Probability Game (Part 2): Finding Order in Chaos on 祝融说。

Appendix: A Record for Sustainable Action

Sun, 13 Sep 2026 00:00:00 +0800

Begin with an Actual Action

This appendix supplies a recording structure, not ready-made answers. Select an action that will actually be taken or forgone, and state its object, deadline, authority, and observable result. A simple action can use a reduced form; expand it for actions with high consequences or multiple stages.

Mark blanks as unknown, not occurred, or not applicable. These differ: an unknown needs material, something not occurred cannot enter the actual ledger, and something not applicable does not belong to the object. Do not invent probabilities or costs to fill a table.

Every person, amount, labor period, device, and reading in D17 is fictional. The template may be transferred; the case thresholds may not be copied as rules.

Chapter 1: Why the Same Probability Can Lead to Different Choices

Sun, 13 Sep 2026 00:00:00 +0800

What Should We Do After Fifty Percent?

At nine on Saturday morning, Gu Ning reopened the R17 record. On Thursday evening she had reported 0.5 for the event that complete confirmation of Version 1 would arrive within the original window. The window ended at 17:00 on Friday, and the available record was sufficient to adjudicate that the event had not occurred. The client finally said that the original receipt period would not work and asked to discuss it again the following week.

Tang Ke asked whether they should now stop. Ye Cheng thought that because the client remained willing to talk, the chance of continuing the following week was not low, so they could reserve some time. Faced with the same record, one leaned toward stopping and the other toward continuing. It looked as if one of them had failed to understand what fifty percent meant.

Chapter 10: Why Multiple Paths Can Fail Together

Sun, 13 Sep 2026 00:00:00 +0800

Two Budgets Do Not Mean Two Independent Paths

At 17:30 Monday, Gu Ning placed G2's two estimated labor slots and two conditional units beside the six slots of existing work. The margin ledger proved that the quantities fit, but not whether the arrangements depended on the same person, device, or time entrance.

Separate budgets encourage people to imagine separate paths. In practice, one unavailable inspection bench may prevent both from being adjudicated, and an executor occupied by rework on the earlier job may cause the later one to miss its window. Separate project names do not create independent conditions.

This chapter maps dependencies before calculating correlation coefficients or joint-loss probabilities. Shared structure can be checked from schedules and resource records; chances of failure require different material. Declaring risk doubled from one shared node exceeds the evidence just as much as ignoring it.

Chapter 11: How Nominal Diversification Conceals Aligned Losses

Sun, 13 Sep 2026 00:00:00 +0800

Why Two Loss Rates of Twenty Percent Are Not Enough

At 18:30 Monday, Gu Ning drew a four-cell table beside the exposure map. If the station later knew the separate probabilities of adverse results for E4 and G2, could it review them one by one and say that two projects already provided diversification?

A marginal probability states how much space each event occupies across all scenarios. It does not say how the two occur together. Two twenty-percent losses may mostly alternate, occasionally overlap, or always be triggered together by one condition. Equal margins can expose a subject to very different concentrations in one realization.

This chapter first uses numbers wholly independent of Chengwan to demonstrate three joint structures, then returns to D17 with a blank-valued table. Demonstration probabilities, loss units, and repetition conditions do not enter E4 or G2. D17 lacks a sufficient batch and adopts no numerical joint probabilities.

Chapter 12: How the Capacity to Wait Changes Choice

Sun, 13 Sep 2026 00:00:00 +0800

Why a Subject with Margin Can Decide Later

At 19:30 Monday, Gu Ning faced three candidates: release G2's two formal slots immediately, wait for bounded node checks and release in segments, or stop V2 and return the first-payment balance. All three were describable from current records and preserved different later choices.

Waiting can bring information about M7 quantity, C1 state, and E4 progress, reducing reliance on guesses. It also consumes Tuesday preparation time and makes Wednesday's window harder to meet. Waiting is neither inherently prudent nor the absence of action.

The station can consider waiting because a current balance of fourteen, three uncommitted slots, and stage exit terms provide short-term bearing conditions. If cash were at a boundary requiring immediate use or the window closed tonight, the available choices under the same information gap would differ.

Chapter 13: A Single Deviation or a Shift in Conditions?

Sun, 13 Sep 2026 00:00:00 +0800

A Failure After the Previous Pass

At 08:30 Wednesday, E4 began on its existing schedule. At 10:20 it produced the handoff-ready result required by its checklist, releasing Ye Cheng and inspection bench C1. V2 did not delay existing work, so A_D in the joint table can be adjudicated as not occurred.

