The official Ottoman statistics compiled for the period between 1915 to 1917-18 were of 800,000 killed. This figure originates from Djemal's bureau. The results were published in the official Ottoman gazette.
It was allegedly the result of a commission formed by the interior minister Mustafa Arif. It is said that they relied on reports and statistics they had compiled in a period of two months; in March 14, 1919, the results were made public by Djemal. This same figure was mentioned in Rauf Orbay's own memoirs. The initial results apparently represented those who were “massacred” during the deportation, without any indication as to the total number of people who perished. Mustafa Kemal, during a conversation with Major General Harbord, the chief of the American Military Mission to Armenia, in September 1919, repeated the same number. The figure of 800,000, it should be noted, excludes Armenian soldiers in the Ottoman army liquidated in the early stages of the genocide, as well as the number of women and male and female children assimilated into Turkish families.
However, following the dissolution of the military tribunal, those figures were reinterpreted. The Turkish author Taner Akçam refers to a Turkish military estimate published by Lt. Col. Nihat in 1928, in which the figure of 800,000 no longer represented those "massacred" or "killed", but simply those who perished. Then the historian Bayur in a famous work wrote: "800,000 Armenians and 200,000 Greeks died as a result of deportations or died in labor brigades." Bayur concluded: "According to our official sources, these numbers are correct.
While the official figures were of 800,000 killed, there were many unofficial numbers presented during the war by some Ottoman authorities--Talat, for instance, presented the figure of 300,000-- but there is no indication as to how those figures were obtained. This figure is currently the one used often by the Turkish government officials.
McCarthy calculated an estimate of the pre-war Armenian population, then subtracted his estimate of survivors, arriving at a figure of a little less than 600,000 for Armenian casualties for the period 1914 to 1922. But as in the cases of his population, his statistics are controversial. In a more recent essay, he projected that if the Armenian records of 1913 were accurate, 250,000 more deaths should be added, for a total of 850,000. And he is also criticized for overestimating the survivor table. Frédéric Paulin goes as far as comparing his methodology with Rassinier's method in calculating the European Jewry losses during World War II.
German sources gave the highest estimates of Armenian losses during the war even though they were the Ottoman Empire's ally. Some speculate that it was due to their access to murder sites.
A report stated that as early as February 1916, 1.5 million Armenians had been destroyed. A report in May 27, 1916, by Foreign Office Intelligence Director Erzberger provided the same figure, as did an October 4, 1916 report by the German Interim Ambassador to Turkey, Radowitz. It seems that the generally cited 1.5 million figure originated from those German sources. German major Endres, who served in the Turkish army, estimated the number of Armenian deaths as 1.2 million. The same figure was mentioned during the Yozgat trial, and before the Permanent Peoples Tribunal and is often cited elsewhere.
The Austrian consul at Trabzon and Samsun, Dr. Kwatkiowski on March 13, 1918 reported to Vienna, restricting himself to the six eastern provinces, Trabzon and Samsun district, that of the million deported, most died, while Austria-Hungary's Adrianople (Edirne) consul Dr. Nadamlenzki reported that for the entire Ottoman Empire 1.5 million had already been deported. The Austrian Vice Marashal Pomiankowski estimated the Armenian losses at about a million.
While the Ottoman official statistics covered 1917-18, and some of German figures, most other figures excluded them. Another problem remains, as to the availability of the sources for what followed 1917. More recent scholars have called this period the second phase of the Armenian Genocide. Melson, for instance, provide' a rough estimate of 500,000. On the other hand, those estimates have no archival grounds, for this reason some researchers considers any such figures could be near to the actual casualty figures or far from it.
Few commissions were formed though, such as the investigations for Kars and Alexandropol. The Alexandropol investigation by its nature is seen as the most serious such endeavor. It presented 60,000 as directed killed, in a total of 150,000 victims which condition would have ultimately led to their death sentences. But the investigation apparently came to an end abruptly. The Germans on the other hand, not presenting any numbers, have reported Russian Armenia condition, in what they considered as an Ottoman attempt to destroy it. Without taking in account the Ottoman excursion of what was considered as Persian Armenia.
Most of the victims could be counted in Cilicia, as well as the Eastern zone, and without ignoring Smyrna (İzmir) during what was reported as massacres and what followed with the burning of the Armenian and Greek quarters of the city (see Great Fire of Smyrna). While the total of casualties in this category is estimated to tens of thousands to over hundred of thousand, the number of victims is not well established.
While there is no clear consensus as to how many Armenians lost their lives during the Armenian genocide and what followed, there seems to be a consensus among Western scholars with the exception of few dissident and Turkish national historians, as to when covering all the period between 1914 to 1923, over a million Armenian might have perished, and the tendency seem recently to be, either presenting 1.2 million as figure or even 1.5 million, while more moderately, "over a million" is presented, as the Turkish historian Fikret Adanir provides as estimation, but excludes what followed 1917.
Far from finding the exact figure of Armenian casualties, some researchers have at least tried to provide some figures of losses during the war and what followed based on some sources. But most of it is rough estimates or are based on calculations of others. An example here might be the cases of Justin McCarthy, since he is one of the rare researcher that has worked with Ottoman records, various Ottomanists have recycled his figures. Scholarly consensus, however, has largely followed the conclusions presented by Levon Marashlian's study (arriving at a figure of 1.2 million), which claimed that McCarthy's approach suffers from a fatal methodological flaw: in basing his results on inaccurate records. Marashlian maintain there was a reciprocal undercounting on the Ottoman's government's part on the one hand, and underreporting by Armenians, on the other. McCarthy, nonetheless, claims that his results and the Ottoman adult male records were accurate. Others go on to criticize McCarthy in not only having undercounted the prewar Armenian population, but also overcounting the survivors. McCarthy, for his part, argue that his works are too easily labeled by academia as a Turkish-apologist, and complains of a lack of scholarly debate.