Appendix I: The Cold-Case Capstone Workbook

This is your workbook. Across forty chapters you have worked one fictional but realistic case — the death of Marcus Diallo, 38, a general contractor found on the morning of 18 October in the partially burned cabin he was renovating on Mill Creek Road in Carrow County. The first responders called it a probable accidental fire. The book spent forty chapters asking whether that was true, adding one kind of evidence at a time, and refusing — every time — to draw the conclusion until the science had earned it. Chapter 39 is the capstone where the file is finally assembled. This appendix is the set of blank tools you use to do that assembly yourself, chapter by chapter, so that when you reach the capstone you are reading your own completed file rather than someone else's.

There are four templates here, and they correspond exactly to the two tables Chapter 39 says a competent analyst actually builds (an evidence log and an exclusion matrix), plus the two documents the work has to produce in the real world (a chain-of-custody log and a final report). The templates are not decoration. As Chapter 39 puts it, the format itself is a guardrail against overstatement: a narrative lets you quietly lean on a soil association as if it were a fingerprint, but a table with a column headed "honest verb" will not let you write proves where you recorded supports. Fill them in honestly and the conclusion takes care of itself.

The one rule that governs every template. Carry each piece of evidence forward at exactly the strength it earned, and not one degree stronger. The book's scale, from page one, is: excludes → consistent with → strongly supports → (the forbidden) proves. Reserve "identifies" for the rare method (good single-source DNA) that has earned it. A finding does not get promoted just because it now has company; convergence is the agreement of independent threads, not the summing of weak ones. When you are unsure which verb to use, use the weaker one — that instinct is the whole discipline of this book.

A note on the worked answers. This workbook is deliberately blank — the templates are yours to complete as you read. The fully worked conclusion lives in Chapter 39 (the capstone): the completed exclusion matrix, the convergence diagram, the honest "strongly supports" conclusion, and the defense's residual-uncertainty case. Do the work here first; check it against Chapter 39 after. If your completed matrix matches the capstone's, you have learned to reason like a forensic scientist. If your verbs are stronger than the capstone's, you have just caught yourself committing the exact error the book exists to cure — which is itself the most valuable result this workbook can give you.

Template 1 — The chapter-by-chapter evidence log

This is the spine of your file. Every chapter (with a few explicitly-noted exceptions for the most specialized topics) ends with a 🗂️ The Case File checkpoint that adds one evidence type. Log each one as you go. The discipline of the log is the "honest verb" column: it forces you to record the weak threads as weak, in real time, before you know where the case is heading.

How to fill it in. For each chapter's Case File checkpoint, write down: the evidence type added; what it establishes (the defensible observation); what it does not establish (the limit, the alternative explanation); and the single honest verb at its true strength. Leave the verb blank until you can defend it. If a chapter complicates the picture rather than tidying it (some are designed to), say so — that is honest reasoning, not a failure.

COLD-CASE EVIDENCE LOG — Mill Creek / Marcus Diallo            Analyst: __________  Date: ________

