# Wander distance of self-avoiding walk that backs out of culs-de-sac

Suppose a self-avoiding walk on $\mathbb{Z}^2$, with random steps each one unit long, backs out of culs-de-sac, but retaining the lattice points on which it stepped marked as unavailable for future steps. For example, in (a) below, a cul-de-sac is reached. Unwalking two steps allows the path to escape and walk in another direction in (b); but note those two points are marked. Eight steps further another cul-de-sac is reached (c). And so on.

The walk can continue forever, because once it unwinds to a point on the bounding box, it is free to step outside that box. My question is:

What is the growth rate of the distance of the path endpoint from the origin, with respect to the total length of (number of steps in) the path?

Perhaps this model has been studied? Here's a longer example:

(Added.) I can't resist one more, extending the above to 11,371 steps:

(Added 29Agu14.) Permit me to point to the new paper j.c. found and includes in a comment to his (knowledgeable) answer. That paper calls these walks SKSAWs: Smart Kinetic Self-Avoiding Walks.

• Culs-de-sac or cul-de-sacs? – Joseph O'Rourke Feb 27 '14 at 1:40
• quickanddirtytips.com/education/grammar/what-plural-cul-de-sac – Anthony Quas Feb 27 '14 at 1:42
• I don't know if you can say much about this model. In the case when you select a self avoiding walk of length n uniformly at random, the distance to the origin is not known (the best result is that it is sub-ballistic: arxiv.org/abs/1205.0401). – Bati Feb 27 '14 at 11:17
• Very pretty! I've seen pictures some-what like this in talks on SLE: en.wikipedia.org/wiki/Schramm%E2%80%93Loewner_evolution -- My (limited) understanding is that you are supposed to stare at the picture, think for a bit, and then hazard a guess about the Hausdorff dimension of the rescaling limit. – Sam Nead Feb 27 '14 at 21:30
• The model reminds me of loop-erased random walk, even though the definitions are quite different. – Eckhard Feb 28 '14 at 0:23

Coniglio et al quote results from percolation which imply that if $N$ is the number of steps taken which don't go into the cul-de-sacs, then the root mean squared end to end distance scales like $N^{4/7}$.
The critical percolation explorer is conjectured to have SLE${}_6$ as its scaling limit, and so that in some sense should be the scaling limit of your curves as well. This was famously proved by Smirnov for the hexagonal lattice and is still open for the square lattice.
• Wow! $N^{\frac{4}{7}}$! – Joseph O'Rourke Feb 28 '14 at 12:09