Data and Mines: Fraud at The Atlantic

I’m a big fan of The Atlantic, but they do step into it every so often, and they did this month in a big way in a story that paints a bleak view of trends in Iraq (part of a trend as they seem to be taking a more negative line on the war).

Man Versus Mine: Iraqi insurgents have perfected the use of lethal explosives, with profound implications for our military operations in Iraq

In Iraq the insurgents are using similar weapons against U.S. forces. Today they are called IEDs—for “improvised explosive devices”—rather than mines, and the insurgents are targeting automobiles rather than trains. But the effect is just as devastating.

The number of mines being used in Iraq, and the share of casualties for which they are responsible, dwarf anything ever before seen by the American military. During World War II three percent of U.S. combat deaths were caused by mines or booby traps. In Korea that figure was four percent. By 1967, during the Vietnam War, it was nine percent, and the Pentagon began experimenting with armored boots. From June to November of 2005, IEDs were responsible for 65 percent of American combat deaths and roughly half of all nonfatal injuries.

They present us with this graph (my version, with data from, which shows the percent of casualties (deaths) caused by IED’s by month:

Pct IED Deaths.JPG

Looks horrible, no? The percentage is high and rising, and obviously our troops are in unmanageable peril.

The growing use of IEDs is forcing America’s military strategists to rethink centuries of military doctrine holding that in warfare, mobility equals dominance. Votel told me that given the success that IEDs have had against America’s fleet of motor vehicles, the Pentagon may need to switch to more foot patrols. An intelligence analyst working on the IED problem agreed, saying, “The answer to the IEDs is to leave the vehicles. It’s obvious. It’s the only choice.” But such a move would expose U.S. soldiers to other risks, including snipers. And the December detonation of an IED in Fallujah, killing ten Marines on foot patrol, shows that soldiers will remain vulnerable to IEDs whether on foot or behind the wheel. As long as the insurgents can use IEDs to inflict damage on U.S. soldiers without ever engaging them directly, they will have a tactical advantage. “Our whole military is based on the idea of overwhelming firepower put on targets,” says William S. Lind, a noted military theorist who has written extensively on asymmetric warfare. “But that doesn’t work in this type of conflict. We are fighting an enemy that has made himself untargetable.” Therefore, Lind says, the insurgents can continue fighting the American military in Iraq indefinitely—regardless of how many U.S. troops are deployed or how quickly they are massed.

Fear and uncertainty, of course, ultimately breed mistrust. That may be the most damaging aspect of the IEDs: they prey on American minds, making soldiers suspicious of the local population and ultimately isolating them.

For Lind and other military theorists, the IED problem in Iraq is insoluble no matter how much time or money is spent. “If we can’t engage the enemy,” he says, “what do we do? The answer is, we lose.”

Somehow I was kind of leery of this conclusion.

So I looked at the author’s credits.

Robert Bryce is the author of Pipe Dreams: Greed, Ego, and the Death of Enron and Cronies: Oil, the Bushes, and the Rise of Texas, America’s Superstate.

Somehow, I don’t see him at a BBQ at Crawford any time soon. And Lind is an interesting character as well, worth some scrutiny as a far-right acolyte of John Boyd who posts on the Lew Rockwell site.

So here’s my first point. The war in Iraq – and the wider war it may presage – is a critically important issue that ought to be treated with some intellectual honesty.

Being a Bush partisan whether pro- or anti- and sifting data to look for the nuggets that support your position is basically being a scrivener – a reader of entrails. We need to look as hard as we can for facts.

So let’s look at the facts presented.

My first thought was that the graph represented deaths/month, in which case it was deeply serious. Then I noticed (in the magazine, it’s a small graph – maybe one column wide) that it was the % of deaths caused by IED’s.

So, in theory, in a month when one soldier died from an IED, we’d be talking about a 100% rate.

Which implies that as the operational cadence changes and fewer soldiers are exposed in combat, if the number dying from IED’s were to just stay constant, the percent would spike, as we see.

So I dug out some other numbers.

Here are the deaths from IED’s against all deaths. Note that the pecent of IED deaths spikes in Sept-Oct of 05 at over 80%. Note that the absolute number of deaths in that period was relatively low, under 50. 40 of them were from IED attacks, hence the spike.

US v IED Deaths.JPG

So look at the graph of two numbers – yes, the trend for IED deaths rose slightly through late 05 (and has dropped off steeply since then). But the trend of overall deaths was trending downward at the same time – hence the rise in the % of deaths.

