In 2012, we showed that the citation count for articles in ecology and evolutionary biology declines with increasing density of equations. Kollmer et al (2015 New J. Phys. 17 013036) claim this effect is an artefact of the manner in which we plotted the data. They also present citation data from Physical Review Letters and argue, based on graphs, that citation counts are unrelated to equation density. Here we show that both claims are misguided. We identified the effects in biology not by visual means, but using the most appropriate statistical analysis. Since Kollmer et al did not carry out any statistical analysis, they cannot draw reliable inferences about the citation patterns in physics. We show that when statistically analysed their data actually do provide evidence that in physics, as in biology, citation counts are lower for articles with a high density of equations. This indicates that a negative relationship between equation density and citations may extend across the breadth of the sciences, even those in which researchers are well accustomed to mathematical descriptions of natural phenomena. We restate our assessment that this is a genuine problem and discuss what we think should be done about it.