King at SCSU Scholars has updated his post attacking the Lancet study with a response to my post. He admits error on one point, but on the rest he has the nerve to accuse me of bringing biases rather than facts to the debate. To see who is bringing facts and who is bringing biases let's look at one of the points he contests:
King originally claimed (my emphasis):
They also chose both to change their list of randomly sampled areas so they didn't have to drive as much; this meant they stayed close to Baghdad and the Sunni triangle, probably oversampling high-casualty areas.
Note that reducing the driving doesn't matter unless it biases the results of the study. King claimed that biases the results upwards by oversampling high-casualty areas. I responding by pointing out that they didn't oversample high-casualty areas:
It is true that sampling was done in such a way as to reduce travel, but this does not mean that they stayed close to Baghdad. They paired Governates that were adjacent and had had similar levels of violence and only sampled from one (randomly chosen) of the pair. This produces no bias towards high-casualty areas.
According to King, that passage contains no facts, just my biases. Now look at how King responds (my emphasis):
See page 1858: "During September, 2004, many roads were not under the control of the Government of Iraq or coalition forces. Local police checkpoints were perceived by team members as target identification screens for rebel groups. To lessen risks to investigators, we sought to minimise travel distances and the number of Governorates to visit, while still sampling from all regions of the country." No Basra and the marsh Arabs. Limited sampling of the Kurdish north (Arbil, for instance.)
Instead of sampling Basrah, they sampled neighbouring Missan (where marsh Arabs also live). If you look at the map (figure 1 in the study) you will see that Missan is not "close to Baghdad". Instead of sampling Arbil they sampled neighbouring (and also Kurdish) Sulaymaniya. If you look at the map (figure 1 in the study) you will see that Sulaymaniya is not "close to Baghdad".
One glance at the map (figure 1 in the study) shows that they did not stay "close to Baghdad", but King continues to insist that they did.
King finishes with:
The Lancet article includes only one violent death of the 7438 preinvasion individuals they interview (see Table 3). That fact alone should indicate the lack of measurement of prison/torture chamber deaths under Saddam.
The Amnesty International Report on Iraq for 2002 estimates that there were scores of killings by Saddam. This number is too small to show up in a survey of the size of the Lancet's.
Update: 
Amazingly, King insists over and over again that they oversampled Sunni areas, despite the clear language of the study and Figure 1, seen to the right, that shows that their samples were not close to Baghdad and did not oversample Sunni areas.
"The Amnesty International Report on Iraq for 2002 estimates that there were scores of killings by Saddam. This number is too small to show up in a survey of the size of the Lancet's."
Yes, but a killing by Saddam regime agents did show up in the post-invasion violent death numbers, and served, ironically, to raise the post-invasion violent death rate ever so slightly.
It's another example of the volatility and potential inaccuracies inherent in some of the violent death subsets from the sample. We also have the unexplained discrepancy between the ex-Falluja bombing death numbers contained in the study's raw data, and the authors' conclusion that approximately 30,000 deaths from the 100,000 estimate were from coaliton bombing.
The study also fails to pick up a single death from an Iraqi regular soldier, yet it seems beyond debate that at least several thousand Iraqi soldiers, possibly ten thousand or more, died during invasion.
All of this does not impact on the aggregate excess death figure, or the rise in the relative risk post-invasion. It does however, raise questions about the manner in which the Lancet study has been portrayed by those who opposed regime change. Too often, we see the " mass slaughter of 100,000 " line being tossed around, deliberately creating the inference that coalition bombs and bullets have been responsible for virtually all the deaths. I want to make it clear that you aren't one of these people, Tim.
This is, however, a valid complaint concerning the study, and one that is independent of the arguments that have swirled around the validity of the study's methodology. The authors appear to have helped advance the misrepresentation I complain of above, by deliberately intending that violent deaths by coalition bombing be singled out as the number one cause of the excess deaths. If the study reflects that, fine. The evidence, however, seems to point to the contrary.
Tim, Try to ignore the King. St. Cloud State University is a fifth-rate school with an emphasis on hockey. That whole team of bloggers from Minnesota has perfected the art of the trawler troll. They cast a wide net of subjects to pontificate on and we feel we have to give a retort because they have close association with Time's Blog of the Year.
I long for the innocent days of the individual troll.