There’s really no other word for what the authors have done – and the Atlantic has condoned – except fraud. I say fraud, which is deliberate, rather than error, which is unintentional, because anyone smart enough to make graphs is smart enough to look at the numbers and come to the same conclusion I came to.

But that doesn’t fit the author’s need to explain why our cause is doomed and why “The answer is, we lose.”

I’m ashamed of the author – who scrives when he should be thinking – and I’m deeply ashamed of the Atlantic, which knows better.

There are real issues about how we’re doing in Iraq and what we ought to do. The Atlantic should be leading the way in asking hard questions to drive that discussion, instead of examining entrails by committing junior-high-school acts of innumeracy.

27 thoughts on “Data and Mines: Fraud at The Atlantic”

  1. Most reporters can’t balance a checkbook or calculate a tip. They simply are innumerate. That’s why they are journalists.

    They also have no knowledge of military operations or concepts. It’s like a mysterious black box to them, comprised of one part My Lai, one part Platoon, Apocalypse Now, and Three Kings, and one part Cindy Sheehan. So yeah of COURSE we have no way of winning, because every reporter “knows” that guerilla warfare is unstoppable and some kind of juggernaut.

    Yes of course it seemed wrong to you, you were suspicious of the source, and did the math. Just as Dan Rather “knew” that of course Lucy Ramirez and Microsoft Word documents from 1972 were the real deal, and that “Million Little Pieces” guy was the real deal (it was Smoking Gun not journalists from MSM that found he was a fraud), this “we can’t win” theme seemed absolutely correct. Because it conformed to what Journalists already “knew.”

    This is par for the course. Real analysis of what’s going on in Iraq comes from Michael Yon, or Adventures of Chester, or Bill Roggio. NOTHING like it happens in the MSM. You wonder why reporters rarely venture out with military patrols despite that being much safer? Because they “know” that the Military is the “enemy” and they “fear” being co-opted that is to write about the bravery, skill, and success of blue collar and middle class young men making life and death decisions every day.

    Joel Stein is part of the problem. His attitudes are widely shared in the Press, which stems from the same elite and pampered/sheltered background.

  2. Well,

    Graphing the number of casualties due to IEDs as a percentage of total casualties tells *a* story, but it certainly doesn’t tell the *whole* story. It looks like another case of lies, damn lies, and statistics.

    I have bought the Atlantic several times in the past year and been deeply dissatisfied with the content. No subscription for me. Someone, somewhere, must still publish a decent magazine.

  3. It’s John Boyd. And yes, Lind is a hardcore paleocon and pretty much an isolationist who has occaisional insights but should generally be taken with a massive grain of salt.

    For the sake of Devil’s Advocacy at least, though, if you want to play the numbers game we can do that. How about the fact that as the number of American casualties has decreased, the number of Iraqi casualties has gone up? Over 2700 Iraqi police and military officers have been killed in the past year — over four times the rate at which American soldiers are dying. Unless you assign greater moral weight to American lives than Iraqi lives, this angle doesn’t exactly have the same “things are looking up!” feel to it.

    I myself am still cautiously optimistic about Iraq’s long-term future in spite of the continuing ugliness, but my point here is that you could just as easily be accused of ignoring broader context because it suits your narrative better. Careful where you point that finger, it’s loaded.

  4. The sad thing is that as soon as I read the axis on the graph, I could see not only the story that was gong to be reported, but the fraudulent means by which it was assembled. It’s really that obvious to anyone even passably acquainted with these matters. (I am certainly no expert!)

    Moreover, I would bet the contents of my just-got-paid wallet that no one in th ePentagon is seriously re-considering the doctrine of superior mobility because of a few IEDs. Not only does the article itself point out some of the idiocies inherent in the position (leaving the vehicles leaves the soldiers more exposed) but there’s a deeper strategic flaw as well: The solution to this is not to give up our mobility, but to inhibit the placement of IEDs by inhibiting the enemy mobility!!


  5. This is one of those “lies, damned lies, and …” situations. The metric is completely bogus and meaningless and is, moreover, misused. For example, consider that time honored statistic of “(some-high-percentage) of accidents happen in the home”. Does this mean you shouldn’t stay at home? That the home is an incredibly efficient death machine out to get you? No. It means that people tend to spend a high percentage of their time in the home, most accidents are due to “pilot error” (slipping, tripping, falling, etc.) and the home contains enough hazards for people to hurt themselves, eventually, if they spend most of their lives there.

