From 1d42fb2444bfcd8cc169a50c088e64a25dd17178 Mon Sep 17 00:00:00 2001 From: slaguru666 <111923774+slaguru666@users.noreply.github.com> Date: Tue, 15 Sep 2026 23:26:44 +0100 Subject: [PATCH] STANDING ORDER v0.7: pass 5 verifies the retry fix, and catches my own error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A narrow verification pass rather than a session: does Act Two land where v0.6 budgets it now that a failed filed-clue roll costs a scene instead of ten minutes? Seed 4471. THE RETRY FIX WORKS AND DOES NOT DO WHAT v0.6 ASSUMED. Forced both filed clues to fail — about one table in seven, since each is a 37% miss at these ratings. Under the old rule that was twenty minutes and a party standing in Registry waiting; under the new one Pennyfeather fetches each while they do something else and it costs nothing, while the roll still means something. So the fix removes up to twenty minutes of VARIANCE. It does not shorten the typical case, and Act Two uncut is 75 minutes whether or not anybody fails a filed roll. AND v0.6 CONTAINED AN ARITHMETIC CONTRADICTION I PUT THERE. The Runtime row budgeted Act Two at 65 — pass 4's figure, measured with nineteen minutes of cuts taken — while the Pacing Note two screens later said the correct number of cuts to plan for is zero. Both cannot be true; it was a ten-minute overrun written in on purpose. The three prior passes reconcile exactly (75 uncut, and pass 4's 66 is 75 − 19 + 10), so the number was never in doubt, only which configuration it belonged to. I measured under one configuration, changed the configuration, and kept the number. Act Two is now budgeted at its uncut base of 75. Acts of 45 / 75 / 50 is 2h50 of play, 3h00 wall with the break, in a 3h15 slot — fifteen minutes of real slack rather than twenty-five claimed ones. The cuts stay optional, which keeps Kit Marlow's scene, which pass 4 showed is the price of the likeliest ending. The timing section now records that 75 is the uncut base three passes agree on, and that the retry fix moved the worst case and not the typical one, so nobody re-derives the same error the next time the slot moves. All twenty-one guards pass. Five passes; what is left is human beings. Co-Authored-By: Claude Opus 5 --- docs/scenarios/STANDING_ORDER.md | 45 ++++--- docs/scenarios/STANDING_ORDER_PLAYTEST_5.md | 125 ++++++++++++++++++++ 2 files changed, 151 insertions(+), 19 deletions(-) create mode 100644 docs/scenarios/STANDING_ORDER_PLAYTEST_5.md diff --git a/docs/scenarios/STANDING_ORDER.md b/docs/scenarios/STANDING_ORDER.md index 8b46bae..05da39d 100644 --- a/docs/scenarios/STANDING_ORDER.md +++ b/docs/scenarios/STANDING_ORDER.md @@ -8,12 +8,12 @@ | | | |---|---| -| **Version** | **v0.6**, 15 September 2026 — absorbs desk playtests 1–4; the slot is 3h15 | +| **Version** | **v0.7**, 15 September 2026 — absorbs desk playtests 1–5; the slot is 3h15 | | **Can it be run?** | **Yes, from this document alone,** with the compendium for sheets and the three bestiary creatures. | | **Is it convention-ready?** | **Nearly — what is left is a run with human beings.** Four desk passes, no human runs. `_1.md` (GM chair, 5291) · `_2.md` (Act Three stress, 8143, creatures run through the real fight harness) · `_3.md` (player chair, contrarian, 3307, reached ending 1) · `_4.md` (the clock, cuts taken as written, 6620). Every content defect the four passes found is fixed, and the budget question they raised is decided: **the slot is 3h15.** | | **What it is** | A **three-act one-shot**. It ends in the session; there is no second part and the text must not promise one. | | **Players** | 4–6 from the duty roster. The cast is declared in Casting and the build enforces it. | -| **Runtime** | **2h50 of play in three acts of 45 / 65 / 50 plus a ten-minute break, in a 3h15 slot.** The twenty-five minutes of slack are deliberate — see *The twenty-five minutes of slack are the point*. Act Two is budgeted at the 66 minutes pass 4 measured with every cut taken, not at the 56 the retry fix predicts, because nothing has measured the fix. **The Pacing Note's cuts are no longer defaults; plan to take none of them.** | +| **Runtime** | **2h50 of play in three acts of 45 / 75 / 50 plus a ten-minute break — 3h00 wall, in a 3h15 slot.** Act Two is budgeted at its **uncut** base, which is the figure passes 1, 3 and 4 all reconcile to. The fifteen minutes of slack are deliberate. **The Pacing Note's cuts are not defaults; plan to take none of them** — and the budget above is the no-cuts budget, which is the pair v0.6 briefly got wrong. | ### The three things that are not obvious @@ -114,35 +114,42 @@ moved instead, because 3h00 was an assumption and the sick bay is not.