Double ropes … why the twist (rgold)?

Using double ropes (Mammut Genesis) for multi-pitch rock (target use is alpine >10pitch). Standard mode is to swap leads, using guide mode w/ plaquette style device. Getting twists in the system as we get further into routes. What are the causes of this (e.g. creating a twist when setting guide mode belay, where/how second comes up to belay, then shoves off to lead next pitch)? Seems like this would be straight forward but continue to create the twist, clearly missing something simple here. Rgold or someone that uses doubles would appreciate insight into the secret here (!).

Thinking about it I always assumed it was turning around at the belays or the second turning around while putting on shoes on the ground etc. But it actually will also happen even if you both keep facing the rock the entire time depending on which side of the leader you swap leads. i.e. if your second comes up to your belay on your right hand side and continues, you've got a twist. If you pass them on the left on the next pitch it will un-do it but if you pass on the right again now you've got two.

How you put your ropes into the belay device won't cause problems, if you make a twist that way it will just undo itself when you remove them, you haven't actually passed the ropes around each other.

Easy enough to un-do if it gets clustered, just go into the anchor, untie one of the ropes and pass it around to undo the twists, re-tie.

Edit: I'm talking here about doubles twisting around each other, not twists in the individual ropes. I assumed that's what you meant

Following

I think your rope is always twisted to some extent. And when you tie in to both sides you lock it off. Then as your belaying each other your just ironing the twists to one end and it becomes more pronounced.

But if you're talking about the ropes twisting around each other. There's some voodoo magic I've never quite figured out 100%. Like even climbing on singles, what is the mantra stack the rope in long loops -> shorter loops for swapping leaders, and short loops increasing size for same leader?? idk magic?

I think the easiest way to understand what's going on is to simulate the the system in action:

Take two pieces of string or cord (different colors would help, or mark one of them). Take two pens or pencils and put them on the floor or a table (to simulate the rock face they will be climbing). Place the perpendicular to the direction of travel and tie one string to the left side of each and the other string to the right side of each.

Don't rotate or turn the pens as you do this next bit. Now mimic the action of the lower pen joining the right pen at the belay on the right side and then passing to the next belay. Now have the following pen join the leading pen at the belay on the right side, and passing. Repeat this a few times and you'll see twists develop.

Now untangle your mess and repeat the process but have one pen always pass on the right and the other always pass on the left. I don't think you have any twist this way.

Essentially, if you both pass on the same side, one of the ropes crosses the other one each and every time you pass.

Edited: I'm pretty sure (providing the belay allows it) that if you do a 360 degree turn towards the direction your partner passed on, you undo the twist. They pass on the right, turn clockwise. They pass on the left, turn counter-clockwise.

Are you speaking about individual strands becoming progressively more kinked or about the two strands wrapping around each other?

... the later, the 'two strands progressively wrapping around one another'. Is most notable as the number of belays accumulate.

It's well-known that if you don't keep an eye on them at all times then double ropes inevitably sneak off into the fourth dimension and twist around themselves.

More seriously, on multi-pitch climbs you often alternate between facing in (when climbing or belaying the leader) and facing out (when belaying the second). Depending on your turn directions your net rotation could be 0 degrees, 360 degrees, or some multiple of 360 after multiple belays. If two climbers don't mimic each other's turns exactly then the result will be twisted ropes.

And then there are weird chimney pitches like the crux on Syke's Sickle where you could rotate through 360 just by climbing the pitch.

Yup, I'd go with Martin's answer. Eliminating the results of 4-dimensional motions, either the leader or second or both have to induce rotations in complementary ways in order to put twists in. It turns out this is easy to do, and the rotations may not be obvious. Wide cracks that the climbers turn around in can create twists, but I'd say the vast majority happen at belay changeovers. It is also possible, when changing from belaying the second to belaying the leader, to turn the plaquette over in a way that creates a twist, and the belayer can initiate a twist by threading the ropes for the leader's belay in the wrong order.

Almost all of the time, when the ropes are taken up, the second will see they are twisted and can undo the twists by an appropriate rotation. Taking one strand and dropping it over the head (perhaps repeatedly) is a worse approach, because this will put kinks the the strand that is being moved and these might eventually get bad enough to cause their own problems.

As for kinks, as a way to keep ropes as kink-free as possible, we don't have the second tie in on the first pitch if is is feasible to be unanchored. The leader takes up the ropes without them being anchored at their ends and so at least some of the kinking can dissipate. It makes good sense to knot the rope ends to prevent them running through the belay plate on long pitches, and once the second is off belay, they can clip each strand through a carabiner on their harness. This allows the ropes to de-kink as they are taken up but still keeps the second in control of the ends, which will lodge in the carabiners.

The main downside of this approach is the second doesn't get a partner check when tying in. I don't care about this at all, but many people might and so would not use this method. Perhaps I should add that I would never leave a novice at the base to tie in by themselves; this is a technique for practiced climbers only.

Thank you for all the feedback - will test it out this weekend.