    The “percentage of deaths” statistic is even worse, because it has too many variables. Namely, total deaths. As AL points out, if there were 1 IED death per, say, year in Iraq, and no other combat deaths, IEDs would account for 100% of deaths. Yet that says absolutely nothing about the effectiveness of IEDs or the degree of danger for the troops. The ONLY thing such a statistic says is about the enemy. It indicates which weapons are most effective for them. In this case it means that IEDs are the most effective weapon for them. However, it doesn’t mean much more than that. When you look at the absolute figures, and especially when you look at the figures as a percentage of deployed troops in Iraq, you’ll see a somewhat different story. Especially when compared against other wars. When you make that comparison you see something interesting, the current rate of combat deaths in Iraq for US soldiers is very, very low. Several times lower than in all past wars the US has fought except for the first Gulf War.

    Which means that The Atlantic has it completely ass-backwards. The truth is that very few of our soldiers are getting killed in action, and that we have managed to almost completely shut down every fighting technique of the enemy, except for IEDs (mines). And even then we’ve significantly deteriorated their effectiveness, but we haven’t shut them down 100% yet, so that technique is still causing casualties. This lack of complete perfection is not cause for defeatism though, it means only that war is still dangerous (surprise) for US troops, even when it’s dramatically less dangerous than it used to be.

  6. Good comments, all. Matt’s point was well-taken, also.

    Iraqi security forces (police and army) are at over 8,000 casualties I believe, and if you think the armored Hummers are inadequate try doing the job in what amounts to a civilian pickup. Sheesh. They’re showing real heroism out there, and I feel bad for them re: the tools they’re given.

    Still kind of scratching my head re: why the USA hasn’t taken about 1,000 of the 2,000 or so M113s armored personnel carriers languishing in storage, added rubber band tracks and cage armor, and given them to high-performing Iraqi army and police SWAT units.

    Heck, I’m sure some US troops would be glad to trade their Hummers for that. A couple more upgrades and you’d essentially have a bargain basement Stryker with better off road capabilities.

    Given the stuff the US military hasn’t done re: the IED mine threat (deficit of continuous aerial surveillance or boots on the ground over key routes until recently, etc.), it’s somewhat impressive that the trends are what they are. Donald Sensing has talked about American adaptability on the front lines, and that has been a big part: using store bought remote-control toy trucks to scope potential IEDs, downloading jihadist attack videotapes from the Internet and using them as teaching aids for each other (how’s that for Sun Tzu in the 21st century?), incorporating lessons into all soldiers’ basic training fairly quickly, etc.

    The jihadists are learning, too, but this isn’t the Russians they’re fighting, and there really is something to all of Victor Hanson’s points about the Western Way of War and “fight as you train as you live.” The Americans will almost certainly continue to work on their rate of improvement, too – it’s one of their culture’s more annoying traits, so why should the jihadists be exempt?

    Needless to say, this kind of background is almost certainly lost on Bryce. As for Lind, I find him provocative on military matters (his all-time political hero, BTW? Kaiser Wilhelm). But this criticism was both dishonest and – for the reasons associated with my expansion of Matt’s point – very sub-par compared to the legitimate, intelligent, and potentially educational criticisms that might have been delivered.

    When stuff like this comes out, we actually lose twice.

  7. Well, total deaths weren’t exactly trending downward–I blogged about this a while back using similar methods just to see how things were going. The trend line that I found showed a steady increase. However, I haven’t updated the data through Jan 06. Still, it is interesting that they somewhat misrepresented the nature of the graph about IED’s. Good catch.

  8. There are alot of such graphs out there. All of them claim to demonstrate something. Most of them are dishonest.

    I’m an amatuer military geek by way of wargames. I love military history. I love operational analysis. I love crunching numbers. This is what I’ve discovered crunching numbers in Iraq for three years. Past performance is no predictor of the future. There are very little in the way of solid trends. Everyone that has tried to make predictions based on trends has been wrong. The numbers are ambigious. Trend lines which seem to be up don’t explain why the number of deaths in the third year of the war will likely be down significantly from the second. You can’t tell from the numbers whose winning or whose losing. Fatality numbers – which everyone focuses on – are particularly ambigious because they are occuring at such a low rate that the overwhelming factor in thier incidence has been random noise. The wound numbers draw a much better picture, but one which remains decidedly ambigious except in hindsight.