G2's Wednesday segment still needs an entry review. At 10:35, Ye Cheng first checks C1's current state, then takes the inspection piece left by Tuesday's first segment and reads the internal deviation quantity d under V2 checklist procedure S-3. The rule is pass when d is no greater than 2.0, and fail when d exceeds 2.0.

The first reading is 2.3. G1's archived Monday reading was 1.6 and passed. The results point in opposite classificatory directions, and the new reading crosses the same 2.0 boundary.

Chapter 14: What Material Can Show That the Environment Changed?

Sun, 13 Sep 2026 00:00:00 +0800

Chapter 15: How Thresholds Make Gradual Change Appear Sudden

Sun, 13 Sep 2026 00:00:00 +0800

What Happened Between 4.5 and 4.7?

At 14:00 Thursday, Tang Ke plots the original h log as a timeline: 4.2, 4.3, 4.5, 4.7, 4.9, 5.2, 5.6, 5.8, 6.1. The numbers move gradually, and no material shows an instantaneous jump in working conditions.

Another table contains only pass and fail. G1's Monday d=1.6 passed; Wednesday's 2.3 and 2.2 failed; Thursday's low readings passed and high readings failed. Classification looks like a sudden switch.

The source of suddenness requires layers. Conditions may change gradually and d may move with them, while the checklist places a line at 2.0 and the action rule connects classification to release or pause. Natural process, measurement, classification, and consequence each have their own shape.

This chapter explains only the fictional structure. h, d, and 2.0 are not real material laws. The few readings cannot fit a complete response curve.

Chapter 16: How Earlier Choices Leave Paths Behind

Sun, 13 Sep 2026 00:00:00 +0800

The Same Balance Does Not Mean a Return to the Same Position

At 15:20 Thursday, Chengwan Repair Station still has thirteen units of working capital. Judged only by the balance, this is one more than the twelve it had at the start on Saturday. Yet the team has not returned to a simple state of being “one unit ahead of the starting point.”

The three-unit initial payment has arrived. Two units have been used for G1 and the Tuesday segment of G2, while one unit remains subject to V2 conditions. Three half-day labor slots have gone to G1, G2’s first segment, and the conditional investigation. Wednesday’s window has closed, and both the inspection record and the relationship with the other party, who is waiting for an explanation, now exist.

Chapter 17: Why Prior Investment Cannot Decide Whether to Continue

Sun, 13 Sep 2026 00:00:00 +0800

What Is Missing from “Only One Last Step”

At 16:18 Thursday, the other party replies to the explanation of the pause: if Chengwan can provide a one-page revised arrangement before 10:00 Friday, the parties can discuss receiving the work again from 14:00 to 16:00 next Tuesday. The original total fee of six units remains unchanged, the final payment is still triggered after item-by-item receipt, and no additional advance will be paid.

This does not mean a new window has been established. The message says “can discuss.” It confirms neither object conditions, environmental entry criteria, nor whether the existing stage result may be carried forward, and it does not settle the remaining unit of the initial payment. It only forms a new proposal with a deadline.

Chapter 18: When a Model Should Be Reopened

Sun, 13 Sep 2026 00:00:00 +0800

Not Every New Message Can Revive an Old Model

At 08:40 Friday, Gu Ning opens the one-page draft sent Thursday. The reply deadline is 10:00, and no new message has arrived. The team must first decide which object should be updated if a reply comes.

R17’s first version was closed in Part I. Its original Thursday report was 0.5, and the result in the original Friday window was Y=0. The later formation of V2, G1, and a new proposal cannot return Y to pending.

V2 has its own history. Its Wednesday receipt window closed, and B_D occurred. Even a successful delivery next week would be the result of a new window; it would not make Wednesday’s delivery timely.

Chapter 19: Why Exit Conditions Must Be Written in Advance

Sun, 13 Sep 2026 00:00:00 +0800

Exit Is Not a Remedy Invented After Failure

At 09:45 Friday, V3’s first stage is executable. On Monday the team will use one slot and the remaining one unit of the initial payment to check working conditions, C1’s reference state, the S-3 reading, and whether G2’s first segment can be carried forward.

If the team waits until it sees Monday’s results to discuss exit, window pressure, the fact that another slot has already been used, and each person’s hopes will all enter the rule. The same reading may be interpreted differently with a different mood.

