Are there timing effects in coordination game experiments?

The timing effects (timing without observability) identified by Weber, Camerer, and Knez (2004) in coordination game experiments are caused by their fixed-matching protocol. When we use a random-matching protocol the alleged timing effects completely vanish.