What are the criteria of inclusion of a conflict in the list of wars?
NOTE: These criteria are used for data from UCDP, that is: almost all wars from 1989.
Inter-state or intrastate wars
I've included data if a conflict in UCDP Battle-deaths Dataset is coded as war, that is if >1000 battle-deaths occurred in one year for this conflict (there are a few exceptions).
I've included fatality data from UCDP Onsided Violence Dataset if:
... the actor that committed that onesided violence is also an actor in a specific war in the list, and the killing took place during the same time period as that war.
... if an armed conflict is categorized as minor onesided violence data will be included in the list, if 1) the above criteria is met concerning actors and time periods, and 2) if the number of battle deaths plus deaths of onesided violence exceeds 1000 in a calendar year.
Non state wars
I've included data from UCDP:s Non-State Dataset if ...
... a dyad with some certainty can be connected to a war in the list, in which case it is presented in the context of that conflict.
... if more than 1.000 battle-deaths occurred in one year for this conflict, in which case it may be added as a separate war.
Note: non-state war data has not been completely implemented in the list.
What definition of war is used on this website?
The list of wars uses, with a few exceptions, the definition of a war as being an armed conflict (interstate or intrastate) with at least 1,000 battle-related deaths in one calendar year, which was put forward by Small & Singer in 1972. Most of the wars in the list meet this criteria in one or several years. A few do not and should rather be classified as minor armed conflicts.
There are, of course, critics against Small & Singers definition which UCDP, PRIO and COW base their datasets on. See for example Herman and Peterson and this article at BMJ and the answer by the Human Security Report Project.
The reason I've chosen to base my list of wars on Small & Singers definition is because it is used by many academic institutions that offer well researched data.
 J. David Singer and Mel Small, The Wages of War 1816-1965: A Statistical Handbook,
What is the base for the fatality numbers in the war list?
The fatality numbers that are presented in the list of wars are predominantly, but not consistently, so-called "battle related deaths" or "battle deaths". For most of the conflicts after 1989 fatality numbers for "onesided violence" have been added to the total number.
The definition of "battle-related deaths" is not conclusive from source to source, and specially when it comes to civilian deaths. UCDP/PRIO include civilian deaths in their numbers in some circumstances. COW on the other hand only include killed military personnel in their fatality numbers as far as I understand. Their data is used as the base for the war list for wars before 1945.
A good definition of battle-related deaths can be found at UCDP:s website: "Typically, battle-related deaths occur in what can be described as "normal" warfare involving the armed forces of the warring parties. This includes traditional battlefield fighting, guerrilla activities (e.g. hit-and-run attacks / ambushes) and all kinds of bombardments of military units, cities and villages etc. The targets are usually the military itself and its installations, or state institutions and state representatives, but there is often substantial collateral damage in the form of civilians killed in crossfire, indiscriminate bombings etc. All deaths - military as well as civilian - incurred in such situations, are counted as battle-related deaths."
Many of the fatality numbers in my list include civilian deaths, but they are not consistently classified and thus not really comparable with each other. Some include battle-related civilian deaths while others include both battle-related deaths and deaths due to famine or disease (often called "excess deaths"). Some include "massacres" and "purges" that were related to a war (onesided violence) but most do not.
Since August 2013 I've started to include numbers from UCDP:s dataset for one-sided violence that records events from 1989 and the killing of more than 500,000 civilians, as well as UCDP:s numbers on non-state conflicts. This work which has not yet been completed. The Konstanz One-Sided Event Dataset (KOSVED) is another interesting effort to gather data on violence against civilians that I have not yet used on this site. The chronological scope for these datasets are quite limited and do not account for excess deaths. I also occasionally include sources dealing with separate conflicts, such as the Iraq Body Count. For excess deaths I occasionally cite Matthew White's findings at necrometrics.org.
 See “The COW Typology of War: Defining and Categorizing Wars (Version 4 of the Data)” (PDF) pages 14-17.
 See Human Security Research Group for a discussion on civilian fatalities here
Why are only wars from 1945 included in the online memorial collection?
Those killed in the First and Second World Wars have not been included here. WWI and II were such disastrous events in the history of mankind that insitutions, like the United Nations and the Geneva Conventions, were set up to prevent such bloodbaths in the future. The Second World War constituted a distinct and terrifying mark on the timeline of the history of mankind and is therefore one (of many) logical starting points for this project. WWII has also been touroughly documented. This project aims instead to remember those that were killed all over the world despite the existence of the protective measures, institutions and security policies that emerged in the late 1940's.
We only exclude sites when we are not sure if they are overlapping with other sources, if they are displaying unlawful content according to Swedish law or if they are in a language that we cannot get translated at the moment. We appreciate any help to locate virtual war memorials that we may not be aware of at this time.
Why isn't this or that genocide/democide/mass-murder included in the list of wars?
Acts of genocides, or similar acts of deadly violence against a large number of a civilian population, are generally NOT included in the list of wars (see definition of war in this faq). I have, however, started to include data about so-called one-sided violence in the fatality data from the UCDP onesided violence dataset (UCDP OSV, see sources). This dataset has data from 1989, which means that many of the large-scale atrocities of the 20th century are not included!
The definition of onesided violence is when a state or a named, armed group commits acts of deadly violence agains unarmed, civilian populations. The following rules applies for inclusion of such acts in the list of wars:
1) Fatality numbers for onesided violence are included for a specific war if they appear in UCDP:s OSV dataset, are ascribed to the same actors that figure in that war, and if the deaths are ascribed to the same time period as that war.
2) If an armed conflict is counted as minor it will be included as a war in the list if the above criteria is met and the number of battle deaths plus deaths of onesided violence exceeds 1000 in a calendar year.
Please also note that one-sided violence is not the same as excess deaths (deaths due to famine or disease as a result of warfare). Excess deaths are not included in the fatality data.
I'm assessing the possibility to also include one-sided violence from before 1989, but that is a huge effort that may run into numerous problems. Also, I have not yet found a suitable, comprehensible and coherent dataset that cover the whole period 1900-1989. For those who want to examine such atrocities further I strongly recommend a visit to Matthew White's website.