It is assumed that each instance would have some kind of documentation, however it is understood that there are cases for which no documentation is available either because it was never recorded or because the record was lost through history. This assumption is justified because these events occurred in modern history, in an area with a relatively free press and a relatively literate (if perhaps somewhat racist) population. We know from other instances of genocide that evidence abounds, even with the official attempts to suppress it, but also know from the archaeological record that there have been instances genocides for which we have no such solid written evidence.
This method does not account for Native Americans who died due to the presence or encroachment of Europeans but who were not killed directly, such as the estimated millions who died due to starvation from displacement or from disease brought by or exacerbated by the presence of the immigrated americans (which by definition are unlikely to have been recorded at all.). However, displacement by European settlers between 1492 and 1699 as a causal factor for any mass Native American deaths is unlikely - there simply weren't that many settlers to push out the larger tribes.
One might argue that these deaths were accidental, or due to the inferiorness of the native culture, medical knowledge, or immune systems. Historians are in disagreement on the total Native American population at the time of first Contact. A number between 20-30 million is considered most credible, though the highest estimates are around 80 million. By the end of the 16th century, credible estimates of the number of survivors is just over 1 million.
This method also does not include native americans killed during military engagements. Certainly Native American culture and European culture are not 100% congruent on the definition of military engagement, since for the Native Americans war was a much more personal affair than for Europeans (for whom it was more of a state-affair), however it is possible to broadly generalize that in situations where both sides were armed and where the majority of deaths were of people who entered the engagement with the knowledge that they might die but who fought anyway, that this is not a "massacre". We do not count such cases. History judges deaths of soldiers in warfare to be an outcome of larger matters.
This system also does not intend to show that the deaths were one-sided or unjustified. In fact, the documented record includes many instances of deaths of Europeans (innocents as well as provocatures) at the hands of the native americans. Also, while the Europeans did much to exacerbate or incite animosity of the Native Americans, the reverse is likewise true. The native populations frequently warred among themselves, and held life and death in a significantly different context than the europeans did. This method merely intends to demonstrate a pattern of direct, state-sponsored mass killing of native populations during peace time, directly associated with expansion into native lands.