Hey guys, so I have been messing around with competitive Melee, and Luigi, for around 8 months, but I cannot seem to consistently wavedash. While I am wavedashing around with Luigi I will randomly full hop, I practice and try to adjust my timing, but I always seem to manage to randomly full hop at the most inconveniencing times. Is there any advice, tips, or comforting words that any of you fine lads can share with me that will help me overcome this problem? It is getting very frustrating and is really holding me back in matches. Thanks
If you're jumping instead of wavedashing, then you're either inputting the airdodge too early (so the input gets eaten by the jumpsquat and doesn't take effect), or you're not inputting the airdodge at all.
Bear with this long section (the tl;dr is "do your airdodge input slightly later so you're more likely to be too late on your wavedash instead of too early without changing how often you time the wavedash correctly"):
Wavedashing without leaving the ground is actually a frame-perfect input, so if you're even one frame too early then you'll jump, and if you're at least one frame too late then you'll enter the air (you'll still get a wavedash in the first few airborne frames; it'll just be shorter and slower). You
will mess up frame-perfect inputs occasionally unless you are literally absolutely perfect (which is humanly impossible). This is because timings are based on frames (which are discrete) and not real time (which is continuous). Therefore, inputs the same distance apart in real time can have different results when made at different points relative to the frame time. The timing of each frame changing to the next affects whether the input is correct, but since you can't possibly know what that timing is, it's random for practical purposes.
A 60Hz frame is 16.7ms long (to one decimal place), and the inputs for wavedashing with Luigi are 4 frames apart. When you wavedash, you could do the airdodge input 67.7ms (only 1ms more than exactly 4 frames' worth of time) after the jump, and if it happened to be that the frame changed on that 1ms then the input would be considered a whole frame off. If the frame didn't change then the input would be considered to be perfectly timed.
Since being too late is much less bad than being too early (you still wavedash, rather than jumping), you should bias your inputs slightly to the slower side. That way you'll still get the perfect timing with the same frequency (almost all the time), but when you're early you'll be more likely to still end up doing a perfect wavedash anyway because you have a larger buffer of early timings that are still frame-perfect. It will also be the case that more of your late inputs are going to be too late rather than falling into the perfect timing zone, but this is worthwhile to not accidentally jump as often. In other words, the relative number of frame-perfect inputs to too early/too late inputs remains identical, but the number of too early inputs (the worst-case scenario) reduces.
This is all assuming that the variance in input timing is the same in both directions; i.e. that you'll be too slow to the same degree and just as often as you'll be too fast.
Here's a simplified visual representation (the vertical bar is the where you're aiming, braces show variance in timing; "too early" or "too late" means an input made there will be considered at least a frame off from frame-perfect; all inputs made in the area in the middle are frame-perfect):
Instead of doing your input here:
-------------------{-----------------|-----------------}-------------------
[.......too early.......][.......................][.......too late........]
You do it here:
-----------------------{-----------------|-----------------}---------------
[.......too early.......][.......................][.......too late........]
Your frequency of doing the frame-perfect input remains the same with both timings, but with the second timing you're more likely to be too late than too early. The smaller your variance the more likely you are to be perfect (and the less likely you are to be too early/too late).
Practicing wavedashing more will still help you, even without changing the timing you're aiming for, since with practice you can still reduce your variance.
Also, being nervous will make your techskill worse, so that may be a factor.
