Markov type 
of a metric space, for 
, is a bound on the expected growth of distances under
a stationary reversible finite-state Markov
chain. The space 
has Markov type 
with constant 
when
 |
for every such chain, every map 
from its state set to 
, and every positive integer 
. The constant
is uniform over the chains, maps, and times. A Banach
space is given the metric induced by its norm. Nontrivial
Markov type means that this holds for some 
.
