Cycloptichorn wrote:ican711nm wrote:Cycloptichorn wrote:ican711nm wrote:Cycloptichorn wrote:Ican,
You are appealing to Authority, a logical fallacy.
You too are appealing to authority. Only in your case, you are appealing to a much lesser authority whose report is self contradictory. How about that for a logical falacy?
I still haven't seen Britannica's data or methodology. Without seeing those, I can't make a rational decision about whether or not their figures are accurate.
You made an irrational decision based on a report that is self-contradictory and that describes a polling process that obtains the unverified opinions of people that may or may not constitute a true statistically accurate representation of the Iraqi population.
Therefore, your argument fails due to reliance on a source which cannot be proven to be true or false. You cannot independently judge Lancet's numbers using a source which doesn't provide methodology or data. There is no reason to believe that their casualty numbers are in fact correct, yet you seem to take them as gospel. Most likely, b/c it supports your position better to do so.
It isn't my argument that fails. It is your reasoning that fails. My source does provide data, lots of data for lots of countries--more than 200 countries in its Britannica Book of the Year 2007, section titled, Britannica World Data.
You and I can independently judge the validity of Lancet's data based on the content of its report without resort to any Britannica reference. As I have already explained, Lancet contradicts its own Figure 4 on page 7 with its own well advertised data given on pages 1 and 6 of its report.
You or anyone else can checkout both Britannica's data and its specified sources (e.g., page 852 of Book of the Year 2007) simply by going to a public library and accessing the relevant Britannica Books of the Year--2004, 2005, 2005, and 2006, and 2007.
You do have reason not to believe the Lancet report--that reason is contained in the Lancet report. You do not have any reason not to believe Britannica's data--data they have been supplying for 22 years, originally for more than 160 nations and currently for more than 200 nations.
Whether you like it or not, the total number of deaths in Iraq since the war started cannot possibly be less than the number of "excess Iraqi deaths as a consequence of the war." But Lancet claims otherwise. 
Cycloptichorn
Bullsh*t. Crying about a 27-person difference in statistics which number in the hundreds of thousands is ridiculous.
...
I'll trust the word of people who know something about the subject over the half-assed analysis of a senile old man any day.
Cycloptichorn
First, thank you for your very creative and enjoyable concession to me via your personal attack on me: "half-assed analysis of a senile old man."
Second, the difference between Lancet's claim on page 1--or page 6--and Lancet's claim on page 7, Figure 4, is not "a 27-person difference."
Even this "half-assed ... senile old man" can subtract better than that.
Specifically, I am challenging the validity of the Lancet report based on the
difference between the
654,965 number on page 1, and the
113,000 sum computed from page 7, Figure 4 (21,000 +35,000+57,000 = 113,000). That's a tad more than 27--
wherever in hell you got that number.
That difference is: 654,965 - 113,000 = 541,965.
Ican, Jeez.
The numbers on the left aren't the total numbers of dead per violence per year, it's the average number of dead per violence per year for the entire course of the war.
Look at the right side of the graph - deaths per 1000, averaged out over the entire war.
20 deaths per 1000.
20,000 deaths per 1 million.
500,000+ deaths per 26 million.
Like I said - I don't trust your analysis, b/c it isn't particular well thought-out. Instead of looking for 'gimmes' like you have been, you would be better off buying a statistics textbook and a copy of the SPSS software. A few months of study might give you some added credibility.
Cycloptichorn
Wrong! In Figure 4 on page 7, the ordinate on the left specifies
number of deaths. The abscissa at the bottom specifes
four different time periods. The Lancet points (little square boxes) and the IBC points (little diamonds) specify the total number of deaths in each of the four time periods. The ordinate on the right specifies
deaths per thousand per year.
For example, Figure 4 shows that in the period March 2003 to April 2004 Lancet claims there were about 21,000 deaths & IBC claims there were about 15,000 deaths--15,000 is about the same as IBC actually claims for that same time period.
According to the ordinate on the right, 21,000 & 15,000 deaths amount to about 8 & 5 per thousand = 8,000 & 5,000 per million, or 80,000 & 50,000 per 10 million, or 160,000 & 100,000 per 20 million, or 208,000 & 130,000 per 26 million. The Lancet-left numbers of 21,000, 35,000, and 57,000 correspond to Lancet-right numbers, respectively, about 8 per thousand, 12 per thousand, and 20 per thousand. For a population of 26 million those would equate, respectively to 208,000, 312,000, and 520,000. Their Lancet-right sum is 1,040,000--not the Lancet-left sum of 113,000, and not the Lancet claimed number of 654,965.
If the Lancet-right ordinate were divided by 10, the result would be more consistent with the sum of the Lancet-left numbers: 104,000 instead of 1,040,000. Or if the Lancet-left ordinate were multiplied by 10, the result would be more consistent with the sum of the Lancet-right numbers. I choose the former, because its more consistent with the IBC numbers. You will probably choose the latter because
