How swiping works

If you’ve followed our instructional material, you know that swiping involves playing through a string while alternate picking. It can be a mistake, but when used systematically it’s also a technique that many players develop subconsciously to simplify string changes — especially if they use single escape picking motions like USX and DSX.

However, when they do this, they’re not just using a single escape picking motion and hitting the string that is in the way. Instead, there is a technique to doing this well. It involves making, essentially, a double escape picking motion. Here’s what it looks like:

When I demonstrate swiping slowly at the start of the clip, you can see how the pick flops over the swipe string. I have dwps here so the pick slides, but the contact is still pretty obvious.

However when I do this fast, the motion changes. You can see clearly that the pick is making a DBX motion. But because it’s on a tilt, there is still some incidental swipe string contact. It’s very slight, due to the fact that DBX is a semicircular motion. On some repetitions, it almost appears to not swipe, even though, ironically, I’m trying to swipe.

When this is done correctly, it feels and sounds very smooth, like there is no string in the way. In slow motion we can see why that is — because the pick is doing its level best (pun intended!) to avoid the string, and only contacting gracefully when it does contact.

Also ironically, I think I have always done this, I just didn’t know it. We introduced the concept of swiping in 2014 with the Antigravity seminar. This where we profiled Paul Gilbert’s more accidental use of it, and Michael Angelo Batio’s more systematic use.

And we have numerous test examples of it in the clips. Re-watching these now, you can see that the picking motion during the swipe pickstroke is different, approximating a DBX pickstroke to lessen the swipe string contact:

The upstrokes escaping are super obvious here. I think many of these pickstrokes actually do escape. In fact, the only really obvious swipe in this lick is actually a downstroke, even though this is an upstroke example. Obviously, I was not aware I was doing this when we filmed this, and didn’t catch it during edting either.


The “best” swiping isn’t just a single escape motion moving through the string with the same trajectory it always uses. There is a different joint motion the alter the trajectory to make less forceful contact. You are trying to escape, just strategically not succeeding.

I didn’t learn to do this by trying to simulate the motion slowly and doing “repetitions”. I didn’t even know I was doing these motions so that wouldn’t have been possible anyway. Even now, I can’t do it slowly even though I am more aware that I’m doing it. When I do ti slowly it just looks like what you see in the beginning of the Instagram clip.

In general, the more complex a technique is, the less likely you will be able to think through or simulate it correctly while going really slowly and repeating it many times. I think this is true of any motor skill, not just instrument technique.

Instead, success will most likely happen when going at some “real world” speed while controlling the bigger picture things you can control: overall form, feeling of the motion, amount of force applied, when you start and stop, etc. I think the idea that you can control smaller pickstroke-level details while playing at even medium speeds is an illusion and not what is really happening or how motor learning actually works.

The advantage we have now is feedback like you’re reading here. This gives you a clear idea how to do something, what form is involved, what motions are involved, and what something is supposed to look and sound like when done correctly. This way, you’ll be more likely to recognize when the end result is correct.


Great stuff! I am wondering if, when playing with that dwps setup as you are in the video, the opposite holds true as well. So, if you started that pattern on an upstroke, would(/could) the pickstroke at the beginning of the repeat be DBX of sorts as well?

In Antigravity I do it with 2wps and no swiping. This is descending inside sixes but the picking motions are identical to the ascending version you’re describing. You can turn off the sound and it would look the same:

There is no swiping here and it wouldn’t be common for there to be swiping here. Swiping can sometimes happen during inside picking but players learn to avoid it because it generally sounds more obvious and feels worse.

As to whether or not there is “DBX”, notice the third pickstroke of this lick uses the same wrist motions as a DBX pickstroke. It makes a semircular motion. It’s just tilted so it doesn’t look like it’s escaping at both ends, even though the joint motion is the same. This is why there is more forearm involved in a 2wps player’s motion.

