diff --git a/docs/scenarios/CLEAN_GROUND.md b/docs/scenarios/CLEAN_GROUND.md index 88ce9db..1aa6b01 100644 --- a/docs/scenarios/CLEAN_GROUND.md +++ b/docs/scenarios/CLEAN_GROUND.md @@ -406,6 +406,9 @@ belongs to nobody in Britain, and dusk comes anyway. - The six **hollow men** walk at the back. The column is kind to them and does not talk about them. `[CUS: Spot — the same six at the back, and the same six yesterday]` - Somewhere out on the flank, something is keeping pace at a distance and on all fours. +- **If violence against the column is even possible at your table, read EXPOSURE in Act + Four first.** Six of them wipe one party in twelve; at four players three of them wipe + two in five. This is the dangerous encounter in the scenario and it is in this act. **Failure case:** if the party refuses contact, the column camps within sight and one of the children is sent over with bread. That works on almost everybody. @@ -453,6 +456,79 @@ requisition number on it and that number is enough to work the rest out from the - **If a PC is taken:** hand that player the understudy sheet and a private note. They keep playing. Do not sideline them and do not let it run more than twenty minutes. +### EXPOSURE — what a fight actually costs. Read before Act Two, not before Act Four. + +**The finding that matters: the understudy cannot hurt them, and the refugees can wipe +them.** That is the scenario's whole thesis, and it is not a claim — it is measured. A GM +who does not know it will pad Act Four to make it frightening and will be padding the one +encounter in the case that is incapable of being dangerous. + +**MEASURED.** `tools/simulate.mjs`, 2000 runs each, seed 11, against the cast of six — +Ashcroft, Bhattacharya, Renshaw, Braithwaite, Pollard, Okonkwo. Taken at commit 5450521 +with a clean tree, so the same command reproduces the same figures: + +``` +node tools/simulate.mjs --creature the_understudy \ + --party ashcroft,bhattacharya,renshaw,braithwaite,pollard,okonkwo --runs 2000 --seed 11 +``` + +| What they fight | Hurt | Down | Wiped | Deaths/run | Median rounds | +|---|---|---|---|---|---| +| **The understudy, alone** | 0.48 of 6 | 0.01 | **0.0%** | 0.00 | **2** | +| A duplicate of an agent | 0.45 of 6 | 0.05 | 0.0% | 0.01 | 2 | +| One hollow man | 0.20 of 6 | 0.00 | 0.0% | 0.00 | 2 | +| Three hollow men | 1.15 of 6 | 0.00 | 0.0% | 0.00 | 6 | +| **Six hollow men** | 2.79 of 6 | 0.15 | **0.0%** | 0.01 | **12** | +| Three of the column | 1.74 of 6 | 0.30 | 0.0% | 0.11 | 5 | +| **Six of the column** | 4.54 of 6 | 2.38 | **8.6%** | 1.18 | 13 | +| **Ten of the column** | 5.94 of 6 | 5.69 | **83.5%** | 2.52 | 15 | +| Fifteen of the column | 6.00 of 6 | 6.00 | **100.0%** | 2.53 | 8 | + +**Read these as a floor.** The harness does not model bleeding, Panic, Coherence, cover, +range bands or fire modes, and every one of those makes a fight worse for whoever is +losing. Both of the first two are live in this scenario. + +#### The four things a GM needs from that table + +1. **Do not fix the understudy.** It is a 1d3 punch with a 60% chance to land, and it goes + down in two rounds. That is correct and it is the point: the horror is that it passed + every test the party had, not that it is hard to kill. A GM who buffs it into a threat + has replaced a scenario about classification with a scenario about a monster, and the + case stops meaning anything. +2. **Adding hollow men buys rounds, not danger.** Six of them still wipe nobody — and + take a **median twelve rounds** to resolve. In a 45-minute act that is the whole act, + and the Close is what pays for it. If Act Four feels thin, the answer is more doubt, + not more bodies. +3. **The column is the lethal encounter in this scenario, and it is not in Act Four.** + Three of them is survivable. **Six wipes one party in twelve. Ten wipes five parties in + six.** The curve between six and ten is almost vertical and a GM improvising bodies + will cross it without noticing. This is why Act Two says the column must never become a + threat: not only because shooting refugees is the moral catastrophe of the case, but + because the table will lose. +4. **The four-player cut is a different game.** Without Okonkwo's rifle and breach suit + and without Pollard's shield, **three of the column wipe two parties in five** and six + wipe virtually all of them: + + | Four-player cut fights | Hurt | Down | Wiped | Deaths/run | + |---|---|---|---|---| + | Three of the column | 3.24 of 4 | 2.09 | **40.2%** | 0.85 | + | Six of the column | 4.00 of 4 | 3.98 | **99.1%** | 1.45 | + + At four players, **any** violence against the column is a party-ending decision. Say + the row out loud before they commit, exactly as THROUGH TRAIN does. + +#### What to do instead of a fight + +- **Act Four's tension is a test, not a threat.** The question is which of two people is + filed, and the clock is the rehearsal counter, not hit points. +- **If the table insists on shooting the understudy, let them, and let it be easy.** Then + hand them the consequence: they have killed something wearing Sgt Prichard's face, the + entry is still filed, and something over there is already practising. That lands harder + than any wound. +- **If they fire on the column, do not soften it.** Use three, resolve it honestly, and + let the survivors keep walking north past them. Nobody at the Bureau will ask. + + **Failure case:** if the party kills it clean in the first ten minutes, it does not matter — the entry is still filed, and the second one is already practising. Go to Close. @@ -686,8 +762,8 @@ twice.* 2. ~~The read-aloud text.~~ **Done** — three blocks, in place in Act Two and the Close. 3. ~~Handouts.~~ **Done** — four sheets, CG-H01 to CG-H03b. Still need a print-ready HTML pack (A4, 11pt minimum) before a convention; the text is final. -4. **A written exposure model for Act Four**, so the GM knows what a fight at the cordon - actually costs before the table starts one. +4. ~~A written exposure model for Act Four.~~ **Done** — measured, 2000 runs a row at seed + 11 against the named cast, reproducible from the command in the section. 5. **Desk playtest**, twice, before any of this is called runnable. ## Open questions for Tim