diff --git a/docs/guides.rst b/docs/guides.rst
index 999f9d4..27d740d 100644
--- a/docs/guides.rst
+++ b/docs/guides.rst
@@ -234,7 +234,7 @@ will be created in each output directory.
 Some useful initial checks on the output are:
 
 - Is the calculation equilibrated, or is the estimated free energy strongly dependent on the total simulation time? See the plots of free energy change against total simulation time. Often, the bound vanish stage shows the slowest equilibration
-- Are there large discrepancies between runs? The overall 95 % confidence interval for the free energy change is typically around 1 kcal / mol for an intermediate-sized ligand in a reasonably behaved system with 5 replicates. If the uncertainty is much larger, identify which leg and stage it originates from by checking the free energy changes for each, and inspect the potential of mean force and histograms of the gradients to get an idea of which lambda windows are problematic. Inspecting the trajectories for these lambda windows is often helpful. Checking for Gelman-Rubin $\hat{R}$ > 1.1 (indicative of substantial discrepancies between runs)(stage output directory) can also be informative.
+- Are there large discrepancies between runs? The overall 95 % confidence interval for the free energy change is typically around 1 kcal / mol for an intermediate-sized ligand in a reasonably behaved system with 5 replicates. If the uncertainty is much larger, identify which leg and stage it originates from by checking the free energy changes for each, and inspect the potential of mean force and histograms of the gradients to get an idea of which lambda windows are problematic. Inspecting the trajectories for these lambda windows is often helpful. Checking for Gelman-Rubin :math:`\hat{R} > 1.1` (indicative of substantial discrepancies between runs)(stage output directory) can also be informative.
 - Are the free energy changes for the bound restraining stage (where the receptor-ligand restraints are introduced) reasonable? As a result of the restraint selection algorithm, these changes should all be around 1.2 kcal/ mol. If they are not, check the plots of the Boresch degrees of freedom in the ensemble equilibration direcoties. Discontinous jumps can indicate a change in binding modules
 
 Running Sets of Calculations