The Military Decision Making Process

How MDMP Actually Works When You're Not in a Classroom

I've run through The Military Decision Making Process more times than I can count, usually with people who have never done it before and a clock that's already running too slow. The process itself isn't complicated in theory. It's seven steps laid out in doctrine, and on paper they make perfect sense. In practice, you're working with incomplete information, conflicting advice, and a staff that is tired. That changes everything. Here's the sequence as the manuals define it: Receipt of Mission

Mission Analysis War Gaming COA Development

COA Comparison COA Approval Orders Production

That's the skeleton. The muscle is what happens inside each step, and most people skip straight to war gaming because that's the part that looks impressive. It's also where plans go to die if you haven't done the work before it.

What Nobody Tells You About Step One and Step Two

Receipt of Mission sounds simple. You get the order, you read it, you acknowledge it. The real work starts the moment you open the fragmentary order and realize the commander's intent is written in vague language. "Seize the high ground" means nothing if you don't know what the high ground actually is or why the commander wants it specifically at 0600 instead of 1400. During mission analysis, you're supposed to identify the operative factors, constraints, and decisive points. What actually happens is your S3 is pulling five different sources and they contradict each other. The intelligence estimate says one thing, the surveillance report says another, and the local liaison officer says the terrain has changed because of recent flooding. You spend two hours just figuring out what "ground truth" actually means for that sector. One trick that saves time: start a running list of unresolved questions the moment the order arrives. Don't wait for the formal mission analysis brief. Write them down as you read. By the time you get to the actual analysis phase, half of them are already answered. In one operation I was involved with, we identified 34 open questions during the initial read-through. Twenty-one were resolved within the first three hours through quick coordination calls. That cut our mission analysis from a six-hour session down to about forty minutes of actual productive work.

War Gaming Is Where Most Plans Break

War gaming gets treated like a tabletop exercise where everyone plays hero. It should be the opposite. You're trying to find the places where your plan fails, not where it succeeds. The standard technique is branch and sequent analysis. You run the main effort through its expected phases, then for each critical decision point you ask what happens if something goes wrong. Not what would happen in a perfect world. What actually goes wrong. The common mistake is war gaming only the favorable scenario. You paint the enemy as passive and predictable. Your maneuver elements move exactly when planned. Supply routes stay open. Weather cooperates. This produces a plan that works in a textbook and fails immediately in the field. I once watched a staff war game run for four hours and come back with a plan that assumed zero communications degradation. When we deployed, the primary net went down within the first hour and the entire sequence collapsed because no one had planned around that specific failure mode. What I do now is force at least one adverse branch at every phase line. Enemy conducts an ambush during phase one. Logistics cannot resupply during phase two. A key leader becomes unavailable during phase three. You don't need elaborate scripts. Just establish one realistic disruption per phase and see whether the plan still holds.

Course of Action Development and the Approval Problem

After war gaming, you develop courses of action. Doctrine says you should produce at least two but no more than three. The reality is your staff will produce whatever they think the commander wants to hear. If the commander has already signaled a preference during mission analysis, you're not developing COAs anymore. You're validating a decision that was made three hours ago. COA comparison is where you apply the criteria the commander established. Speed, risk, feasibility, acceptability. The problem is these criteria are usually written in general terms. "Acceptable risk" means different things to different people. I learned to force the commander to rank them explicitly before comparison begins. Not choose between acceptable and unacceptable. Rank them one through five. This small change eliminates about thirty minutes of argument during the comparison phase and produces a much clearer result.

