AI-NATIVE SERIES · SHEPHERD'S ACADEMY OF WAR, PART 5 OF 5

A Scorecard for a Family Academy: One Car-Ride Drill, Four Numbers, Six Tripwires

By Paul Jialiang Wu · agentic-portfolio-lovat.vercel.app · 2026-10-12 · 11 min read

This piece does not answer: What the name means — that is part 1. What the last battle asks of a child — that is part 2. The evidence for each practice — that is part 3. Who may teach and how — that is part 4.

How to read the labels: EVIDENCE is sourced, DESIGN is how SAW is built, OPEN is not yet tested, and LENS is my own conviction.

A note before we start: these are my own views, not my employer's. I'm a Christian father, not a clinician or a researcher, and nothing here is medical advice. Verses are quoted from the World English Bible (WEB), which is public domain, as bible-api.com returns them; a translated edition of this article translates the WEB.

What this part covers: a scorecard for a three-week-old academy, and who it is for

This part is for parents and leaders who run anything meant to build character and want to know whether it works. It gives one drill, four numbers to count, and six conditions, written in advance, that would make us change course.

Four parts of this series made claims. This one says how I will find out whether they were true.

Shepherd's Academy of War is three weeks old. It has two students, my sons, who are 11 and 7. It has no confirmed instructor and no building, and nothing has been measured yet. So this scorecard is empty today, on purpose and in public. An empty scorecard with dates on it is more honest than a full brochure.

I suspect, about my own project, that enthusiasm is being counted as evidence.

What does run today is the field lab this grows out of, at daniel-and-david.vercel.app. Its free Builder Loop app takes a family through five short build-show-learn cycles and keeps every note on your own device. Its measure, which I am borrowing, is what a child can show they did without being pushed (theory of change). The SAW drill below is not in the app. It is an index card and a car.

The picture: a scoreboard that can show a loss

No extended analogy. One test: can this number go down, and would I show you if it did?

Why measure a family project at all

EVIDENCE Because larger and better-funded people than me skipped it. The U.S. Army gave resilience training to its soldiers under a program called Comprehensive Soldier Fitness. Critics writing in the field's own journal described it as a 125-million-dollar initiative launched without pilot testing (Eidelson, Pilisuk & Soldz, 2011). Two years later, reviewers noted it had reached more than a million soldiers and had still not been independently and objectively reviewed (Steenkamp, Nash & Litz, 2013).

LENS If an army with a research budget can roll out character training to a million people before testing it, a father in a car is certain to. I am the least objective evaluator this academy will ever have. I named it, and I love the students. So the test has to be written before I look at the results.

The drill

Here is the drill as I have designed it. We have not yet run it as written. It takes twenty minutes and fits a drive a family is making anyway, which is how the oldest instruction on the subject says to do it:

These words, which I command you today, shall be on your heart; and you shall teach them diligently to your children, and shall talk of them when you sit in your house, and when you walk by the way, and when you lie down, and when you rise up.
— Deuteronomy 6:6-7 (WEB, World English Bible)

"When you walk by the way." We drive by the way. Close enough.

DESIGN The drill has five moves.

1. Report (5 minutes). What happened with last week's rescue? What did it cost? This comes first because it is the only part that matters.
2. Story (3 minutes). One brave person, told short. A verse, a relative, someone in the news.
3. Question (5 minutes). Who was protected? What did it cost? And one from part 1: how would the attack have worked? You cannot guard against what you do not understand.
4. Rehearse (7 minutes). One small hard thing, practised safely. Said out loud, acted out, or planned for when we arrive.
5. Rescue (the week between). One real act of help, chosen by the child.

DESIGN The arena rotates, one a week, through the five from part 2. This is a plan. We have not yet run a full rotation.

Every row keeps the laws from part 1, and each is sized for the child: a seven-year-old watches and names; an eleven-year-old does. The attack is studied so it can be stopped. A fake message is written by the adult, on paper, and never sent. No row has a person as the enemy.