    There are two solid things I think that can be seen in the numbers. One is that the conflict in Iraq is not one war, but two wars mashed together. The first war is a war to oust the regime of Saddam Hussein. It’s pretty much over by Feb. ’04, and Saddam Hussein’s side lost. The second war is a war to control how the vacuum created by that first war’s success will be filled, and its a much more intense and bloodier war and its still raging inconclusively.

    The other is that in every year of the war, Ramadan was the bloodiest month of the year.

    Anyone that claims to have proven something about the future from the numbers is I think bogus or self-deluded. I would not at this time make any predictions about the 4th year of the war – either that the situation will improve, or worsen, or even stay the same. If I were forced to bet though it would be that it will be closer to the latter option.

  9. ..and the uptick in Iraqi Security Forces is a good (yet morbid) indicator.

    it is a positive sign of more involvement of Iraqis getting down to the business of providing security for the Iraqi people. i thought thats what everybody wanted?

  10. In other words using truth in a way to mislead into a false Propaganda conclusion like this article anti-war article at the Atlantic has done.

    The Mainstream media has since the beginning been using propaganda in the style of truth in a misleading way mixed with some pure false hoods since the beginning and it is wrong to be allowed to do such with no challenge short the blogosphere.

  11. As Gerbils once stated “good propaganda is part truth”.

    In other words using truth in a way to mislead into a false Propaganda conclusion like this article anti-war article at the Atlantic has done.

    The Mainstream media has since the beginning been using propaganda in the style of truth in a misleading way mixed with some pure false hoods since the beginning and it is wrong to be allowed to do such with no challenge short the blogosphere.

  12. We had a saying in one of my stat/methods classes: “torture the data long enough, and it’ll confess anything.” Unfortunately, that phenomenon isn’t all that uncommon.

    Lind’s useful sometimes as an iconoclast, but he’s such a 4th generation warfare True Believer that he advocated creating armed militias in the U.S. to operate as an anti-4GW force. To use another saying, “To the man who only has a hammer in the toolkit, every problem looks like a nail.”

  13. I know we’ve discussed this before but I think this article demostrates quite clearly why many view the enemy as inexhaustible. Through the limited prism of taking into account only US deaths without concern of enemy casualties or the relative effectiveness of a weapon system most are left with the impression that there is no way out. However, while we don’t do body counts most analysts agree that we’ve killed thousands of jihadists and ex-Baathists and from what I hear on the ground the amount of IED’s that actually cause harm to a target in relation to how many are planted or exploded is quite low. In fact, if anything the use of uparmored vehicles and changes in tactics have brought the success rate down quite considerbly. Again, admittingly I pull from a limited sample but nonetheless it gives some perspective.

    RE: M113

    The M113 didn’t get the nickname “death box” for nothing, its slow, extremely loud and ackward. Better than a Toyota pickup truck? Possibly, but the tradeoff in agility can have consequences.

  14. A.L.,

    I went ahead and did a chart incorporating 90 and 120-day moving averages myself–see here. The 120-day line is in yellow and hard to see, but it looks pretty much like the 90-day average.

  15. Nice chart. Fatalities are up from the start of the period, but well off their peak and trending downwards. I’d like to see this chart with Vietnam casualties overlaid during a similiar time period. That would be good amunition against that interminable comparison.

  16. “Nice chart. Fatalities are up from the start of the period, but well off their peak and trending downwards. I’d like to see this chart with Vietnam casualties overlaid during a similiar time period. That would be good amunition against that interminable comparison.”

    Heh. You’d think, but you’d find that you’d get into a huge argument about what constitutes a ‘similar time period’. Liberals will go though back flips and contortions to try to convince themselves that two unlike things are alike. Believe me, I know. I’ve had this debate.

    But still, if someone were to create such a graph, it would be nice to have. I already know what it would show though. Second Indochina (Vietnam War) was about 5 times as intense as Gulf War 2, with fatalities in the 100’s per month for an equivalent number of deployed troops.

    What I would really like to have that I don’t have is a list of American casualties during Operation Southern Watch and Operation Northern Watch. I know that there were some pretty bad days during that period (I recall one incident of like 30 US KIA do to friendly fire), but I’ve never seen a full listing or even a summary of how many people we lost during the sanctions period to all causes (hostile and non-hostile). It’s my conjecture that the casualties of the sanctions period of the Iraq conflict will be comparablely to the pre-Gulf of Tonkin resolution period of the Vietnam Conflict, but that the intensity will scale roughly the same way that the post Gulf of Tonkin scales with the post invasion period of GW2.

  17. Thanks for the corrective. I read the Atlantic piece a couple of weeks ago and had a vague feeling that Bryce’s article was simply dated and I had read better information on blogs.

