# Strategy Guide — Scenario 63: "Cancelled, Again"

*Game documentation for a fictional simulation in "The First 90 Days" (Tactic Watch). The family, the case, and every choice below are fictional. This guide is general information about how the scenario is built — it is not legal advice for any real case.*

## (1) The win

**Best ending: "Ending — Sixty Percent, Corrected."**

State it honestly: this is a **partial** win, and it is the best this scenario offers. The review adopts the corrected attendance table on the record — 12 scheduled, 3 cancelled by the center, 9 held, 9 attended — the "family attended 60% of scheduled visits" finding comes out, and the three lost visits are ordered made up, with dates. Felix does **not** come home at the end of this ending. He remains placed, visits still run Tuesdays and Saturdays at the center, and the case continues. The win is that the finding every future summary will quote now describes the parent Simone actually is — and that three real afternoons with her son, which the old number had silently deleted, exist again on a calendar. There is no ending in this scenario where the review itself reunifies the family. Anyone selling you one is selling you a different game.

## (2) The route

Verified by walking the scenario graph from the start node (see the verification note at the end). Key decision points, in order — node title, winning choice, and what it changes:

1. **Tuesdays and Saturdays** — the premise: the center confirms visits by phone call, so every confirmation evaporates. The scenario is about what that evaporation costs, two months later, in a single percentage.
2. **The Day Before** — *Start confirming in writing — a short message the day before every visit, replies saved.* Sets **confirmWritten**; hidden item: visits confirmed in writing from the first week. The center's own two-word replies become the first layer of the eventual correction.
3. **Two Words, Dated** — the thread begins. Dull correspondence; load-bearing paper.
4. **The Good Weeks** and **The Drawing, Panel Four** — visits held and attended; Ms. Kessler calls Simone "one of my reliables," a sentence that appears nowhere, because praise in this building is spoken and counts are written.
5. **Cancelled, the First Time** — *Answer the same day in writing, requesting a make-up date.* Sets **cancelTrail**; hidden item: center cancelled by phone, morning of; parent answered in writing the same day. The reply — "when there's a slot," no date — still proves who cancelled and that she asked.
6. **A Reply with No Date in It** — the promise, examined: it confirms the cancellation was the center's and confirms nothing else. Saved anyway. Percentages are computed from less.
7. **A Held Visit** — attended, as every held visit is attended. The file cannot see the room; it sees counts kept by the party that cancels.
8. **Chasing the Make-Up** — *Follow up in writing that evening — second request for a make-up date.* Re-sets **cancelTrail**; hidden item: the center replied "when there's a slot," no date given. The asking, twice, with dates on it, is the trail.
9. **Cancelled, Again** — *Answer in writing the same day, as before.* **cancelTrail** again; the second cancellation joins the first in writing, not in memory.
10. **Sixty-Something Percent** — the foster parent repeats what she heard on a call: Simone's attendance described as "sixty-something percent." The denominator problem announces itself: Simone has attended every visit ever held, so someone else owns the bottom half of the fraction.
11. **Halfway There** — the third cancellation, on a Saturday, with Simone already on the bus. *Text the center from the bus stop, in writing.* Cash −7 (the spent fare); **cancelTrail** set a third time; hidden item notes she was already travelling when the center cancelled.
12. **The Long Way Back** — cash −6 (the return fare). Summaries assign owners to unowned events, and the nearest name in an attendance paragraph is always the parent's.
13. **The Count** — the kitchen-table arithmetic for the report window (the first Tuesday visit through the third cancellation): 12 scheduled, 3 cancelled by the center, 9 held, 9 attended. Nine of twelve is seventy-five percent — so even the center's cancellations, counted against her, do not reach "sixty." Something else is in their denominator.
14. **Nine of Nine** — her count is perfect and worthless in the same breath: it is *hers*. What pierces the agency's number is the agency's own paper.
15. **Their Number, Their Paper** — *Sign and send the written request — attendance summary and scheduling log.* Cash −4. Sets **recordsRequested** (55 Pa. Code § 3130.44(d); counsel throughout, 42 Pa.C.S. § 6337).
16. **Fifteen** — the packet explains everything: the summary counts 15 "scheduled" — the real 12 plus 3 make-up slots the center logged but never gave her dates for, from which she was marked absent — and folds the center's own 3 cancellations into her column. Nine of nine held visits became 60% of fifteen. Every step of the transformation is on *their* paper, which is what makes it answerable.
17. **The Corrected Table** — *Build the corrected table with your lawyer and attach it to the review packet.* Cash −5. Sets **mathAttached**: four columns — scheduled, cancelled by the center, held, attended — every row dated, confirmations and replies attached, the center's own log lines quoted.
18. **Four Columns** — the table, admired briefly for what it is: the other side's arithmetic, corrected out of the other side's own file.
19. **Sixty Percent, in Print** and **Panel Ten** — the review report quotes the 60% line, hardening by quotation — the table's last chance to be lying next to it when the room convenes.
20. **The Review** — *Present the corrected table, with the confirmations, the same-day cancellation replies, and the center's own summary and log attached.* This choice is flag-gated: it is only available with **confirmWritten** + **cancelTrail** + **recordsRequested** + **mathAttached** all set. It leads to **Ending — Sixty Percent, Corrected**.

