4.1: Data Skeptics

Author’s photo. Afghanistan, 2010.

In the year since my first article, people across the army and DoD have reached out to share their support and some of their own vignettes. You’ve met four of them: X.4’s Zach Griffiths, X.5’s Kevin Horrigan, X.6’s Matt Moellering, and X.7 & X.8’s Natalie Stone. The article has shown up in the curriculum at CGSC, SAMS, the AWC, and even at the SMA.¹ The latest round of CAP is starting up this month, and I’m waiting to hear from peers if they’ve incorporated a data literacy assessment.

But not everyone has agreed with me. There have been plenty who offered their own counterarguments. Minus one Colonel who mistook my argument to be we didn’t need commanders who can pass an ACFT², these responses have generally fit into one of four critiques:

• Paralysis by analysis

• I have a staff for this

• What about when EW hits?

• I haven’t needed it thus far…

I’ll offer my responses to each in turn.

Paralysis By Analysis

To those who think using data will jam up us commanders and our headquarters, I’d remind them this isn't new. History has plenty of commanders that dithered like Hamlets and incurred their own blood baths as a result. Being data-driven doesn’t mean ignoring the time constraints of the decision we need to make. It’s about trying to make better ones.

Combat makes decisions even harder. We learn by heuristic, but very few commanders get enough reps and sets in to really develop the tacit knowledge you need to have a qualified intuition. This does not stop any of us from having a gut instinct though. Data helps by giving us more than just that undereducated gut to go on.

Just gathering data isn’t enough though. The key is to be data literate. Forgetting how small your dataset is, or cherry picking the data that confirms your gut can both lead to poor decisions. A data literate commander can cheat ahead by identifying what data we need to make a critical decision and then develop a collection plan to get the best information we can in the limited time we have.

I Have A Staff For This

I addressed this in 3.6: Data Toddler6, but I’ll briefly recap. Staffs are part of the process, but if I as a commander relies sole on my staff (or worse, a single data officer) to be data literate, I am slowing everything down. Only commanders can ask commander questions.³ Worse, if a commander isn’t data literate, they’re surrendering some of their command authority to the staffer who is, because they have to uncritically trust the data they’ve been given. A commander who lacks data literacy can’t develop a solid data plan to help make the decisions they can’t make yet, and is left with only their gut. You get your gut for free; spend the effort to get the data.

Not everything in the army moves at the speed of combat. But when seconds matter, data literacy saves time by allowing us commanders to engage directly with the data our staff has given us. When you’re not under as tight a deadline, those seconds can compound into hours and even days, buying your unit the time it needs to experiment and innovate.

What About When EW Hits?

Prepper-porn fanfic isn’t new. Throughout my career there has been this pervasive idea that we'll all end up fighting like the Amish in a future war because ‘all the digital systems will go down’. I’m fairly certain the first human to string a bow heard something similar. Steam powered ships, cartridge ammunition, night vision goggles; I’m confident with every innovation, some nagging soldier was asking, ‘What’re you going to do without coal / black powder / batteries?’

Data systems do go down, and EW is absolutely a thing.⁴ But look on the battlefield in Ukraine today. The most contested EW environment in the world isn’t pulling back from digital and data. Quite the opposite, both sides have unleashed it from antiquated acquisition systems and are moving faster and flatter. The Ukrainians innovate constantly on how to regain access to spectrum and how to fight through EW challenges. And they are killing every day despite a very capable EW adversary, one the US Army has rarely trained for. Any future fight will require our forces to be just as adept.

The data transition doesn't mean we go fully digital anyways. You'll find a large and well used dry erase board in every office I've ever had. I still rely on an analog map, just like I did in my Bradley days when building a map board was a self-taught art.

Land navigation serves as another great example. You’ll find soldiers today who think they don’t need to do land navigation training because they have Garmins and other GPS. They’re wrong, because if you can’t read a map, then you can't plan an offence or a defense. It's not either or. You use both, GPS and the map. GPS might have been the first small step in our future of centaur ‘human machine teaming’.

We stopped relying on solely iron rifle sights decades ago. We adopted a plethora of new, more accurate optics. Be we keep a backup iron sight, or the optic has a fixed point. We use the best tech we can, while understanding that means having to work to maintain it. You’ll have to fight for your data, but you’ll kill with it as well.

I Haven’t Needed It Thus Far…

There’s one expression you’ll hear all throughout PCCs and at the AWC: ‘What got you here can’t get you there.’⁵ It’s supposed to remind us that we’re entering a new phase in our careers. One that requires us to drop old tools and pick up new ones. To swallow the pride that comes with being selected for command and recommit yourself to curiosity and humility.⁶

The character of war is changing, and data is no longer a choice. You see it in Ukraine, and you see Israel using it right now on the battlefield. We are behind our adversaries here, because we haven’t been forced to adapt yet. US Army leaders keeping data at arms distance is why we still don’t have human machine teaming, a decade after third offset. New domains mean new data sets.

Centaur fights, or human machine teaming, are data fights. EW fights are data fights. You can't win in these conflicts with commanders who are waiting for everything to get converted back down into PowerPoint. If you're going to use machine learning, and more critically if you're going to deceive enemy ML, you need to understand data.

So, the real question is why are you, commander, not working on your data literacy? What are you doing that’s so much more important? And why are you not demanding more from yourself?

1. Command and General Staff College. The Army’s school for majors at Fort Leavenworth Kansas. School of Advanced Military Studies, also in Leavenworth is where the army trains ‘jedis’. The Army War College, the army’s senior service college for colonels. The Sergeants Major Academy in Fort Bliss Texas, where the Army trains most of its SGMs.

2. It was very much not:

3. Ostensibly only commanders can write commanders intent too. I got this wrong in ILE when I answered the S3 did, because apparently, I’m the only one in the army that’s been in that sort of a unit in the army. 😏

4. Electronic Warfare is using directed energy to control the spectrum or deny it from your opponent

5. Pre-Command Courses. One-to-two-week training events newly selected commanders and CSMs attend to prepare them to take command.

6. Officers who are selected for command often think we are the best of the army, when really, we’re what’s left of the army.