Four numbers

DESIGN One index card per child per month, four tallies.

1. Rescues. Acts of help the child chose, each with what it cost.
2. Refusals and rebuilds. Times the child decided beforehand not to do something, and didn't, or questioned a rule the right way and offered a better one. The Daniel count.
3. Fears faced. Rungs attempted on the Brave Ladder from part 3, chosen by the child.
4. Free-choice minutes. Time the child spends on any of this when nobody asked. The play series explains why this is the one I trust most: it is what a child does when no one is watching.

And one number for me: how many of the four tallies I filled in for them. The target is zero.

OPEN Today all four read zero, because counting starts with this article. On 11 January 2027 I will publish what the cards say, including if they say nothing happened. Sample size: two. And the experimenter hands out the snacks, which is one more reason there are no prizes.

One more rule, because of the word "war"

A program that uses the word war owes the public one rule in writing.

DESIGN Illness is never a spiritual verdict. See a doctor, and pray. Both. An illness of body or mind is never described as a spiritual failure, or as anything other than an illness. The academy teaches children to stand against fear and lies. It does not diagnose anyone.

EVIDENCE Two of the largest church bodies have it in writing. In 2013 the Southern Baptist Convention resolved: "we support the wise use of medical intervention for mental health concerns when appropriate" (SBC, 2013). The Catholic Catechism, in the very paragraph about its rite against evil, says: "Illness, especially psychological illness, is a very different matter; treating this is the concern of medical science." (Catechism 1673).

EVIDENCE The common failure is silence. In a 2014 survey, 66% of Protestant senior pastors said they seldom speak to their congregation about mental illness (Lifeway Research, 2014). LENS Silence leaves children to fill the gap with their own explanations. So any program like this one should say it plainly, in words a seven-year-old can hold: being sick is not being bad, doctors are a gift, and we pray too.

Five whys: why do character programs so rarely find out whether they work?

1. Why don't they find out? Because they count what is easy: attendance, smiles, testimonials. Evidence: my own brochure column above; I wrote every line of it first.
2. Why count the easy things? Because the program was launched before anyone decided what failing would look like. Evidence: a 125-million-dollar program with no pilot (Eidelson et al., 2011).
3. Why launch before deciding? Because the people launching it already believe. Evidence: more than a million soldiers trained before an independent review (Steenkamp et al., 2013).
4. Why does belief block measurement? Because a believer reads every result as support. A good week proves it works; a bad week proves they need more of it. Evidence: one father, told as a story.
5. Why does anything break that loop? Because a condition written before the result cannot be reinterpreted after it. Evidence: the MYRIAD trial in part 3 ran at scale, found no average benefit and signs of harm for high-risk students, and said so (MYRIAD, 2022). The root: you only learn whether a program for children works if you write down, before you look, what would make you change it or stop.

Operate: the thing that runs

In one sentence: the date on which I owe you numbers is written into this page, and the page is checked. DESIGN The re-check date, 11 January 2027, sits in this article's TRUE section, and `scripts/check-saw.mjs` fails the build if any part of the series lacks a dated re-check. The same script refuses the series if it stops stating its honest status, or claims an instructor it does not have. The drill itself is a card. I would like it to run in the field lab's app one day; that is a build, and this series comes before any build.

Feedback: the numbers, and what they cannot see

DESIGN Four tallies a month per child, read against last month's. Free-choice minutes is the lead signal. If it is zero, the others are probably me.

OPEN What the card cannot see is most of what matters. It cannot see who a child is at fourteen. It cannot see what they do when they are alone, which is the whole point of the Daniel house. It cannot separate the drill from an ordinary good month. And with two children there is no comparison at all. Readers' cards in the comments are the only way this becomes more than a family story, and even then it will be a collection of stories. I will describe it that way.

Feedforward: six tripwires, decided now

DESIGN What I commit to before the results exist.

