is vague, all encompassing, and shrouded in mystery across different wiki pages.
5) The automaton gets "stronger" offensively and defensively.
ffxiclopedia ( https://ffxiclopedia.fandom.com/wiki/Overdrive ) says:
"Vastly increases the automaton's accuracy, attack, defense, evasion, attack speed, and possibly more."
BGwiki ( https://www.bg-wiki.com/ffxi/Overdrive ) says:
The Automaton parameters will be greatly increased: Damage dealt with melee hits and ranged attacks, as well as accuracy, ranged accuracy, damage reduction and evasion will significantly increase.
But I think most insightful might actually be the JPwiki ( http://wiki.ffo.jp/html/954.html ) which says:
_Increase the effect of some attachments . You need to be familiar with what is affected and what is not.
What is effective T. Increased attack power of Spring I-II . Increased magic attack power of loud speakers I-II . Increased avoidance rate for accelerators I and II . Increased accuracy of stabilizers I and II . Galvanizer counter rate . Increased magic defense of Mana Jammer I-II .
Ineffective Possibility to detect the magic defense of the scanner . T. Processor status judgment ability. Turbocharger haste + . Drum magazine shooting interval . The number of replicator illusions . The turbulent shooting effect of the barrage turbine . Capacitor overload prevention capability . HP up of auto repair kits I-II . MP up for Mana Tank I-II . MP recovery amount of mana converter . The amount of MP recovery of the economizer . Disruptor dispel effect .
Other attachments are unknown.
The effect varies greatly depending on the attachment and maneuver used. The main use is Turn on 3 wind maneuvers to make haste + 50%. Turn on 3 Thunder Maneuvers and counter with a galvanizer .
Etc. are common._
To me, this is what the other wikis are pointing at with the vague comments, but I have no idea how we would figure out the specifics. Note that the list has Drum magazine and Turbocharger in the ineffective category, yet they specifically call it out in the example below saying a main use of overdrive is to get the 50% haste from Turbo charger (25% from 1 wind that is treated as 3, and 25% flat from overdrive passive). They are highlighting that Turbocharger does not get FURTHER enhancements from Overdrive, just that it works normally with the 3 maneuver benefit. Note that they also call out Tension Spring 1 & 2 as EFFECTIVE. This to me says that not only do they get their 3 fire maneuver benefits, but they get boosted even further during overdrive. I believe that THIS is the extra damage that other wikis reference vaguely.
Lastly, just for completeness, here is the jpwiki from era ( http://wiki.ffo.jp/wiki.cgi?Command=HDetail&articleid=119171&id=954 ) where they haven't quite figured it out yet and say:
In addition, there is a view that the attack power, defense power, hit, and evasion may have increased internally, but the details are unknown.
Unfortunately there are no hard numbers to implement anywhere here but I think the most important fact so far is that patch notes never indicate that any adjustment to Overdrive. Here are all the patch notes that relate to Overdrive that I could find:
April 2006 when PUP was added to the game and they were referencing its 2hour.
The same patch as above, when they added Overdrive to the auto-translate
.... (truncated)...
May 14 2015 they increased the duration from 1 minutes to 3 minutes and changed the effect of a job point category from +3 to stats to +5 to stats.
Since it doesn't appear overdrive was ever reworked, I'm pretty sure we can just do retail testing to uncover all of the bonuses of the attachments during overdrive. I know I just made that sound simple, which it isn't, but it does sound obtainable.
Anyways, here are some of the comparisons that I've done between retail videos and Eden, although it is exceptionally hard to find old youtube videos where a PUP actually uses overdrive. Even in a lot of PTS/LB5 videos I watched I struggled to find one where someone uses their 2 hour.
The video below is from 2010 and shows a PUP doing an ENM run of Pulling the Strings. https://www.youtube.com/watch?v=Jmy6SEIpBTI
Just focusing on the defensive side of things, I recorded the damage that the Fantoccini dealt with his auto attacks to the players puppet before overdrive and during overdrive. I then ran the ENM twice on the Eden test server and recorded my results as well. The 2nd run was just to get more data for our automaton getting hit because I wasn't sure if valoredge blocking was coming into play an unlucky or lucky amount of times. Here is the data, with averages displayed the bottom row.
Note that I used 0 maneuvers when gathering my data. Here is my checkparam before overdrive
And here is my checkparam after overdrive
Unfortunately I am not positive which 2 rows of attachments I had equipped, but I did a hasty equip of basic attachments that I assumed the person in the video would have. I made a point to put on Armor Plate 1 + 2, Accelerator, Tension Spring 1 and 2, but I don't recall the other 3 I had on. I believe it was 2 thunder related attachments (not relevant to what we are testing here) and Flashbulb. I am happy to reproduce this test again if needed.
Observations: The average damage our automaton takes on Eden is notably less than in the retail video. The two lowest hits they take are 42 and 44, while on Eden the two lowest hits I took were 21 and 23. Possible explanations could be our block amount/rate, but I think what's important here is just that these averages were "reasonably" close for the purposes of comparing the overdrive numbers.
When we look at the overdrive averages, its night and day between Eden and retail. It is a very small sample size and could use more data, but although there was a defensive improvement on retail, it is nowhere near the defensive improvement that is gained on Eden.
I have not personally run LB5 on Eden, but this video is from 2013 and has a retail PUP doing the LB5 fight. https://www.youtube.com/watch?v=DWGB5ykbDiU
I did not parse this entire video, but some examples of damage taken during the overdrive ( timestamped at https://youtu.be/DWGB5ykbDiU?t=636 ) are 27, 39, 31 41 34, 28, 76 (crit), 34, and 33. Average is 38.11
Once their overdrive wears off we see that the new numbers are 39,48, 35, 43, 32, 34, 33, 38, 33, 33, 33, 33, 45, 36, 28, 29, etc. ---- Average is 35.75. These numbers are VERY similar.
I'm going to stop there for now and perhaps we should rename this report to be specific about the defensive capabilities of Overdrive, but the offensive ones are staggering, and possibly equally/more alarming. For quick reference, in the first PTS video his damage was roughly 50 b4 overdrive and high 50s maybe 60ish in overdrive. My damage went from around that to upwards of 170 on Eden.
I'm available to help as much as I can with this outside of retail testing, so please let me know how I can assist, I realize this is a huge undertaking and very convoluted.
Version: 30181205_0