In Special Forces, an 85 percent pass rate meant the instructors had to go back and look at the course. It meant the standard was too easy. That is not a typo. If almost everyone passed, something was wrong with the evaluation, not with the students.
At FLETC, the target is 99 percent. The system is built around it. If a student fails, there is retraining, a retest, additional instruction. There is no mechanism to send them home and wait for the next class. The economics do not allow it. So the institution builds scaffolding around every student until they clear the bar, and the bar stays exactly where it needs to be to make that possible.
I am not describing negligence. I am describing a system that has made a rational decision about what it can afford. The problem is that the decision gets made on behalf of the institution, not the officer.
If almost everyone passed, something was wrong with the evaluation. Not with the students.
The average student at FLETC receives roughly six hours of basic marksmanship instruction before live fire: stance, grip, drawstroke, the fundamentals. Then approximately twelve hours of live fire spread over three weeks, a pre-qual, and a qualification requiring 70 percent or better. About 1,000 rounds total. If they fail, four more hours of instruction and a second attempt. The pass rate runs above 98 percent.
In SFAUC, SFARTEAC, and SOT pipelines, the goal was 10,000 rounds in the first ten days. Every day started with a B8 target at 25 yards. A student needed a minimum score of 700 on those B8s just to enter the first shooting evaluation, before CQB ever began. The standard for instructors was harder still. Super Dave Harrington required instructors to shoot speed bulls: 10 rounds, 10 seconds, 25 yards, at 90 percent minimum. Not to qualify. To teach.
That gap is not a resource problem. It is a philosophy problem. And it has consequences that do not show up until years after graduation, in a different fiscal year, attributed to a different cause.
The lawsuit is cheaper. Not because anyone says that out loud. Because the math works out that way.
Force-on-force training is where this gets most visible. In the current federal law enforcement model, students suit up in roughly $700 worth of protective gear: helmet, gloves, groin protector, full coverage. They move from covered position to covered position. They are shot at with paintballs. They return fire with FX rounds at an instructor position behind glass. There is no pain compliance. Unless a paintball hits the face plate directly, most students do not know they have been hit.
That is not a training failure. That is the design.
When I ran force-on-force as an SFAUC instructor, we played OPFOR. We suited up the same as the students: glasses and a towel in the groin, standard duty gear. Both sides used FX. At close-quarters distances, FX rounds communicate something a paintball does not: consequence. I have former teammates who still carry sunburst-shaped blue marks from FX rounds taken to the face. They made a tactical mistake once. They did not make it again.
Pain modifies behavior. That is not a training philosophy. It is physiology. The mind needs a reference. Once it has one, the learning retention is permanent. The LE model has systematically removed that reference because the cost and liability of producing it cannot be justified inside the institution's current structure.
Fifteen years ago, FX was used across every training division at FLETC. The retreat from it was not sudden. It was incremental, driven by cost. FX rounds run about 50 cents each versus a penny for paintball, plus the added expense of the protective equipment itself and cleaning protocols after MRSA outbreaks. The decision to move toward blank-firing guns requiring only safety eyewear was not made by anyone who wanted training to get worse. It was made by people managing a budget.
The record is clean. The performance was temporary. The capability is unknown.
Here is what that budget decision actually purchases. By graduation, the institution has invested several hundred thousand dollars in each recruit: background investigations, security clearances, travel, housing, months of instruction. Multiply that across a department's entire force. That investment creates a structural interest in the outcome. Failing a student at the finish line is expensive in ways that are immediate, visible, and traceable to a specific decision by a specific person.
Fielding an officer who falls apart under real pressure is also expensive. But that cost arrives later. In a different budget cycle. Often attributed to a different cause. Almost never traced back to the training program that produced the officer.
Every officer and agent is taught early: every bullet has a lawsuit attached. That is true. What is also true is that if an officer acts within the scope of their authority, they are protected by law. The agency can still be sued, but government organizations absorb those costs, or settle, because the metrics that produced a bad outcome belong to the administration that set them, not the training budget that was managed three years earlier. The liability is diffuse. The training cost is immediate and concentrated. Run that as a basic cost-benefit analysis across an entire force, where the statistical odds of any individual officer being involved in a shooting incident are low, and the math comes out the same way every time. Not because anyone made a cynical decision. Because institutions respond to incentive structures, and this one has a clear incentive structure.
That math holds as long as nothing exceptional happens. The problem is that law enforcement failures are not linear events. They are low-frequency, high-impact events. One incident, handled poorly under pressure, does not behave like a routine settlement cost. It behaves like a multiplier. It draws national attention, political pressure, and institutional consequences that do not resemble the steady-state model the system was built around. The lawsuit math is not secret. It just does not get said out loud. And it does not account for the Black Swan.
I coordinated the Advanced Pistol program at FLETC for ten years. Across those classes, when officers came in from the field, we routinely spent two of the five days getting them back to a minimum skillset before we could begin actual instruction. These were credentialed officers, qualified on record, carrying weapons in public. Shooting is a perishable skill. Most officers shoot their absolute best on qualification day and erode from there. The qualification does not measure readiness. It measures performance on a clean range, rested and prepared, on a day everyone knew was coming.
The Crucible exists for a reason. When a person is sleep-deprived, hungry, and physically exhausted, they cannot hide who they are. The best SOF operators will perform at roughly 70 percent of their best range days under those conditions. Among the officers I have evaluated under stress conditions, performance without prior stress inoculation is often closer to 20 percent of their best range day. That is not a criticism of the individual. It is the predictable result of a system that has never required them to find out.
Stress inoculation is real. Once the mind has a reference for operating under genuine pressure, that reference is available when pressure arrives. The current training model does not build that reference. It builds a record.
The officers and agents who come through these programs are not the problem. They trained to the standard they were given. They performed to the level the institution measured. The record is clean. The performance was temporary. The capability is unknown. And somewhere down the line, in a situation that does not resemble anything on the qualification course, the gap between those three things becomes the determining factor in how the encounter ends.
That is a systemic failure, and it is one the system has quietly decided it can sustain. What changes it is not a single instructor, a single class, or a single reform proposal. What changes it is the moment an institution decides that verifying capability matters as much as documenting compliance. Some already have. More will. The question is what it takes to move from one side of that line to the other before the cost of not moving becomes the kind that cannot be absorbed.