Ultimately, 2wps doesn’t absolve you from the need to make occasional semicircular picking motions with the wrist. It’s plainly visible how common these curved motions are in everyday playing, even among players who may complain they “can’t do DBX”.


Super cool. I have been enjoying CtC since you first released those episodes about ten years back on Youtube. Really gave me a new look on how to approach playing the guitar. My takeaway from your response is that you do need to change the setup via forearm rotation in order to hit the strings that deviate from the single escape order. I have tried to plow through it whilst in dwps (so after a downstroke and needing to change to a string above and then hitting the note you just played but muting on the way up) with elbow motion as an alternative. It works, but feels quite forcefull.

Great post Troy! When it comes to the systematic use of swiping like in MAB’s case, would you say he is also trying to escape the strings and not succeding or in his systematic use, it’s deliberate to not try to escape and essentially turn it into a single escape/single primary motion picking?

To my understanding his secondary motion for escaping the strings is only used ascending with a USX string change and then avoiding it completly when descending.

Is swiping faster than DBX? I would guess that it is. Also, I’ve seen Steve Lukather do swiping as well, but I can’t remember the exact video. It was a 4-note pentatonic cliche, swiping through the g string:

g: 14(d)----------------------

Repeat the above as many times as you want. Sounds really cool and aggressive compared to the legato + sweep version (d-d-h-p OR d-d-u-p). You can also do that lick swybrid as well (u-pluck-d-u). Many choices!

What are you referring to, how to play the ascending or descending inside sixes lick? This depends on what your technique is. Are you a wrist player or some other joint? Which form and which escape? Film and determine as per the Primer.

Just as one example, for a wrist player who is USX, has an obvious upstroke escape when playing a fast tremolo, and has a downward pickslant — and only if these criteria are met — you can use the “primary down 2wps” approach as described in Antigravity.

If you are a DSX player with a zero-degree pickslant (this is important), you can use the “primary up 2wps” approach described in Antigravity.

However, since this form is essentially indistinguishable from the DBX form we teach as our default for wrist motion now:

…then you can also just play the line with DBX and do nothing super obvious with the forearm. However I recommend starting with outside sixes and making sure you see the sequence of motions shown here. This is what correct for DBX looks like:

Other joints / forms / etc. other instructions. But this is just to give you an idea of what is possible.


Swiping is DBX, just tilted so that it hits the string. At least the way I’m showing you in the Instagram clip. That’s the point of this post — sorry for burying the lede if it was not explained clearly! Good swiping makes a DBX pickstroke and tries to get over the string, even if it ultimately doesn’t. I can’t feel this but the slow motion video shows clearly that this is what is happening.

Yes Steve is a USX player and probably does something like what I am showing in the Instagram clip, where there is some attempt to get over the string that may or may not make it, but succeeds in lessening the audibility of unwanted string contact if it happens. Displacement is also a super common thing and lots of USX / DSX players do it without knowing. You would need to film him to know which of these things are happening.

Players that get the most clarity are players who tend to stick with lines that work for USX and don’t stray from that. This is one of the reasons Yngwie and Eric made such impacts. They somehow knew, these are the lines that work, I’ll play these and avoid the others. The difference between their results and even very good contemporaneous players was dramatic.


Oh, I think you linked the wrong video; it shows the USX version. I did see your other video with the DBX one though. I used to think of DBX as 50% DSX and 50% USX for ascending sixes (DSX D-U-D, rotate, USX U-D-U). Probably why I never got anywhere with it, since it is really 5 notes DSX with one note done by the helper motion.

That USX one was interesting; was the fast clean and distorted take around the same speed? Correct me if I’m wrong, but you said you swiped the clean take and did displacement for the one with gain. Do you find displacement happens at random, or is at a certain speed?

Yea, displacement is another possibility for that lick I didn’t think of. I guess the picking would be d-u-d-u on the b string, with the last upstroke occuring at the same time as a hammer-on on the g string. You could get that going faster than the swiped version since all of the picking is happening on the same string.

