Spell Queue Window Guide & Simulator

A small buffer can remove a repeated delay

Suppose your GCD lasts one second and you press a key every 250 ms. With the documented default of a 400 ms queue window, one of those presses can arrive during the final 40% of the GCD and wait for the exact finish. With the queue disabled, the presses made before the finish do nothing, so the first useful press may not arrive until a fraction of a second afterward.

That leftover time is the rotational gap. A single gap may look harmless, but the same delay repeated across a long encounter reduces the number of actions you can fit into the fight.

Your keypress rhythm matters as much as the number

A queue window only helps if one of your inputs reaches the server while it is open. That makes the spacing between presses important. A player pressing every 100 ms gets several chances to enter a 250 ms window. A player pressing every 300 ms can skip over a 100 ms window entirely, even with stable latency.

This is why copying another player's number is unreliable. Their network, hardware, class pace, and button rhythm may not resemble yours.

Wide and narrow windows solve different problems

A wider window favors continuity. You can commit the next action earlier and are less likely to leave the GCD empty. That is useful when the next choice is already clear or when your keypresses are relatively far apart.

A narrower window favors late decisions. The queue only ever holds your most recent press, so whichever action you choose last before the GCD ends is what fires regardless of window size. What a narrow window changes is how early that final press can land: an attempt made too soon arrives before the window opens and is dropped outright, so you are forced to commit closer to the deadline. That gives you a shorter runway to change your mind, but it does not make an early press "stick" — you can still overwrite it with a later one as long as it lands inside the window. The cost is a smaller target for your next press. If your cadence cannot hit it consistently, gaps appear.

The goal is not the smallest possible value. It is the smallest value that still gives your real input pattern enough room to produce reliable back-to-back casts.

A practical way to choose a setting

1. Start at the 400 ms default and play content where your rotation is familiar.
2. Use the simulator below with your normal GCD, world latency, and an honest estimate of how quickly you repeat a key.
3. Lower the window in modest steps, such as 50 ms, and test for several minutes rather than a few casts.
4. If casts begin to leave small empty spaces, move back to the last stable value.
5. Repeat the test during high-haste effects. A shorter GCD gives you less total time to place the next input.

Do not treat a popular formula such as “ping plus a fixed amount” as a universal rule. Latency is only one part of the timing chain, and the game's displayed round-trip latency is not a perfect measurement of when every input reaches processing.

What this simulator leaves out

The model deliberately reduces the system to GCD length, a queue deadline, estimated one-way travel, and repeated keypresses. The live game also has frame timing, variable network conditions, server processing, ability-specific rules, off-GCD actions, channels, cast times, macros, and cases where an action is not eligible when pressed.

Use the numbers to understand the relationship and compare settings. Validate the final choice by feel and combat logs in the content you actually play.