 Ch │ Evidence type added            │ What it ESTABLISHES        │ What it does NOT establish   │ HONEST VERB
────┼────────────────────────────────┼────────────────────────────┼──────────────────────────────┼─────────────
  1 │ the file received (bare facts, │ a question, not a finding  │ anything — no analysis yet    │ (n/a)
    │ accidental-fire assumption)    │                            │                              │
  2 │ ............................   │ ........................   │ ...........................   │ ...........
  3 │ ............................   │ ........................   │ ...........................   │ ...........
  4 │ ............................   │ ........................   │ ...........................   │ ...........
  5 │ ............................   │ ........................   │ ...........................   │ ...........
  6 │ ............................   │ ........................   │ ...........................   │ ...........
  7 │ ............................   │ ........................   │ ...........................   │ ...........
  8 │ ............................   │ ........................   │ ...........................   │ ...........
  9 │ ............................   │ ........................   │ ...........................   │ ...........
 10 │ ............................   │ ........................   │ ...........................   │ ...........
 11 │ ............................   │ ........................   │ ...........................   │ ...........
 12 │ ............................   │ ........................   │ ...........................   │ ...........
 13 │ ............................   │ ........................   │ ...........................   │ ...........
 14 │ ............................   │ ........................   │ ...........................   │ ...........
 15 │ ............................   │ ........................   │ ...........................   │ ...........
 16 │ ............................   │ ........................   │ ...........................   │ ...........
 17 │ ............................   │ ........................   │ ...........................   │ ...........
 18 │ ............................   │ ........................   │ ...........................   │ ...........
 19 │ ............................   │ ........................   │ ...........................   │ ...........
 20 │ ............................   │ ........................   │ ...........................   │ ...........
 21 │ ............................   │ ........................   │ ...........................   │ ...........
 22 │ ............................   │ ........................   │ ...........................   │ ...........
 23 │ ............................   │ ........................   │ ...........................   │ ...........
 24 │ ............................   │ ........................   │ ...........................   │ ...........
 25 │ ............................   │ ........................   │ ...........................   │ ...........
 26 │ ............................   │ ........................   │ ...........................   │ ...........
 27 │ ............................   │ ........................   │ ...........................   │ ...........
 28 │ ............................   │ ........................   │ ...........................   │ ...........
 29 │ ............................   │ ........................   │ ...........................   │ ...........
 30 │ ............................   │ ........................   │ ...........................   │ ...........
 31 │ ............................   │ ........................   │ ...........................   │ ...........
 32 │ ............................   │ ........................   │ ...........................   │ ...........
 33 │ ............................   │ ........................   │ ...........................   │ ...........
 34 │ ............................   │ ........................   │ ...........................   │ ...........
 35 │ ............................   │ ........................   │ ...........................   │ ...........
 36 │ ............................   │ ........................   │ ...........................   │ ...........
 37 │ ............................   │ ........................   │ ...........................   │ ...........
 38 │ ............................   │ ........................   │ ...........................   │ ...........
────┴────────────────────────────────┴────────────────────────────┴──────────────────────────────┴─────────────
 39 │ CAPSTONE — assemble the whole log; do not add; WEIGH. (Chapters 1–38 above.)
 40 │ (no Case File — careers reflection)

Worked example of one row, to set the standard (this is the kind of entry the chapters model; it is the Chapter 11 beat, stated at honest strength):

text 11 │ Autopsy: no soot in the deep │ Diallo was DEAD BEFORE THE │ does not name a killer, a │ strongly │ airways; blunt-force skull │ FIRE; a homicide, not an │ time to the minute, or how │ supports │ fracture │ accident │ he was overcome │ (homicide)

Notice what the entry refuses to do. It records a genuinely decisive finding — the hinge of the whole case — and still writes only "strongly supports (homicide)," not "proves," and still lists what it does not establish. That restraint, on the single most important finding in the file, is the model for every row.

Template 2 — The exclusion matrix

This is the most important tool in the capstone, because it embodies the book's first theme: forensic science excludes far more reliably than it proves. You make progress not by proving guilt but by eliminating possibilities until the field is small enough that whatever convergence remains actually means something. Build this matrix as the exclusionary evidence comes in (most of it lands in Parts II, IV, V, and VI), then read it whole at the capstone.

The cast (persons of interest). A small, fixed set recurs through the book:

Roy Keller — Diallo's business partner and co-owner of the property flip.
Dana Whitfield — Diallo's estranged partner; an old domestic-violence report exists (handled soberly in Chapter 37).
Victor Salas — a subcontractor owed money, openly angry.
Cody Renner — a 22-year-old local who falsely confesses under a long interrogation (Chapter 33). The book's wrongful-conviction near-miss.
(plus the unknown-stranger theory — the possibility that someone never identified did this, which an honest assembly must confront and answer, not ignore.)

How to fill it in. Down the rows, the persons of interest (and the stranger theory). Across the columns, the lines of evidence that bear on who. In each cell write one of: EXCL (this evidence excludes the person), CONS (consistent with / not excluded), n/a (does not bear on this person), or (—) (not tested / unknown). Then read each row to an overall status. The point of the grid is that it narrows the field by exclusion first, before anyone reasons about who remains.

EXCLUSION MATRIX — Mill Creek case        Analyst: __________  Date: ________

 Legend:  EXCL = excluded by this evidence   CONS = consistent with / not excluded
          n/a  = does not bear on this person   (—) = not tested / unknown