Yep, although USX swiping is really nice since you can’t hear it compared to the DSX version.

Aren’t you doing forearm rotation there during the “DBX moment”? Just look at the palmaris longus (or is it the flexor carpi radialis? You know, the obviously sticking out tendon) in the slow-mo part of the video at the end.

No, Mike just goes through the string. It sounds pretty good considering but is maybe somewhat more audible at times than what I’m doing in the Instagram clip.

The escape possibilities are determined by which joints are available. I don’t know what Mike’s motion really is — it’s complicated and appears to change. But the escape itself looks like an elbow escape, which is to say, DSX and relatively flat.

1 Like

Sorry again for the confusion! I meant to link to the USX video, so at least I got that part right.

What the USX video shows is that when I play ascending sixes with swiping, I am NOT actually doing USX joint motion all the time. Instead, at the moment the swipe happens, the wrist makes a semicircular pickstroke instead of the usual straight line motion. This semicircular pickstroke is what allows the pick to almost get over the string. So even if there is a swipe, it is less aggressive because of this.

It just so happens that this type of semicircular pickstroke is * the same joint motion * used for DBX playing. The overall form is simply tilted so the pickstroke doesn’t always escape. So it doesn’t actually “doubly escape”, but it would if the form were different. In a perfect world we would have a term for “semicircular pickstroke done by the wrist”, but we don’t. We just call that “DBX” because I didn’t forsee that we would get to this point. But here we are.


A player doing what I am doing in the USX Instagram clip is doing virtually the same thing as a player who does what I am doing in the DBX Instagram clip. Mainly the overall form is different. But the joint motions are very similar.

So, when you ask which one is faster, my answer is, they are the same because the motions are (nearly) the same. Let me know if I’ve explained it this time!

Not materially. The difference in escape trajectory between the upstroke and downstroke escape in that clip is about 30 degrees. In no way is 30 degrees of forearm rotation happening — barely any is happening. The change in escape is because the wrist is changing its direction of motion.

Again, if you’re trying to learn to play phrases, and you want to get actual results, try not to get lost in these tiny details because it’s easy to miss the bigger picture. Assume the overall form, do the best approximation you can of the motion at a normal speed, and shoot for what feels easiest and sounds good. Then film as a test and see if it looks like it’s supposed to look.

Yes, thanks for the clarification! I can see the different motion now, but it is very subtle. I definitely wouldn’t have picked up on it just watching, even with the slo-mo footage. I wrongly assumed USX swiping didn’t have that motion and it was only DBX that did.

Do any USX players do swiping without a helper motion? Seems like it would be possible if you just plow through the strings. Or maybe that semicircular motion lessens the amount of resistance you have when you swipe?

Yes, the DBX motion makes it less swipey, I think that’s why I do it. I wouldn’t call this a “helper motion” per se, it’s just the picking motion that the hand has decided to do at that point because it works. Many wrist styles involve mixing and matching linear and semicircular motions — this is common.

Just keep in mind this stuff is all learned by feel and I had no idea was I doing this until I filmed it. I can only assume that “plowing” as you say sounded and felt too aggressive so the hands figured out another way, maybe even accidentally at first, but kept doing it because the results were good.

This is why it’s important to operate with a top-down approach, mainly paying attention the bigger picture aspects like form, force, speed, etc. Now that we have good information on what the overall form should look like for various techniques, that information should get you close enough to minimize the trial and error that happens as you actually try to perform the techniques.


@Troy I think this is what I have referred to as “skating” on this site before. You’re not really plowing through the strings more than just gliding.

Exactly, lots of people probably do things liks this and don’t realize it. This just provides more clarity on what the actual motion is that is happening, so players have a reference for determining if/when they’re doing it right. Knowing you should be seeing a specific sequence of motions when you film yourself is at least something actionable.