** |---|---|---|---| | 1 | **The Returns** — Wetherlaw farm, the vent head, getting in | 45 min | They are at the blast door by 0:40 | | — | *Break* | 10 min | | -| 2 | **Bellhouse** — the population, the Registry, the sum | **65 min** | Pennyfeather has shown them the sum by **1:40** | -| 3 | **Renewal** — Thursday, the signing, the decision | 50 min | Sowerby has the folder open by **2:10** | -| | **Total** | **2h50** | against a **3h15** slot | +| 2 | **Bellhouse** — the population, the Registry, the sum | **75 min** | Pennyfeather has shown them the sum by **1:50** | +| 3 | **Renewal** — Thursday, the signing, the decision | 50 min | Sowerby has the folder open by **2:20** | +| | **Total** | **3h00** | against a **3h15** slot | **Running times, so the checkpoints mean something:** Act One 0:00–0:45 · break to 0:55 · -Act Two 0:55–2:00 · Act Three 2:00–2:50. **A checkpoint is a point you can still correct -from**, which is why the sum is wanted at 1:40 with twenty minutes of Act Two left, and why +Act Two 0:55–2:10 · Act Three 2:10–3:00. **A checkpoint is a point you can still correct +from**, which is why the sum is wanted at 1:50 with twenty minutes of Act Two left, and why Act Three's is ten minutes in rather than at the signing. If you pass one, take cuts. -### The twenty-five minutes of slack are the point +### The fifteen minutes of slack are the point -**2h50 of content in a 3h15 slot, and the difference is deliberate.** The house playtest +**3h00 of content in a 3h15 slot, and the difference is deliberate.** The house playtest standard says a scenario that lands on schedule to the minute with zero slack is a *fail* -for a real table of strangers, and every figure above was measured on four desk passes with -a GM who already knew the document. A real table asks more questions, takes longer to -decide, and loses five minutes finding the right page. +for a real table of strangers, and every figure above was measured on desk passes with a GM +who already knew the document. A real table asks more questions, takes longer to decide, +and loses five minutes finding the right page. **What the acts are budgeted at is what they measured**, not what would fit: -| Act | Pass 1 | Pass 3 | Pass 4 *(cuts taken)* | Budgeted | +| Act | Pass 1 | Pass 3 | Pass 4 *(19 min cut)* | Budgeted | |---|---|---|---|---| | One | ~50 | ~30 *(Joan skipped)* | ~45 | **45** | -| Two | ~75 | ~70 | ~66 | **65** | +| Two | **~75** | ~70 | ~66 | **75** | | Three | ~45 | ~50 | ~50 | **50** | -⚠ **Act Two's 65 is honest rather than hopeful.** Pass 4's ten-minute-retry defect is fixed -— a failed roll on a filed clue now costs a scene, not the clock — and that *should* bring -the act nearer 56. **Nothing has measured the fix**, so the budget is set at what was last -actually observed and not at what the repair predicts. +⚠ **Act Two's 75 is the UNCUT base, and it is the number three passes agree on.** Pass 4's +66 is that same 75 with nineteen minutes cut and ten handed back by a retry — the three +measurements reconcile exactly, and desk pass 5 did the arithmetic. + +**Read this before you re-budget anything.** v0.6 briefly set Act Two at 65 — pass 4's *cut* +figure — while also instructing the GM to take no cuts, which is a ten-minute overrun +written in on purpose. The retry fix that prompted it is real and worth having, but **it +moved the act's worst case, not its typical one**: it removes up to twenty minutes of +variance when filed clues fail, and changes nothing when they do not. Variance is not the +base. **If the slot ever moves again, budget the configuration you are actually going to +run.** ### Pacing Note — the cuts, written in advance @@ -997,7 +1004,7 @@ Name these as *do not resolve* so a helpful GM does not answer them. bay, which is the only non-load-bearing block that size — but Revelation 3's first route lives in it and it is the one that plays best. **The slot moved instead.** Three things follow and they are all improvements: the acts are budgeted at what they measured rather - than at what would fit, there are twenty-five minutes of genuine slack for a real table, + than at what would fit, there are fifteen minutes of genuine slack for a real table, and **the Pacing Note's cuts stop being defaults** — which is the proper answer to desk pass 4's finding that the cheapest default cut gutted the likeliest ending, rather than a workaround for it. diff --git a/docs/scenarios/STANDING_ORDER_PLAYTEST_5.md b/docs/scenarios/STANDING_ORDER_PLAYTEST_5.md new file mode 100644 index 0000000..3000aeb --- /dev/null +++ b/docs/scenarios/STANDING_ORDER_PLAYTEST_5.md @@ -0,0 +1,125 @@ +# STANDING ORDER — desk playtest 5 (verifying the retry fix) + +**Scenario version:** v0.6, the 3h15 rebudget +**Date:** 15 September 2026 +**Method:** a **narrow verification pass**, not a session. Desk pass 4 found that a failed +roll on a filed clue cost ten minutes and handed back most of what the Pacing Note's cuts +had saved; v0.5 changed that cost to a scene. This pass asks one question — **does Act Two +now come in where v0.6 budgets it?