    That said, William Langewiesche’s piece on A.Q. Khan is well worth reading.

  18. It’s just what Robert Heinlein said so long ago: The best way to lie is to tell part of the truth. And Mark (#17) I did see a chart that had both lines on it. The Vietnam line was like a steep mountain cliff compared to gently rolling hills for the Iraq line. One variable I’ve not seen anyone consider (at least in the media and blogs I’ve read, a minor piece of a percent, I’ll admit) is to look not only at the Vietnam war line, but analyze both based on the moving average number of troops actually exposed to the enemy. Using Vietnam as an example, even though the US had as many as half a million in country and stationed in places like Thailand and on ships in the South China Sea, the total number actually involved in ways that exposed them to harm never exceeded 80,000 or so up against VC/NVA strength of 400,000 or more (source: Unheralded Victory by Mark W. Woodruff, p. 252). These days in Iraq, even the people who are stationed in the Forward Operating Bases (FOB) are out running convoys and doing other things that expose them to danger. That’s why the combat troops in Iraq have stopped referring to rear area troops as REMF (and no, I won’t explain the acronym, ask a veteran). Now they’re called FOBbits, a play on the FOB and hobbits, creatures that live mostly underground and are not war-like, but fight quite bravely when forced.

  19. Just for posterity, I added the numbers from Dec and Jan and it does seem to reveal a slight decline in troop deaths. For a large version of the chart (sorry about the first one) see here. Also, what I find to be interesting more so than death numbers, is the fact that US wounded are also trending downward, from 618 in October, to 466 in Nov, to 408 in Dec, to 259 in Jan (the lowest total since Feb of 04). What is more encouraging, as has been mentioned, is that the number of Iraqi police/military are also down, the lowest since Feb 04. Do these numbers tell the whole story–no, but it gives you something to chew on.

  20. M113? This is another one of those worn out arguements that has been going on for a couple of years. Slow, death box, won’t stop an RPG. Ask one of the older vets about these, he will tell you that they were not that great.

  21. >>M113? This is another one of those worn out arguements that has been going on for a couple of years. Slow, death box, won’t stop an RPG. Ask one of the older vets about these, he will tell you that they were not that great.

    My older vet friends definitely consider the M113 a &*^%$ deathtrap. They’d far rather sleep on TOP of one than ride in one. I seem to remember them saying something about “50 cal goes right into it, and small arms fire causes armor spalling on the inside.”

  22. Right. Given the M113s’ presence in storage doing nothing, one might think they would be useful to folks whose alternative was a new GM pickup (which are bought through MAC/Mansour in Dubai). Or a left-over Toyota pickup. Or, frankly, an “up-armored” HMMWV.

    “Iraq’s lone armored division currently uses BMPs,”: which are familiar to them but have weaker engines than even the Vietnam M113s. They are also even more of a death trap due to fuel placement that can block the exit when hit (on the plus side, BMPs have 30mm-76mm guns up top). Iraqi forces also operate Soviet MT-LBs, introduced in 1970. Troops have to exit that one through the roof. Sounds like fun under fire in urban terrain.

    If they wish to keep that stuff, fine it’s their country. But there should be a mechanized element for all other divisions, especially the ones that are performing well. M113s won’t replace the pickups entirely, but if upgraded slightly they’d be a supplement that would make a big difference.

    The M113 has since been up-engined to give it good speed, plus there’s now slat armor or active armor to stop an RPG. Rubber band tracks have longer wear and less maintenance, and are as easy on roads as wheeled vehicles. They aren’t your father’s M113s any more, and some are serving with US forces – “and serving well.”:

    The Israelis have also used M113s since Vietnam with success, though they have also developed heavier alternatives like the Achzarit (turretless T-55s converted into APCs and given remotely operated weapons up top and heavier add-on armor protection) for full-on urban combat. And of course, there’s their Merkava main battle tank’s ability to swap out some storage space and carry 1-4 infantry.

    The Iraqis will never have Merkavas. But they could, and should, be getting better equipment than they have. Especially since “better” could be cheap and is available more or less immediately.

  23. My Atlantic subscription just expired, and I was thinking about renewing. Then I looked at the Jan/Feb issue, and in addition to this bogus article about IEDs, there was fatuous defense of mutually assured destruction, an infantile and unfunny “humor” piece on the Bush presidential library, and some lame parodies of the national anthem by the tiresome Garrison Keillor. Do I need to pay for this? Nope. No renewal…

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>