While obviously working on dinosaurs means that you are primarily looking at fossil bones, these are not always the whole story when it comes to the skeleton. Obviously bits go missing, and whole skeletons are a rarity, unossified parts can also make themselves scarce. However, not all parts of the skeleton always even ossify – the skeleton is not synonymous with bones. Chickens have a mostly cartilaginous sternum for example.
So while a fossil might preserve all of the bones, it won’t necessarily preserve all of the skeleton. Add to the fact that as noted with ossified tendons, not all things that can ossify always do (and of course with tendons, most of them do not) making things harder. Coupled to this is the fact that, as with ossified tendons, there is likely to variation if how such a feature is expressed. You might have half a dozen good specimens with only one preserving a given feature. That might not be missing in the others, it simply ossified in one of them and not the rest. Part of this variation also comes from age, with typically older animals having more bones or parts of bones that are ossified than juveniles (or even other adults if it’s a particularly old individual) so this has to be factored in too.
This is an especially important consideration why trying to track evolutionary changes in groups where certain characters or bones seem to come and go regularly. An example are the clavicles / furcula that I touched on recently. If you match a phylogeny to the specimens we have with furculae / clavicles the picture is a little confusing with the elements apparently disappearing and reappearing multiple times. Part of this will be down to the incompleteness of some fossils, but it’s also likely to be in part due to how these bones ossify. Based on their inconsistent appearance and incomplete preservation in at least some taxa, it’s likely that they were largely present in most, if not all, theropods and merely remained as cartilaginous elements and did not ossify, hence their apparent (but not genuine) absence.
Characters can of course be lost, and reappear (we’ll be dealing with this issue next up) but caution should be taken in assuming that this is the norm. Obviously knowledge of the vertebrate skeleton really helps as some elements are more prone to this pattern of only occasionally being ossified, or ossification patterns changing during ontogeny etc. but the first assumption should not be that an element is genuinely absent, even when there is an apparently complete set of bones present. There’s more to the skeleton than just those lumps of calcium and phosphate based crystals.