Daily Attacks Chart

I stole this graphic from Brian Beutler. It serves as a reminder of how far things have slipped in Iraq. In April 2005, people generally thought we were having a difficult time of it in Iraq. And if you'd suggested then that the daily number of attacks in Iraq would get to around 100, people would have understood you as predicting a dramatic worsening of the situation. From today's vantage point, however, 100 would be major progress. But progress toward what? Toward a return to the unacceptably horrible conditions of early 2006, I guess.

But you can see it all on the chart or any other years-long metric of the war -- if there ever was a time when the situation was amenable to "fixing" it was a long, long time ago before things metastasized and anything resembling the current dynamic took hold.

Comments

There's only one cure for this: invade Iran!

Posted by Ugh | July 19, 2007 8:11 AM

Note also that there's always a dip this time of year. Let's see what August-October brings.

Posted by dbomp | July 19, 2007 8:19 AM

attacks are up = the surge is working

Posted by cleek | July 19, 2007 8:41 AM

Anything going up is good.

Especially for the Republicans.

Posted by gregor | July 19, 2007 9:10 AM

Sen. Vitter is in favor of the 'surge', that's for sure

/rim shot

Posted by r€nato | July 19, 2007 9:12 AM

This graph clearly shows a downward trend from the recent high of 180.

Posted by Tony | July 19, 2007 9:44 AM

"This graph clearly shows a downward trend from the recent high of 180."

Tony, I hope you were trying to be funny as I would be amazed that anyone this stupid could figure out how to find this site in the first place.

Posted by frankly | July 19, 2007 9:59 AM

Man! The insurgents sure are getting good at that "last throes" thing.

Posted by Shawn K. | July 19, 2007 10:11 AM

Am I reading correctly that this data set ends in April of this year? Are the numbers from May and June not available?

Posted by Midwest Product | July 19, 2007 10:30 AM

I remember arguing over the early numbers in this chart with the hacks over at Jane Galt's place way back when there were only a few data points. Wonder how they spin it now?

Posted by decon | July 19, 2007 10:46 AM

"Are the numbers from May and June not available?"

My guess would be is that they are in the same neighborhood as December-April since the number of U.S. dead is about the same.

Posted by Jose Padilla | July 19, 2007 11:31 AM

This graph is clearly statistical, and since you're not a statistician or a "statistical correspondent," you shouldn't feel qualified to comment. In fact, yours is a special category of wrongness. [/McArdle]

You should know, for instance (or would if you were qualified), that one of van Belle's Statistical Rules of Thumb is (and I quote) "Always think of an alternative to a bar graph." (p. 162).

In my alternative version, which I refer to as a "Plot O'Gold", the evidence shows that everything is going swimmingly in Iraq.

And I'm a statistician. So there.

Posted by rb | July 19, 2007 11:46 AM

decon: "I remember arguing over the early numbers in this chart with the hacks over at Jane Galt's place way back when there were only a few data points. Wonder how they spin it now?"

Faster and faster.

Posted by Barry | July 19, 2007 1:57 PM

Is that a Laffer Curve I see....?

Posted by Stefan | July 19, 2007 3:20 PM

The problem with graphs like this is in the nature of the stuggle.

When enemies operate covertly it is hard to know what efforts they can make or how long they can continue them. I certainly don't.

It would be interesting to see militia attacks separated from insurgent or unknown. There was a major change in tactics (the so-called surge to control Baghdad) at the time of the big rise.

Posted by K | July 19, 2007 3:24 PM

If the number of attacks goes up, the surge is working because there is more contact with the enemy. If the numbers go down, the surge is working because the number of attacks has decreased. It's classic Sophistry 101.

Posted by Bill Jensen | July 21, 2007 11:20 PM

If the number of attacks goes up, the surge is working because there is more contact with the enemy. If the numbers go down, the surge is working because the number of attacks has decreased. It's classic Sophistry 101.

Posted by Bill Jensen | July 21, 2007 11:21 PM

Two points:

1) The surge technically has only recently started to operate; troop levels did not reach what was contemplated until about a month or so ago, IIRC. Which makes both the chestthumping about how well the surge was working and the hair-tearing about how poorly it is faring equally misplaced.

2) You can't really tell anything about the effectiveness of a take-clear-and hold plan (which is what the surge is) based on raw numbers. If attacks are going up in Iraq as a whole, but down in the areas of Baghdad that the surge is concentrated in its initial push, then the surge is doing what it was meant to do: create areas of safety in the city that is the literal and metaphorical heart of Iraq. And those zones of safety can be built on and expanded outward, hopefully. If that is the case, I'd *expect* the number of attacks to go up, as insurgents or al-qaeda seek to get the US to abandon the plan and go chasing after them in the latest hotspot; that type of "whack-a-mole" strategy is a recipe for failure.

On the other hand, if the attacks are distributed across areas in roughly the same proportions next month as they were 4 months ago, and the surge isn't impacting that, then the surge is failing in its objective.

Either way, tallying up the total number of daily attacks across Iraq doesn't give you a meaningful basis to draw conclusions one way or the other

Posted by Akiva M. | July 22, 2007 12:14 AM