 PERSON              │ DNA mixture │ Alibi /      │ Confession   │ Motive      │ Physical trace  │ OVERALL
                     │ (Ch.8–9)    │ cell-site    │ analysis     │ (Ch.18,27)  │ soil/pollen/    │ STATUS
                     │             │ (Ch.25,32)   │ (Ch.33)      │             │ CCTV(Ch.13,24,26)│
 ────────────────────┼─────────────┼──────────────┼──────────────┼─────────────┼─────────────────┼──────────
 Victor Salas        │ .........   │ .........    │ .........    │ .........   │ .........       │ .........
 (subcontractor)     │             │              │              │             │                 │
 ────────────────────┼─────────────┼──────────────┼──────────────┼─────────────┼─────────────────┼──────────
 Dana Whitfield      │ .........   │ .........    │ .........    │ .........   │ .........       │ .........
 (estranged partner) │             │              │              │             │                 │
 ────────────────────┼─────────────┼──────────────┼──────────────┼─────────────┼─────────────────┼──────────
 Cody Renner         │ .........   │ .........    │ .........    │ .........   │ .........       │ .........
 (false confessor)   │             │              │              │             │                 │
 ────────────────────┼─────────────┼──────────────┼──────────────┼─────────────┼─────────────────┼──────────
 Unknown stranger    │ .........   │ .........    │ .........    │ .........   │ .........       │ .........
 (theory)            │             │              │              │             │                 │
 ────────────────────┼─────────────┼──────────────┼──────────────┼─────────────┼─────────────────┼──────────
 Roy Keller          │ .........   │ .........    │ .........    │ .........   │ .........       │ .........
 (business partner)  │             │              │              │             │                 │
 ────────────────────┴─────────────┴──────────────┴──────────────┴─────────────┴─────────────────┴──────────

Two interpretive habits, drilled in Chapter 39, that make or break this matrix:

"Not excluded" is not "guilty." When the matrix clears the field and leaves one person un-excluded, the matrix has done its proper job — clearing the field, not convicting. What that remaining "not excluded" means requires the convergence reasoning of Template 3; the matrix alone cannot get you there.
Beware asymmetric scrutiny (the bias trap). The matrix is only a safeguard if you apply the same evidentiary standard to every cell — asking of your remaining suspect's "consistent with" exactly the doubting questions you asked of the people you excluded. A matrix filled in knowing the answer you want can launder a hunch into the appearance of method. Fill it in with the suspect's identity held as far from the reasoning as you can manage.

One row, worked, to set the standard (the most morally important row in the book — Cody Renner):

text Cody Renner │ EXCL │ EXCL │ CONFESSION │ n/a │ (—) │ EXCLUDED (false confessor) │ (excluded) │ (cell-site │ SHOWN FALSE │ │ │ (exonerated │ │ elsewhere) │ (contradicted)│ │ │ by science)

Sit with that row. A young man confessed — and the science excluded him anyway, three independent ways (DNA, cell-site, and the dissection of the confession against the physical timeline). On television a confession ends the case; in an honest file it is not exempt from exclusion. This is the single most important thing forensic science ever does, and your matrix should capture it cleanly.

Template 3 — The chain-of-custody log

The science is worthless if the evidence cannot be trusted, and trust is documented, not assumed. The chain of custody is the unbroken, written record of everyone who handled an item, when, and why — the record that lets a court believe the thing tested is the thing collected (Chapter 2). A break in the chain is a gift to the defense and, in the Mill Creek case, a real vulnerability: the scene was treated as an accident in its first hours and was not held to a homicide standard, so its early integrity is permanently compromised (Chapter 2, and a residual uncertainty named in Chapter 39).

How to fill it in. One log per item of evidence. Record the item and a unique identifier, where and when it was collected and by whom, how it was packaged and preserved (remember Chapter 2: biological evidence goes in breathable packaging, never sealed plastic, which traps moisture and ruins DNA), and then every subsequent transfer — released by, received by, date/time, and purpose — with no gaps.

CHAIN-OF-CUSTODY LOG                                          Case: Mill Creek / Diallo

 ITEM: ____________________________________   ITEM No.: ____________
 Collected by: __________________  Location collected: ________________________________
 Date/time collected: ____________________
 Packaging / preservation: ____________________________________________________________
   (biological evidence: breathable packaging, air-dried, refrigerated as appropriate — NOT sealed plastic)

 ┌──────┬──────────────────────┬──────────────────────┬───────────────────┬───────────────────────────┐
 │ Seq. │ RELEASED BY (name)   │ RECEIVED BY (name)   │ DATE / TIME       │ PURPOSE / NOTES           │
 ├──────┼──────────────────────┼──────────────────────┼───────────────────┼───────────────────────────┤
 │  1   │ ....................  │ ....................  │ ................   │ ........................  │
 │  2   │ ....................  │ ....................  │ ................   │ ........................  │
 │  3   │ ....................  │ ....................  │ ................   │ ........................  │
 │  4   │ ....................  │ ....................  │ ................   │ ........................  │
 │  5   │ ....................  │ ....................  │ ................   │ ........................  │
 └──────┴──────────────────────┴──────────────────────┴───────────────────┴───────────────────────────┘

 Integrity flags (note any gap, reseal, or compromise): ________________________________________________

Use one of these for each major item in the cold-case inventory (Chapter 3): the gas can, the tools, the charred documents, the cabin door, the victim's phone, the stray cartridge case, the toxicology specimens, the soil and trace samples. Then ask the question the defense will ask of every one: can you account for this item's whereabouts at every moment between the scene and the bench? Where you cannot — and at this scene there will be such places — that gap is part of the honest record, and it belongs in the residual-uncertainty section of your final report, not buried.