Gu Ning decides to complete exit card X17-1 on Friday. It neither predicts Monday’s failure nor expresses a lack of confidence. It turns unfavorable paths into executable actions.

Chapter 2: There Is More Than One Desired Outcome

Sun, 13 Sep 2026 00:00:00 +0800

Whose Success Is Reaching an Agreement?

At ten on Saturday morning, Tang Ke wrote "try to reach an agreement" in the goal field of the D17 action question card. To her, this was clear enough: the client was willing to talk again the following week, so of course the repair station hoped to win the job. Ye Cheng asked whether it would still count as success if the new job delayed the six slots already promised to other work.

Tang Ke added, "Reach an agreement without affecting other jobs." Rather than adding still more conditions to the sentence, Gu Ning separated what an agreement contained. Winning a new commission, fulfilling it as promised, maintaining a communicative relationship, learning about a new kind of need, and preserving the capacity to adjust are all possible desired outcomes. They do not always arrive together.

Chapter 20: What Does It Cost to Switch Paths?

Sun, 13 Sep 2026 00:00:00 +0800

Alternative Equipment Does Not Make an Alternative Path Available

At 10:40 Monday, V3’s entry and reuse check for G2’s first segment have both passed. Tuesday has only one planned work slot left, and working conditions must remain within E0.

Ye Cheng asks, “Since Q1 can keep h in the low range, wouldn’t moving directly to Q1 tomorrow be more stable?” The station also has candidate device C2, which someone suggests as a third route.

A path is more than a named device. It includes procedures, object compatibility, personnel capability, the time window, quality rules, recording interfaces, and an exit. Any missing element may add a new failure point during switching.

Chapter 21: Why a Correct Judgment Can Still End in Execution Failure

Sun, 13 Sep 2026 00:00:00 +0800

After All Three Entry Conditions Pass

At 13:00 Tuesday, Ye Cheng reads V3’s entry: h=4.4, within 4.2–4.5; C1 reference r=0.2, with absolute value no greater than 0.3; and d=1.9 under S-3, no greater than 2.0.

The 14:00–16:00 window, scope, and version are unchanged, and the seven existing commitments have not increased. X17-1’s cash, capacity, version, and entry constraints all pass.

At 13:10, Gu Ning releases one Tuesday slot and one unit of original unallocated capacity. Path A becomes executable; B and C remain unauthorized, and D is not triggered.

The entry decision is correct: records match the rule. It does not promise error-free execution or make final delivery automatic.

Chapter 22: Why a Good Result Can Still Come from Poor Judgment

Sun, 13 Sep 2026 00:00:00 +0800

The First Rehearsal Selects the Correct File

At 09:00 Wednesday, V3 is archived. Gu Ning creates internal rehearsal T22 to test the handoff-file entrance: Tang Ke enters the current index from the object workspace, and Ye Cheng checks the object, version, entry page, and item list.

The rehearsal sends nothing externally, occupies no production slot, and spends no working-capital unit. Success requires all four fields to match within ten minutes.

At 09:07 Tang Ke provides the correct V3 index. All four match in seven minutes, so the result is successful.

That line alone makes the process look reliable and Tuesday like a chance error. But the screen record shows Tang Ke still opened the first item in recent files.

A Good Result Answers Only What Happened

T22’s first result is truly good: the right file was found, its content was complete, and time met the standard. Criticism of the process cannot turn success into failure.

Chapter 23: How to Preserve Recovery After Consecutive Failures

Sun, 13 Sep 2026 00:00:00 +0800

Two Nondeliveries Are First of All Two Facts

At 10:30 Wednesday, rehearsal T22-C is complete. D17 has already missed two windows: V2 closed after failing Wednesday’s entry, and V3 closed after a version error in its handoff the following Tuesday.

Neither triggered the final payment or achieved the principal goal of a suitable new engagement. At the level of result events, two consecutive windows produced zero.

Their direct paths differ. V2 stopped at conditional entry without releasing production; V3 completed technical items and failed in versioned handoff. Calling both “quality failures” would erase the direction of recovery.

Chapter 24: How to Hand Off a Sustainable Action Record

Sun, 13 Sep 2026 00:00:00 +0800

Complete Files Can Still Be Impossible to Take Over

At 17:20 Friday, R23-1 is complete. D17’s directory contains action cards, objective and consequence tables, stage gates, resource ledgers, exposure maps, evidence pages, threshold diagrams, path ledgers, exit cards, and result pages.