1. A person in the enemy slot. If either child names a person or a group as the enemy and I trace it to our language, the war vocabulary stops that week.
2. Fear. Dread before the drill, bad dreams, or anxious repeated praying: the frightening content stops that day, and I ask a paediatrician and a pastor for help.
3. No fuel. Free-choice minutes at zero for three weeks running means this is my project and not theirs. I hand them the design and wait.
4. An injury. Anything beyond a scrape, or any contact to the head: the body arena pauses until someone qualified has looked at what we were doing.
5. A missed date. If I cannot show the cards on 11 January 2027, I say so here. A missing result is a result.
6. A verdict. If illness ever gets spiritual-blame language in this academy, or the drill is ever used in place of medical care, it stops and the correction is made in the open.

Where this part sits in the series

Same story, five exits

Build it in 30 minutes

• Input: one weekly journey, one child, one index card, one pen.
• Output: three signed tripwires on the back; four empty tallies on the front.
• Invariant: the child keeps the card. No prizes. No person as the enemy. Illness is never a spiritual verdict.
• Stop rule: any tripwire. Read them before the tallies.
• Minimal record: `month, rescues, refusals, fears_faced, free_choice_minutes, tripwire_fired`.
• Deliberately missing: ranks, a leaderboard, and a comparison between siblings.

TRUE

• Transferable. Any program meant to change people (onboarding, a leadership course, a small group) can use the same card: one behaviour that is chosen freely, counted by the person, with stop conditions signed in advance.
• Recursive. This article is wrong if the four numbers cannot tell a good month from a bad one, or if a tripwire fires. Re-check on 2027-01-11, here, with the cards.
• Understandable. In one sentence: the plan is one short drill a week, the kids count four things on a card, and we wrote down in advance what would make us stop. The move it asks of you is to prototype: run the small version before you believe in the big one.
• Experience it. Open the Builder Loop app at daniel-and-david.vercel.app to see the build-show-learn cycle this borrows, then write three tripwires on a card. What would prove me wrong: January's cards.
• Built. What runs: the field lab's app, and the dated check on this page. The drill is a design on a card, in an academy three weeks old.
• Taught. Have your child explain the four tallies to another adult and decide, themselves, whether something counted as a rescue. If they overrule you, good.

5S self-score, with evidence

Sources

• Eidelson, Pilisuk & Soldz (2011), critique of Comprehensive Soldier Fitness, American Psychologist: https://pubmed.ncbi.nlm.nih.gov/21967209/
• Steenkamp, Nash & Litz (2013), review of Comprehensive Soldier Fitness: https://pubmed.ncbi.nlm.nih.gov/23597815/
• Southern Baptist Convention (2013), resolution "On Mental Health Concerns and the Heart of God": https://www.sbc.net/resource-library/resolutions/on-mental-health-concerns-and-the-heart-of-god/
• Catechism of the Catholic Church, paragraph 1673: https://www.vatican.va/archive/ENG0015/__P58.HTM
• Lifeway Research (2014), pastors and mental illness: https://research.lifeway.com/2014/09/22/mental-illness-remains-taboo-topic-for-many-pastors/
• MYRIAD trial (2022): https://pmc.ncbi.nlm.nih.gov/articles/PMC9340034
• Daniel & David field lab, theory of change and North Star measure: https://github.com/wjlgatech/daniel-and-david/blob/main/docs/vision/theory-of-change.md
• World English Bible via bible-api.com (Deuteronomy 6:6–7): https://bible-api.com/
• The research dossier behind this series: https://github.com/wjlgatech/daniel-and-david/blob/main/docs/research/shepherd-warrior-formation.md

Related

• Part 1 — An Academy of War for Children: Learn the Attack, Live as a Protector
• Part 2 — Revelation 19 Has One Fighter, and the Long War Has Five Arenas: What a Child Trains For
• Part 3 — Courage Training for Children: What the Evidence Supports and What Backfires
• Part 4 — Guest Instructors for Children: Content, Not Contact