Mineral resource estimation is commonly based on the accumulation of a series of local estimates. At times, the distribution of these grade estimates is not appropriate to the grade distribution of the data on which the estimates are based. Mineral resource grade estimation has an increased potential of being in error as the grade variability increases. As the mineral commodity value increases, greater grade variability within the mineral resource is tolerated. This paper directs its attention to mineral properties where the data average grade is 20 per cent or more than the data median grade at a specific cut-off grade. A mineral resource estimate is enhanced if it is supported by graphs which demonstrate that the sample data have been separated into appropriate geological domains. Each geological domain has a unique grade distribution signature, which can be graphically demonstrated simply and objectively. This signature will usually be evident well before the Indicated mineral resource estimates are available. The paper looks at the use of the log probability and MedAve plots to assess how effective domaining has been. It further outlines how the MedAve plot can be used to assess the integrity of the mineral resource grade estimate within a geological domain with the data on which it is based. The paper suggests that the overall grade of a mineral resource at any cut-off grade will be closer to the data median grade rather than the data average grade. Tips are provided to ensure that the odds reflected in the data grade distribution are not ignored and hence the risk that the project is affected by unrealistic estimates of the overall mineral resource grade is reduced.