Evidence that Macros Do NOT Cause GCD Delay or Lost Casts

[LilimoLimomo]

Your second video showing you using a macro for Blizzard isn't the evidence you thought it was. Just watching the video, I noticed you clipping your GCD several times. It's not a huge impact per se but it's absolutely noticeable if you're looking for it.

Thanks for pointing this out! This gives me an idea of how to make a macro that will at least make it clear to me if I clip my GCD; I'll repeat my performance with that, and hopefully I'll have a better video to upload soon!

Nevertheless, as others have said, it's more when you combined oGCDs with GCDs that macros become a much higher detriment. Put Dragoon's baseline rotation in a macro and you'll see how Life Surge or your Jumps either get misaligned or skipped entirely.

There are a variety of different ways I can imagine this working, so can you be more specific regarding what you're talking about? With a single weave, there are macros to make that work, but if you're talking about trying to macro double-weaves then that would be somewhere between tricky and impossible depending on the way it was implemented.

[LilimoLimomo]

I added an edit to the end of my original post to clarify a common misconception; posting it here for everyone's convenience:

Edit:
I've seen numerous replies talking about how the /wait command results in drift because of how it only operates in multiples of whole seconds. This is correct! If you want to queue up a whole 1-2-3 combo with a single button press, the result will be lost DPS compared to just pressing the buttons at the right time. This is why generally speaking, I would not recommend trying to use macros (and specifically /wait) for that purpose. However, that kind of shoddy macro isn't the main thrust of this post. Macros have so much more potential beyond the poorly-constructed macros you may be used to seeing; there are many ways to make good, helpful macros that do not result in drift. You can make macros that cast a skill when GCD is available, but cast an ability if GCD isn't available. You can make macros so that when you cast Fire III, your most comfortable buttons are populated by Fire spells instead of Ice spells, and vice versa. You can make macros so that alternating presses between only three buttons will perform every step of Dragoon's 1-2-3-4-5-6 combo, without any drift (add one more button and all of the AoE combos can fit in there, too, again, without drift). You can make macros that automate a single weave after a specific GCD skill. Macros can do many useful things that make gameplay more convenient for people who don't need it, and more accessible for people who do.

And that's why I did this research, did these tests, and made this post: because I would like to correct the misconception that macros innately lead to drift. However, what I didn't realize when I initially made this post was that many people seem to have only ever seen bad macros, the kind that lead to drift not because they are macros, but because they are trying to do things macros can't do well. To make an analogy, if you judge the capability of a boat based on whether you can fly it to the moon, you will probably feel like boats suck. But if you identify the strengths of boats and use them for things like water travel, you'll see that boats can be pretty useful The same is true for macros. So if you've only seen bad macros before, I hope that this post can let you know that there are plenty of other ways to use macros. I don't want to bloat this post any further, but if you ask me I'll be happy to give you some demonstrations of what those can look like.

[Cabalabob]

To make an analogy, if you judge the capability of a boat based on whether you can fly it to the moon, you will probably feel like boats suck. But if you identify the strengths of boats and use them for things like water travel, you'll see that boats can be pretty useful The same is true for macros. So if you've only seen bad macros before, I hope that this post can let you know that there are plenty of other ways to use macros. I don't want to bloat this post any further, but if you ask me I'll be happy to give you some demonstrations of what those can look like.

A more apt analogy is that it was already tested and the case was closed 10 years ago that manual cars can reach up to 100mph when driven properly while automatic cars can only go up to 80mph and also if you try to do any fancy tricks in them they’ll explode.

Now you’ve come going “actually, I just test drove both a manual and an automatic car for 10 mins each and I only got both up to 80mph, I think all these professional race car drivers that tested this are wrong and one is just as fast as the other”.

And people have come to tell you, no… it’s not… and also if you try to do any tricks in it, it’ll explode…

Can the automatic car be helpful for people who can’t focus on a gear stick and drive? Yes, no one has ever argued against that. But an automatic car will NEVER reach the same top speed as the manual car and if you use an automatic car you’ll never get better at using a manual car because you’re not getting any practice with it.

[TheRealMadruck]

snip

You are right. Macros do work. I can attest to it going back 10 years since beta! I still use macros on my combat jobs to this day from dungeons to 24 mans being on the top of aggro everytime! Back in 2016, I made this post discussing this very thing even including a no longer available video I posted with these facts from my own test: https://forum.square-enix.com/ffxiv/...er-Rotation%21. My methods have been refined since then to make the best combat macros I can make today from those I created in 2016.

Yet let me be frank with you Lilimo...no matter how much evidence you produce showing something I have known for 10 years regarding macros, you will always have a set of players on here saying you are wrong. They will act on cynicism instead of skepticism on any claim because they feel there is only one way to do things. These players are from a specific group who use a certain type of illegal third party addon in which I have known/spoken to directly from various FCs I have been in who make these specific points about macros.

Combat macros ensures you do not miss using OCD, macros eliminates human and latency errors 99% of the time for me. It takes practice like everything else in FF14 to perfect how to use the macros a player has created. But once you have built the muscle memory, I perform just as well in normal combat as any other player who does the "norm" with more consistant dps.

Bottomline, even if it is true that using combat macros "is a dps loss", the amount of dps loss you experience in 95% of FF14 content is minuscule! O no!!! I did only 990 damage per sec instead of 1k in my dungeon roulette... all macros bad. These players disregard their own latency vs anothers, they placate their skill is on par with others, and act like all have the ability or want to be the same as them. Use macros, dont use macros...at the end of the day no player should be able to dictate others on how to play a amazing game like FF14. Yet that is exactly what others are doing relegating this nuanced topic on combat macros as being a black/white topic when it factually isn't.

