Through a Glass, Darkly


Played to the first opium harvest.
Played to the first opium harvest.

Here’s the final position after four turns and maybe five hours of semi-chatty play. Max and I could have knocked it out faster. I wanted to wait to post further impressions until the experience percolated a bit more. I won’t write about the cut-and-thrust of game play. A deeper current needs wading into. After some correspondence with Brian where we chatted over the wheels-within-wheels nature of the current Afghan War, how to design processes, and the complexities of real-world COIN campaign design, I believe Kandahar is the darkest of all the counterinsurgency designs I’ve seen.

“How do you win?” Max asked. “On joue jusqu’à l’épuissement,” I explained. We play until we quit playing, would be another way of putting it. This game, unlike Train’s other Box4 designs, depicts warfare in an unhappy corner that has known serious conflict for more than thirty years now, where the culture is as broken as the irrigation system the Soviets destroyed, where conflict has become the culture. In that it reminds me of Ben Madison’s  courageous Liberia: Descent into Hell.

Of Brian’s designs with which I’m familiar all attempt a similar level of honesty. They are deeply skeptical of ideology and other power fantasies. Their mechanisms give you plenty of rope with which to hang yourself. Glorious blitzkriegs, chevauchées, razzias, and shit-hammerings are rare and never engaged without careful calculation of the downsides. If he had designed OGRE, there would be the possibility of co-opting the cybertank by surrendering power stations along with a good wash and detail. I still say his Algeria is way too hard on the French, although I’ve been able to win playing both sides.

These are, in their way, anti-games. They resist commercialization in the best way by raising the bar for their audience while keeping their author impoverished and angry. And they queer up the taste of a Saturday afternoon. I’m not sure if I share Train’s (and everyone else’s) evident pessimism about counterinsurgency. The undergraduate anthropologist and graduate cartographer in me want someone to say it ain’t necessarily so. But I have heard the frustration expressed in professional quarters for going on ten years now. After a certain point, social network analysis diagrams, or DIME models, or whatever science-po construct you prefer all begin to look like Gordian knots to audiences where everyone carries with them a brigade’s worth of swords in addition to levels of impatience and frustration which can only be calculated by the f*ckton.

To me, Train is of the camp that holds COIN is war as we would like it to be, not as it is. There is nothing of the revolutionary struggle in Kandahar. (My skepticism of revolutionaries very likely outstrips Brian’s.) Anyway, hold on to the concept of war as we would like it to be, for it clears away so much fog in the wargame world that I’m certain to return to this important distinction in future posts.

Here’s the position in Iraq when Barack Obama stepped away from the table:

LLOs from FM 3-24. Sure, sure. How do we revise our model to account for Da’esh?


Yeah, but where’s the nearest pizzeria?


Part of 315 FA's fire plan supporting the capture of Gourbesville. Our summer quarters highlighted in yellow.
Part of 315 FA’s fire plan supporting the capture of Gourbesville. Our summer quarters highlighted in yellow.

In case you were wondering what was going on in the vicinity of the spot where we’ve selected to camp out for a week, here is a map I’ve made from 3/357IR and 315FA’s battalion journals. The highlighted spot is Les Tourelles farm at Orglandes. It seems clearly situated on the German LOC. Judging by the lines drawn 15 June 1944, and the fact that German battalion mortars are about a kilometer from the line of contact (the ones that 315FA spotted anyway) it looks as if the enemy were already withdrawing from the sector by the time Sam Williams’s attack was launched. Also of note, this was the first time 343 and 315FA were able to get it together to fire preparation on Gourbesville center.


Defense Linguistics's CP, vic Orglandes 11-18 June.
Defense Linguistics’s CP, vic Orglandes 11-18 June.

Announcing the From the Ground Up tour and playtest. We’ll be staying near Orglandes in La Manche, 10-18 June, seventy-two years to the day that 90ID disembarked at UTAH and began feeding itself into the grinder that was Lower Normandy.

Plan is to roll out every morning, roam where permitted, get rained on, photograph, eat sandwiches, and lots of fussing with maps and binoculars. Evenings at liberty, drink, or fussing with games or all three. Reserved a whole week so there’ll be plenty of time free to visit other, more well-known spots. La Fière and Amfreville (Timmes’ Orchard, Hill 30, the causeway, etc.) are not far. Nor are Ste. Mère Eglise, UTAH, Carentan.

Racks enough for five more. Don’t bother to call ahead, y’all — just make like 82ABD and drop in. For those of you who can’t make it but who might have questions about the cartography, here’s a list of points of interest I’d like to visit. Feel free to add to it!

Sèves St. German, Catchment Basins

Catchments prior to calculating flood plain. Using QGIS/GRAS, SRTM30 DEM, and OSM as stream centerline.
Catchments prior to calculating flood plain. Using QGIS/GRAS, SRTM30 DEM, and OSM as stream centerline. Click through to a manipulable 3D model.

Taking out a little time to look at Grass GIS. The objective here is to do a simple (heh-heh) floodplain map for Sèves St. Germain. These are the catchment basis as calculated from SRTM30 data via QGIS/GRASS, which will serve as a basis (I hope) for a flow model. Downstream is to the north east (right in the static image. Here’s a 3D model.)

So why does an historian, even an amateur one care about land areas contributing to runoff toward a given flow? Well, this is the terrain that first and second battalions of 358IR attacked across. The plan called for crossing tanks, only there was a pretty big rainstorm in the evening. The crossing was already contested and the water high, but should have been possible the following morning. If I can track down divisional G2 meteorological information from 1944, I should be able to feed it into the model and come up with a better visualization of exactly how high the water was.

Lots of caveats here. The SRTM was taken in the 1990s and the streambed has probably changed a bit since. Metaled (paved) roads are now the norm, which changes the flow coefficients. But enough. I just want a better idea of how high the water was back then. And it is keeping with my design objective to bring as much data to this design as I can.