Template 4 — The final forensic report skeleton

The capstone produces a report. Its grammar is the examination: the whole book lives in the difference between a report that says "strongly supports" and one that says "proves." Use this skeleton to assemble your conclusion from the completed log and matrix. The cardinal rule, repeated because it is everything: state the conclusion at the strongest verb the threads jointly earn — and never one verb beyond it.

FINAL FORENSIC ANALYSIS — Mill Creek / Marcus Diallo          Prepared by: __________   Date: ________

 1. THE QUESTION
    State the legal question the analysis addresses, exactly as posed on day one.
    ____________________________________________________________________________________________
    (e.g., "Was the death of Marcus Diallo an accident, or a homicide staged to look like one?")

 2. SCOPE AND LIMITATIONS UP FRONT
    What this report does and does not attempt. Note the compromised scene and any chain-of-custody
    gaps HERE, at the front, not buried at the back.
    ____________________________________________________________________________________________

 3. SUMMARY OF EVIDENCE (from Template 1)
    Each evidence type, in order, at its honest verb. Do not argue yet — list and weigh.
    ____________________________________________________________________________________________

 4. EXCLUSIONS (from Template 2)
    Who the science rules out, and the specific independent findings that exclude each. (Exclusion is
    the surest thing the science does — lead with it.)
    ____________________________________________________________________________________________

 5. CONVERGENCE
    For any person NOT excluded: lay the INDEPENDENT threads side by side (motive / opportunity /
    means / biological anchor). State explicitly WHY they are independent — independence is what gives
    convergence its weight; non-independent threads (bias-driven) cannot be multiplied.
    ____________________________________________________________________________________________

 6. THE CONCLUSION — AT ITS TRUE STRENGTH
    One sentence, on the scale excludes → consistent with → strongly supports. NOT "proves."
    ____________________________________________________________________________________________
    "The evidence ____________ that ____________ ."

 7. THE DEFENSE'S CASE / RESIDUAL UNCERTAINTY
    Build the strongest honest case against your own conclusion. (A conclusion is only as honest as the
    best argument against it.) Name every limit: mixture vs. single-source match; class vs. individual
    evidence; "in the area" vs. "at the scene"; the compromised scene; any original bias.
    ____________________________________________________________________________________________

 8. FURTHER TESTING
    What would STRENGTHEN — and, just as importantly, what could WEAKEN or overturn — the conclusion.
    (Name this list BEFORE the verdict, not after. A verdict that cannot say what would change it is a
    belief, not a finding. The symmetry — every test could cut either way — is what makes it science.)
    ____________________________________________________________________________________________

 9. SIGN-OFF
    Analyst: ______________________   Reviewed by (independent/blind, ideally): ______________________

Two reminders for sections 6 and 7, because they are where reports go wrong:

"Strongly supports" is not a hedge. It is the strongest honest verb a converging, exclusion-cleared case earns — far stronger than the "consistent with" of any single thread, and a great deal more durable on the witness stand than "proves," which the first hostile question turns into a liability. You are being exactly as confident as the evidence permits, which is the entire skill.
The ability to write section 7 is the surest sign you did section 6 honestly. A competent assembly can name, out loud, everything its own conclusion does not establish. If you cannot fill in the residual uncertainty, you have not finished the analysis — you have written an advocacy brief.

Closing the workbook

When your log is complete, your matrix is filled, your custody is documented, and your report is written, turn to Chapter 39 and read the capstone against your own file. The capstone states the conclusion the science has earned — a homicide staged as an accident; the evidence strongly supports one person; it does not prove it; and here is everything it does not establish. If your work reached that same place, by the same honest verbs, you have done what the entire book was training you to do: you assembled a complete forensic case, let the evidence exclude what it could, and then said exactly as much as it would let you say — and not one word more.

That restraint is not the weakness of the field. It is its integrity. It is a capital mistake to theorize before one has data. You have the data now. Close the file at its true strength.