The Prediction

A FOIA request is pending for completion data from a high-load stenography training program.

Before that data arrives, I’m publishing what the math predicts:

Completion rate: 5-12%

Attrition rate: 88-95%

If completion exceeds 25%, my model is wrong. I will say so publicly.

Today is January 3, 2026.

The Framework

For 140 years, stenography has had no measurement framework. No way to quantify what a method demands from the human brain. No way to predict outcomes. No way to compare systems objectively.

I built one.

Total Effort Cost (TEC) quantifies cognitive and physical load:

TEC = (SDS + DLS) × CEF × PFF

Each of these formulas and what they measure have been thoroughly explained in previous articles.

The variables measure stroke difficulty, decision load, context pressure, and prediction failure. They’re grounded in established cognitive science—working memory limits, motor learning theory, decision fatigue research.

The formulas synthesize existing research. They don’t invent new physics.

The Architecture

The program relies on thousands of arbitrary briefs, phrase-based prediction under real-time pressure, and speed-first progression before accuracy stabilizes. When retrieval fails, there’s no phonetic fallback.

Traditional stenography methods require approximately 303% of sustainable human cognitive capacity. They produce 85-92% attrition.

This program adds load on top of that baseline. Higher retrieval burden. Higher conflict density. Higher failure cost when predictions miss.

Estimated capacity requirement: 350-400%+

The Math

Additional load produces additional failure.

If 303% capacity produces 85-92% attrition, then 350-400%+ cannot produce better outcomes. It must produce worse.

Working memory holds 4-7 chunks. Decision friction compounds under pressure. Error reinforcement locks in wrong pathways. Fatigue degrades retrieval.

These are biological constraints. Design either respects them or humans fail.

What the Data Will Show

5-10% completion: Framework validated. Architecture unsustainable for mass adoption.

10-15% completion: Partial mitigation—extreme self-selection or undisclosed scaffolding. Still worse than lower-load alternatives.

15-25% completion: Model needs adjustment. Something is mitigating load.

Above 25% completion: Model is wrong. I will say so publicly and investigate.

A Note on Data Integrity

FOIA data comes from the institution being measured. Unflattering numbers have a way of not arriving.

Outright falsification, however, is risky—records submitted to accrediting bodies carry legal weight. More likely: definitional games. Counting partial completers. Excluding early withdrawals. Reporting only aid recipients.

Even manipulated numbers tell a story. If reported completion is 30-40% but the math predicts 5-12%, something doesn’t add up. Either the load measurements are wrong or the numbers are manipulated.

Both are investigable. Both are findings.

The math creates a constraint. Numbers that violate it demand explanation.

Why Publish First

Anyone can explain numbers after seeing them.

Prediction before data is how science establishes credibility. The framework is public. The math is public. The prediction is public. When the data arrives, we’ll all see it together.

If the data matches, load predicts outcomes. If the data contradicts, I learn something.

Either way, the prediction can’t be adjusted after the fact.

The Implication

High attrition is the system exceeding what humans can sustain.

The architecture produces the outcome. The math predicted it.

Now we wait for the denominator.