We thank Dr Tan for his comments regarding our study. Statistics are used to test the validity of aims and hypotheses of a scientific study. Therefore, we used statistics to assess factors affecting the survival of bone grafts and these gave us a number of interesting and unexpected results.1 We agree the significance of diabetes was a surprise with the numbers involved, however it is consistent as a risk factor for bone graft failure, both in the clinical and scientific setting.2 The p value for this was 0.006 which is very significant. Diabetes is, however, not a contraindication for implant placement although significant complications related to risk of infection, can occur with placement of implants in diabetics with poor glycaemic control.3, 4 The other interesting finding the statistics revealed from our present study was that smoking was not a significant risk factor for graft failure. This was unexpected and counter-intuitive, particularly in relation to the vascularity issues required for revascularization related to smoking5 and the number of failures found in the smoking group. Some of these peculiarities with the statistical findings may be related to the modest numbers used in the study. We do, though, plan to revisit this topic when numbers are increased. This study was in effect, a preliminary audit of our results to date within the setting of an oral and maxillofacial surgical training programme. With statistics, you win some and you lose some, but they are there to serve a purpose. Sometimes you get some unexpected results, but these aren’t lies, they are just unexpected. If these results are relevant to the aims of the study, then they should be included.