A new recipient might read from first page to last and retell the story, yet still not know whether any task is active, what money is usable, whether an old model may be updated, or who authorizes the next step.

A sustainable record is measured by whether someone can locate the object, adjudicate state, take bounded action, and preserve feedback after the original author leaves, not by file count.

This chapter organizes the products of the first twenty-three chapters into an action-loop template. It summarizes only relations formed within D17 and does not promote fictional values into general operating standards.

Chapter 3: Putting Gains and Losses on the Same Table

Sun, 13 Sep 2026 00:00:00 +0800

What Does a Commission Earn?

At eleven on Saturday, Gu Ning put the D17 goal hierarchy on the table. Tang Ke said the next step was to compare whether the job would make money. If the money from a new commission exceeded materials and labor, continuing would make sense. Ye Cheng asked where the empty half day would be recorded if they reserved a bay and did not win the job.

Tang Ke paused. She had planned to record failure to win the job as zero income, without recording the bay that had been occupied and could not be used for other work. If they did win it, she intended to record all receipts as gains before deducting the materials and labor needed to fulfill it.

Chapter 4: What Expected Value Answers

Sun, 13 Sep 2026 00:00:00 +0800

Is an Average Gain of One Unit Worth Pursuing?

At noon on Saturday, Gu Ning copied several results from the consequence table onto a calculation page. Tang Ke asked whether D17 could be decided simply by multiplying each consequence by its probability and checking whether the sum was positive or negative.

This method captures a core relation in expected value, but it does not complete action judgment. First, D17 has no Version 2 and therefore lacks concrete consequences and corresponding probabilities. Second, a weighted average answers for an average within a given model. It neither says that any single realized result will equal that average nor establishes that the failure is bearable.

This chapter uses several explicit, independent demonstrations to explain expected value. Every number belongs to a thought experiment and does not enter the Chengwan storyline. At this point on Saturday, the client has not replied, and the station has not reserved a bay, bought materials, or sent a clarification.

Probability Game (Part 2): Finding Order in Chaos on 祝融说。

Appendix: A Record for Sustainable Action

Begin with an Actual Action

Chapter 1: Why the Same Probability Can Lead to Different Choices

What Should We Do After Fifty Percent?

Chapter 10: Why Multiple Paths Can Fail Together

Two Budgets Do Not Mean Two Independent Paths

Chapter 11: How Nominal Diversification Conceals Aligned Losses

Why Two Loss Rates of Twenty Percent Are Not Enough

Chapter 12: How the Capacity to Wait Changes Choice

Why a Subject with Margin Can Decide Later

Chapter 13: A Single Deviation or a Shift in Conditions?

A Failure After the Previous Pass

Chapter 14: What Material Can Show That the Environment Changed?

More Failed Readings Are Not Yet Environmental Evidence

Chapter 15: How Thresholds Make Gradual Change Appear Sudden

What Happened Between 4.5 and 4.7?

Chapter 16: How Earlier Choices Leave Paths Behind

The Same Balance Does Not Mean a Return to the Same Position

Chapter 17: Why Prior Investment Cannot Decide Whether to Continue

What Is Missing from “Only One Last Step”

Chapter 18: When a Model Should Be Reopened

Not Every New Message Can Revive an Old Model

Chapter 19: Why Exit Conditions Must Be Written in Advance

Exit Is Not a Remedy Invented After Failure

Chapter 2: There Is More Than One Desired Outcome

Whose Success Is Reaching an Agreement?

Chapter 20: What Does It Cost to Switch Paths?

Alternative Equipment Does Not Make an Alternative Path Available

Chapter 21: Why a Correct Judgment Can Still End in Execution Failure

After All Three Entry Conditions Pass

Chapter 22: Why a Good Result Can Still Come from Poor Judgment

The First Rehearsal Selects the Correct File

A Good Result Answers Only What Happened

Chapter 23: How to Preserve Recovery After Consecutive Failures

Two Nondeliveries Are First of All Two Facts

Chapter 24: How to Hand Off a Sustainable Action Record

Complete Files Can Still Be Impossible to Take Over

Chapter 3: Putting Gains and Losses on the Same Table

What Does a Commission Earn?

Chapter 4: What Expected Value Answers

Is an Average Gain of One Unit Worth Pursuing?