In the paper "A Diagrammatic Approach to the Categorification of Quantum Groups I" (arXiv, journal, MSN), Khovanov and Lauda put a strand with a single dot in degree 2, and put the crossing operator in degree 1, and set these to be the generators of a graded ring.
However, when they define an action of their (graded) diagram ring on a given (graded) polynomial ring, applying the crossing operator a homogeneous polynomial with integral coefficients in this graded ring yields a polynomial in odd degree, which shouldn't be possible (as attaching dots to strands represents raising the degree by 1, and thus all polynomials sit in even degree). Is there something I'm missing here?