What the Process Cannot Handle

The Military Decision Making Process assumes a certain amount of time. Five to eight hours for a full run-through against a relatively stable threat. It assumes you have a complete staff. It assumes the battlefield environment doesn't change drastically during the planning cycle. None of these assumptions hold in fast-moving operations or when you're working with a truncated team. When I've had to run MDMP with only three people and a twelve-hour window, I compress the process aggressively. Mission analysis becomes a focused discussion lasting thirty minutes maximum. War gaming drops to two branches instead of five. COA development happens in parallel rather than sequentially. The output is less polished but it's viable. The alternative to skipping steps is missing the timeline entirely and producing a perfect plan for a battle that already happened. There's also the issue of information turnover. In contemporary operations, the intelligence picture can shift significantly between the time you finish planning and the time you execute. A drone feed showing a route clear at 1800 may show it contested by 2200. MDMP produces a plan based on the best available information at a specific moment in time. That information ages poorly. The workaround is building checkpoints into the plan itself. Brief decision points where the commander evaluates whether to commit, adjust, or abort based on current conditions rather than conditions from six hours ago.

Practical Workflow for a Functional Planning Cycle

Here's how I structure it when I'm running the process under normal conditions with a full staff and about six hours available. Hour zero to one: Receipt and initial briefing. Distribute the order. Identify the critical tasks. Establish the operating timeline with the commander. This timeline is your most important output at this stage because every decision after this point depends on it. A wrong timeline ruins the entire process. Hour one to three: Mission analysis. Work through the factors systematically. Identify the center of gravity. Pinpoint decisive points. Resolve as many open questions as possible. Produce the commander's critical information requirements list. This list drives the rest of the planning process.

Hour three to four thirty: War gaming. Run the approved COA through each phase. Identify branches and sequels. Document where the plan is fragile. This should be uncomfortable. If everyone agrees the plan is solid, you didn't war game it correctly. Hour four thirty to five thirty: COA development and comparison. Write up the remaining courses of action. Apply the commander's weighted criteria. Produce a recommendation with clear reasoning. Hour five thirty to six thirty: Approval and orders production. The commander makes the decision. You begin producing the operation order using the approved format. This is not the time to redesign the plan. This is the time to translate it into executable documents.

The total time varies based on complexity. A division-level attack with multiple supporting operations might require two days. A battalion-level raid might take three hours if you compress aggressively. The process is scalable, but the quality of the output depends on how honestly you apply it rather than how thoroughly you follow the format.

Documentation and the After-Action Reality

Every MDMP run should produce a document package. At minimum, the operation order, the intelligence estimate summary, and the war game results with identified branches. I also keep a planning journal tracking every major decision point and the reasoning behind it. When we debrief afterward, that journal is the single most useful artifact for understanding why certain choices were made and what assumptions proved incorrect. The most common failure point I see is insufficient focus on sustainment. Every planning cycle I've observed puts disproportionate attention on the Maneuver element and the scheme of fires. Logistics gets a paragraph. Medical gets a paragraph. Communications gets a paragraph that assumes everything works. The plan fails later not because the tactical concept was wrong but because the resupply schedule had no contingency for degraded routes or the casualty evacuation timeline didn't account for contested airspace. Building sustainment constraints into the war game early prevents this. Not as an afterthought. As a primary input. Another thing worth noting is the difference between an expeditionary order and a deliberate operation order. The military decision making process is designed for deliberate operations. When you're in an expeditionary context with dynamic conditions, you fall back to a lighter version. The key steps remain the same but the time allocation shifts dramatically. Mission analysis takes twenty minutes. War gaming might not happen at all. The commander makes a decision based on the situation report and the staff executes. This is normal and acceptable. You're trading depth for speed, and in some situations that trade is the correct one.

What to Look for When Evaluating Your Own Process

If you're running this process and want to assess whether it's working, check a few specific indicators. Are the open questions from mission analysis being tracked and resolved? Is there a clear record of why one COA was selected over others? Do the branches identified during war gaming have assigned responsibilities and timelines? Is the sustainment plan integrated into the scheme of maneuver or appended as separate documentation? Are decision points explicitly stated in the order? If the answer to any of these is no, the plan has gaps that will surface during execution. Not necessarily immediately. Maybe not even in the first phase. But they will surface. The Military Decision Making Process is a tool for reducing uncertainty, not eliminating it. The goal is to make better decisions with worse information faster than the opposing force can respond. If your planning cycle produces a document that looks good but hasn't been stress-tested against realistic failure scenarios, it's a document. Not a plan.