## (3) Why this wins — the law as it is supposed to work

Only principles from the scenario's verified set are stated here:

- **Agency records are obtainable on request** — 55 Pa. Code § 3130.44(d) ("shall be released upon request"). Steps 15–16 are that right, exercised in writing, and they are the hinge of the scenario: the percentage was computed from a summary and a scheduling log, and only the summary and log could show *how* — the phantom make-up slots and the folded cancellations are invisible from outside the file.
- **Counsel builds the correction out of the agency's own lines** (42 Pa.C.S. § 6337). The table cites no fact the center did not itself record; that is why a room can adopt it without deciding anyone lied. Arithmetic is not a legal doctrine, and this scenario invents none: the legal moves are the request, the filing, and the attachment. The math is just what those moves were *for*.
- **A framing note, kept record-dependent as Scenario 14 teaches:** attendance findings are permanency findings in embryo — they are quoted into plans, and plans are what later stages of a case are argued from, including, at the filing end, the federal timeline under which a child in care for 15 of the most recent 22 months changes the agency's filing duty unless an exception is documented (42 U.S.C. § 675(5)(E)). This scenario states no placement date and computes no such date. Its point is narrower and prior: exceptions and findings alike are record-dependent, and a record that misstates the family's attendance by folding the provider's cancellations into it is the kind of record those later arguments are lost on. Correcting it in October, at review, is the whole game.

## (4) The traps

1. **Keeping the phone system at "The Day Before."** Cost: no **confirmWritten**. The visits still happen; their confirmations evaporate. At **The Review** the full-table lead closes, because the table's first layer — the center's own dated replies — does not exist.
2. **Accepting "when there's a slot" verbally at "Cancelled, the First Time" (and again at "Cancelled, Again").** Cost: no **cancelTrail** — plus the bus fares (−6 each time) spent on visits that never occur. The cancellations become unowned events, and unowned events in an attendance paragraph acquire the parent's name. Without the trail, even the make-ups ask becomes a story instead of a paper claim.
3. **Skipping the record request at "Their Number, Their Paper."** Cost: no **recordsRequested**, no packet, no "Fifteen." Her table, if built at all, is built from her own paper — authored by the one person the room treats as an interested party — and the denominator is never exposed. The correction cannot be made out of the center's own lines if the center's lines were never obtained.
4. **Going in with the story at "The Corrected Table."** Cost: no **mathAttached**. The story is true, ordered, and believed the way stories are believed in rooms that run on exhibits. It reaches **Ending — Make-Ups, Eventually**: the lost visits are restored with dates — real hours with Felix — but the 60% finding stands uncorrected, because a percentage is a document and only another document corrects it.
5. **Saying the visits "mostly happened" at "The Review."** Cost: everything. The phrase concedes the number's framing; the finding hardens into the plan: **Ending — Attendance: Poor**.

## (5) What no strategy controls

The center's staffing. A fully staffed center never cancels, and this scenario never happens — the shortage is the weather, not a villain. Whether a reviewer corrects a percentage on a complete table or notes it and keeps the figure; reviewers differ, and some treat summaries as entitled to their conventions. Ms. Kessler's memory across forty families, the report writer's quoting hand, and luck all hold pieces of every ending here. The route above controls none of that. It controls only whether Simone's attendance exists in the form October reads: dated confirmations, dated answers, the provider's own log obtained on request, and the arithmetic attached. Outcomes in this system turn on staffing, habits, and luck as much as on any parent's choices — the strategy's whole claim is narrower: that a parent who attended nine of nine should not have to take anyone's word for it, including her own.

---
*Verification: this route was walked programmatically against `scenario-63.json` — each named choice was followed from the start node in order; every intermediate node and choice exists; every flag-gated choice's requirements were satisfied by flags set earlier on the walk; the walk terminates at the ending node titled "Ending — Sixty Percent, Corrected" after visiting 23 nodes, with $128 of the $150 starting cash remaining. Walked path: start -> confirm_choice -> written_start -> good_weeks -> second_saturday -> cancel1 -> cancel1_written -> held_visit2 -> makeup_chase -> cancel2 -> rumor_beat -> cancel3 -> bus_ride_home -> monday_after -> tally_night -> records_beat -> packet_arrives -> prep_beat -> table_built -> review_notice -> last_visits -> review -> end_corrected.*