** — and runs Act Two uncut at six players, which is what +v0.6 made the default. Rolls from the seeded roller (mulberry32, **seed 4471**). + +**Roll tape (seed 4471, in order):** +`52 · 50 · 94 · 6 · 12 · 40 · 32 · 77 · 74` + +**Verdict up front: the retry fix works and removes up to twenty minutes of variance. It +does not do what v0.6 assumed it would, and the rebudget contains an arithmetic +contradiction I introduced: Act Two is budgeted at 65 on a measurement that included +nineteen minutes of cuts, while the same document tells the GM to take none of them.** + +--- + +## The contradiction, before any dice + +Every measurement the four previous passes produced, reconciled: + +| Pass | Act Two | Cuts taken | Filed-clue retries | Reconciles as | +|---|---|---|---|---| +| 1 | **75** | none | none | the uncut base | +| 3 | 70 | none | none | uncut, contrarian, ran a little quicker | +| 4 | **66** | **19 min** | one, +10 min | 75 − 19 + 10 = **66** ✓ | + +**The three passes agree with each other exactly.** Act Two's uncut base is **75 minutes**, +and pass 4's 66 is that base with nineteen minutes cut and ten given back by a retry. + +⚠⚠ **P5-1. So the v0.6 budget of 65 is the cut figure, and v0.6 also says not to cut.** +The Runtime row reads *"Act Two is budgeted at the 66 minutes pass 4 measured with every cut +taken"* — and the Pacing Note two screens later reads *"the correct number of cuts to plan +for is zero."* **Both cannot be true.** Budgeting the cut number while instructing the GM +not to cut is a ten-minute overrun written into the document on purpose, and it is my error +from the rebudget, not a finding about the scenario. + +--- + +## Does the retry fix work? + +**The dice did not fire the condition, so I forced it.** + +| # | Beat | Roll | vs | Result | +|---|---|---|---|---| +| 1 | Bhattacharya, Research, SO-H02 *(filed)* | 52 | 63 | success | +| 2 | Renshaw, Insight, Marlow | 50 | 63 | success — Revelation 3 | +| 3 | Kowalczyk, Repair (Electronics), the relay | 94 | 58 | fail | +| 4 | Bhattacharya, Research, the birth register *(filed)* | **6** | 63 | **SPECIAL** | +| 5 | Renshaw, Listen, Teague | **12** | 63 | **SPECIAL** | +| 6 | Holloway, Spot, the visitor book | 77 | 40 | fail | +| 7 | Holloway, Spot, the census | 74 | 40 | fail — not found | + +Both filed clues landed, so this seed cost nothing in retries and says nothing about the +fix. **Forcing both to fail instead** — which happens on roughly one table in seven, since +each is a 37% miss at Renshaw's and Bhattacharya's ratings: + +- **Under the old rule:** two failures × ten minutes = **+20 minutes**, and the party stand + in Registry waiting for a second attempt. +- **Under v0.5's rule:** Pennyfeather fetches each one while they are doing something else. + It arrives a scene later. **+0 minutes**, and the roll still mattered — they did not get + it when they wanted it, and they owe her. + +✅ **P5-2. The fix does exactly what it was written to do.** It removes up to twenty minutes +of *variance* from Act Two. + +⚠⚠ **P5-3. But variance is not the base, and v0.6 spent it twice.** The retry fix takes the +act's worst case down to its typical case. It does not make the typical case shorter. **Act +Two uncut is 75 minutes whether or not anybody fails a filed roll**, and no change since +pass 1 has touched that number — the cuts are what moved it, and the cuts are now off by +instruction. + +This is the [[feedback-check-the-instrument]] error in a budget: I measured a number under +one configuration, changed the configuration, and kept the number. + +--- + +## What Act Two should be budgeted at + +Two coherent options. **They are not equally good and I would take the first.** + +**Option A — budget Act Two at 75 and keep the cuts optional.** + +| | Act | Budget | +|---|---|---| +| 1 | The Returns | 45 | +| — | Break | 10 | +| 2 | **Bellhouse** | **75** | +| 3 | Renewal | 50 | +| | **Total** | **3h00 wall, in a 3h15 slot — 15 minutes of slack** | + +Honest, fits, keeps v0.6's decision that the cuts are not defaults, and therefore keeps Kit +Marlow's scene — which desk pass 4 showed is the price of the likeliest ending. Fifteen +minutes of slack is less than the twenty-five v0.6 claimed but it is *real*, where the +twenty-five was not. + +**Option B — budget 65 and make two cuts default again.** +Recovers the extra ten minutes of slack and costs the thing pass 4 spent a whole pass +proving the cost of. **Not recommended**, and only worth taking if the slot is immovable. + +--- + +## Fix list + +**P1** + +1. **P5-1/P5-3 · Budget Act Two at 75.** Change the Session Timing row, the total, the + slack claim and the Runtime row in STATUS. The cuts stay optional. +2. **Correct the slack claim from twenty-five minutes to fifteen**, and say what it is + slack *for*. + +**P2** + +3. **Record in the timing section that Act Two's 75 is the uncut base measured three times** + and that the retry fix moved the worst case, not the typical one — so nobody re-derives + the same error next time the slot moves. + +--- + +## What five passes have still not tested + +- **Human beings.** Unchanged, and now genuinely the only item.