Thank you for your test Lilimo by confirming something I already knew back in 2016 in which the mass of players of FF14 have been misinformed about!

[Connor]

Why does it matter whether macros are good or bad? If a disabled person struggles to play without them isn’t it then better that they do use macros? Still seems like a net gain to me lol. I highly doubt anyone is turning up to savage/ultimates with their oGCDs macro’d to their basic attack

[Absurdity]

Why does it matter whether macros are good or bad? If a disabled person struggles to play without them isn’t it then better that they do use macros? Still seems like a net gain to me lol. I highly doubt anyone is turning up to savage/ultimates with their oGCDs macro’d to their basic attack

It matters because the initial premise wasn't that macros can work for people with disabilities, nobody is disputing that, do whatever you need to do to play the game comfortably.
The claim was that macros are not actually worse than just pressing the ability itself which has been proven false.

There are certain exceptions to this, Sprint for example has no skill-queueing so making a macro for it is actually an improvement as it reduces the likelyhood of your button press simply ghosting.

[LilimoLimomo]

The claim was that macros are not actually worse than just pressing the ability itself which has been proven false.

By chance have you had a chance to read my drastically re-worked initial post? I think if you read that you'll see that macro use isn't innately better or worse than just pressing the ability, but rather that there are advantages and disadvantages to each.

But one thing I wanted to make clear, and have provided evidence of, is that macros do not innately result in reduced casts/actions.

[FuzzyJCats]

Thank you for all your hard work Lilimo! It shows that you don't have to worry too much about missing dps if you need to macro due to disability reasons. It seems to only be an issue if you're doing ultimates and synced savages. Luckily the community is very understanding when I write in chat that I have problems, it's never been an issue in normal dungeons.

The macros I use are ones where you have to move your cursor and then press X for the ability (ex Earthly Star). This takes a long time on controller, as you have to move the target manually with analog stick, whereas, a macro will automatically put down the ability where you target. I use your technique of writing multiple lines for queue and it works flawlessly.

Ironically, using a macro this way actually shaves off seconds b/c it takes time to move cursor as a controller player.

That's such a wealth of information that you've given and will reread more in depth and watch your YT videos!

[Sindele]

Decided to take a deeper dive. Turns out it's a real pain to get millisecond precision from chat messages, but the result I got was kind of a no-brainer in retrospect: macros execute one line per frame. This was consistent across every test at every fps breakpoint tested.

As a quick refresher, frametimes at different common fps breakpoints, plus the resulting 'macro queue' window with a micon, error off, 13 cast macro:
144fps = 6.94ms frametime = ~90ms window after ~14ms
120fps = 8.33~ms frametime = ~108ms window after ~17ms
60fps = 16.67~ms frametime = ~217ms window after ~33ms
30fps = 33.33~ms frametime = ~433ms window after ~67ms
15fps = 66.67~ms frametime = ~867ms window after ~133ms(!)

In my opinion, this isn't a good thing for macros at either end of the curve, and shatters some of the initial claims outright.

To get an acceptable "macro queue" time, you need 60fps or lower - but worse fps also means that multi-purpose macros with things like oGCDs get delayed, potentially for extreme lengths of time that can lead to undesirable clipping. With better fps, your "macro queue" time narrows - at 144fps your window is less than 1/5th the length of the standard queue's half-second - which can lead to oGCDs not on waits being brought up so close to the cast that they will likely fail on animation locks. Worse, because fps in this game isn't exactly predictable or reliably controllable, it's going to screw with muscle memory, too.

One factor I'm not able to test for at present is how this interfaces with ping. I have my theories from what I know and what I've heard, but... I prefer data.

[LilimoLimomo]

Decided to take a deeper dive. Turns out it's a real pain to get millisecond precision from chat messages, but the result I got was kind of a no-brainer in retrospect: macros execute one line per frame. This was consistent across every test at every fps breakpoint tested.

As a quick refresher, frametimes at different common fps breakpoints, plus the resulting 'macro queue' window with a micon, error off, 13 cast macro:
144fps = 6.94ms frametime = ~90ms window after ~14ms
120fps = 8.33~ms frametime = ~108ms window after ~17ms
60fps = 16.67~ms frametime = ~217ms window after ~33ms
30fps = 33.33~ms frametime = ~433ms window after ~67ms
15fps = 66.67~ms frametime = ~867ms window after ~133ms(!)

In my opinion, this isn't a good thing for macros at either end of the curve, and shatters some of the initial claims outright.

To get an acceptable "macro queue" time, you need 60fps or lower - but worse fps also means that multi-purpose macros with things like oGCDs get delayed, potentially for extreme lengths of time that can lead to undesirable clipping. With better fps, your "macro queue" time narrows - at 144fps your window is less than 1/5th the length of the standard queue's half-second - which can lead to oGCDs not on waits being brought up so close to the cast that they will likely fail on animation locks. Worse, because fps in this game isn't exactly predictable or reliably controllable, it's going to screw with muscle memory, too.

One factor I'm not able to test for at present is how this interfaces with ping. I have my theories from what I know and what I've heard, but... I prefer data.

Thanks for taking the time to gather this great data! I've added that FPS affects the duration of the "macro queue" in my post, with credit to you.