## Idle question: are Kakeya sets winning?

Jayadev Athreya was here last week and reminded me about this notion of “winning sets,” which I learned about from Howie Masur — originally, one of the many contributions of Wolfgang Schmidt.

Here’s a paper by Curt McMullen introducing a somewhat stronger notion, “absolute winning.”

Anyway:  a winning set (or an absolute winning set) in R^n is “big” in some sense.  In particular, it has to have full Hausdorff dimension, but it doesn’t have to have positive measure.

Kakeya sets (subsets of R^n containing a unit line segment in every direction) can have measure zero, by the Besicovitch construction, and are conjectured (when n=2, known) to have Hausdorff dimension n.  So should we expect these sets to be winning?  Are Besicovitch sets winning?

I have no reason to need to know.  I just think these refined classifications of sets which are measure 0 yet still “large” are very interesting.  And for all I know, maybe there are sets where the easiest way to prove they have full Hausdorff dimension is to prove they’re winning!

## Idle questions: Diophantine approximation and complex dynamics

Laura DeMarco gave a beautiful talk at the Joint Meetings about her work (with Matt Baker) on the postcritically finite locus in the moduli space of polynomial dynamical systems.  (Here are her slides for a similar talk.)

To say only a tiny bit about what that means:  The dynamical systems in question are those coming from a polynomial map f: C -> C.  Like, say,

$f(z) = z^2 + c$

for some complex number c.  The set of c such that the forward orbit of 0 stays bounded is the Mandelbrot set — you know, this guy:

One way an orbit to be bounded is for it to be eventually periodic; when z^2 + c has this property, we say it is postcritically finite, or PCF.  More generally, the postcritically finite polynomials are those whose critical points all have finite forward orbits.  Number theorists like these because they’re the ones whose inverse iterates generate big interesting number fields with finite ramification.  But that’s not what I want to mention now.  DeMarco mentioned the very interesting fact (sorry, I don’t know who proved this or whether I’m stating it correctly) that as you range over PCFs with longer and longer period, the set of PCFs approaches the uniform distribution (with respect to a standard measure called bifurcation measure) on the boundary of M.

The PCFs, DeMarco told us, should be thought of as special points in the space of all polynomials — in this simple case of quadratics, the PCFs are special points in the complex plane.  They’re kind of like CM points on the j-line, or torsion points on an abelian variety.  The main thrust of DeMarco’s work with Baker concerns dynamical analogues of the Andre-Oort conjecture, which aims to classify those subvarieties of the moduli space of dynamical systems which contain a (Zariski-)dense set of PCF points.  Their striking results demonstrate the unexpected ways in which arithmetic dynamics and complex dynamics have truly started to engage with each other, after a fairly long period of separate development.

But that’s also not what I want to mention now; I just wanted to record a simple thought that a number theorist might have while watching DeMarco’s talk.  (Warning:  as usual with math posts, this is not thought through carefully.)

The PCF points are perhaps sort of like torsion points in C^*, which is to say roots of unity; and just as PCFs of larger and larger period converge to uniform distribution on the Julia set, roots of unity of larger and larger order converge to uniform distribution on the unit circle.  Equivalently: rational numbers of bounded denominator look roughly uniformly distributed on R/Z.

But there are lots of more refined questions one can ask about the way in which the rational numbers sit densely in R/Z.  For example, one can ask about Diophantine approximation; given an irrational point alpha on R/Z, we know there are infinitely many “pretty good” rational approximants to alpha; fractions p/q such that

$|p/q - \alpha| < 1/q^2$.

Are there theorems guaranteeing that any point x on the boundary of the Mandelbrot set has infinitely many PCFs which are “pretty good approximations” to x in the above sense?

What is the most badly approximable point on the Mandelbrot boundary — i.e what is the “golden dynamical system” that plays the role of (1/2)(1+sqrt(5))?

Does x have a canonical sequence of PCF approximants which play the role of continued fraction convergents?