S.M.M. Field Manual

Stop Manipulating Me · A Field Guide to Psychological Influence

FIELD MANUAL FM 1-13
SECTOR Cognition · Bias
CLEARANCE Training Use
EDITION 01
Part I · Foundations Chapter
Operation · Blind Spot · Callsign COMPASS-13

Ch 13 — Cognitive Bias

Systematic Error · Predictable shortcuts, predictably exploited
Blind Spot

Mission Brief

Objective

Map the systematic architecture of error. Human judgment does not fail randomly — it fails in patterned, direction-specific ways because the mind runs on fast heuristics that misfire under identifiable conditions (Tversky & Kahneman, 1974). A cognitive bias is a structural feature of the judgment system, present in experts and novices alike, and exploitable precisely because it is predictable — Category 8 (T8.1–T8.31) is this science turned into technique.

Why It Matters

The analyst who knows the architecture stops treating "I just felt sure" as evidence and asks which shortcut produced this certainty, and has someone arranged the situation to trigger it? Because biases are systematic they are anticipatable — and each defense is the same move: import the input the shortcut skipped (the base rate, the independent reference, the forward value, the discarded alternative).

Sector
Cognitive Bias · Cat 8 (T8.1–T8.31)
Doctrine Ref
Mechanism §2.18 · §2.25 · §2.26
Prerequisite
FM 1-1 §1 — bounded rationality / mismatch
Confidence
C4–C5 core (contested magnitudes flagged)
Word Order
Analytic & defensive throughout
Operational Priority
Systematic Error
Biases are stable targets — known, repeatable, and aimed at.
Governing DiagnosticThe tilt is reliable, and someone can arrange the situation to use it. Whether a bias shifts judgment by a lot or a little, its direction is dependable — and the direction is exactly what a manipulator arranges. A decision that depends on you deciding now is a decision that depends on your biases going unchecked.

1. Situation · The Terrain

The mind uses error-prone shortcuts because it must. Simon’s principle of bounded rationality established that real decision-makers have limited time, information, and computational capacity, so they cannot optimize — they satisfice, searching until an option is good enough, then stopping (Simon, 1955) [C4]. Heuristics trade guaranteed accuracy for speed; bias is the systematic cost of that trade — the error the shortcut leaves where it does not fit.

A useful organizing frame is the two-systems description (Kahneman, 2011; Stanovich & West, 2000) [C3, a metaphor rather than two literal brain systems]. System 1 is fast, automatic, associative — it produces the impression and the feeling of certainty. System 2 is slow, effortful, and lazy — it monitors System 1 but usually rubber-stamps it. Most biases are System 1 outputs System 2 fails to catch, and the manipulative implication is precise: techniques that keep System 2 disengaged — speed, load, fatigue, arousal, fluency — let the biases pass through uncorrected.

A caveat frames all of it: psychology’s replication crisis reproduced only about a third to a half of tested effects (Open Science Collaboration, 2015) [C4]. The core biases here — anchoring, framing, loss aversion, sunk cost — are among the best-replicated findings (Klein et al., 2014) [C4]; specific magnitude claims are contested and flagged inline. The defensive lesson survives the caveat: the direction is what a manipulator uses.

2. Enemy Forces · The Biases, Weaponized

Anchoring (T8.2) is the most robust and most exploited: an arbitrary first number pulls estimates toward it even when transparently irrelevant (Tversky & Kahneman, 1974) [C5]; experienced real-estate agents anchored appraisals to a planted list price while denying its influence (Northcraft & Neale, 1987) [C4]. It is the engine of reference pricing — the crossed-out “original,” the extreme opening offer (T20.13, T21.10). The strong “coherent arbitrariness” claim is [C3, contested] (Ariely et al., 2003).

Framing & loss aversion (T8.8, T8.6): prospect theory shows people judge outcomes against a reference point, and losses loom roughly twice as large as gains (Kahneman & Tversky, 1979) [C5]. The “Asian disease” problem flipped choices with identical statistics framed as lives saved vs. lost (Tversky & Kahneman, 1981) [C5]. The endowment effect (T8.7) is loss aversion applied to what you already hold — why auto-converting free trials and “you’ll lose what you’ve built” retention appeals work (Thaler, 1980) [C4].

Availability & representativeness (T8.3, T8.14): people judge frequency by how easily examples come to mind, and probability by similarity to a stereotype rather than base rate — the Linda conjunction error (Tversky & Kahneman, 1973, 1983) [C4]. A vivid anecdote is fed precisely where a base rate belongs. Decoy effect (T8.9): adding an obviously inferior option shifts choice toward the option that dominates it (Huber, Payne & Puto, 1982) [C4] — the three-tier pricing menu (T20.3). Sunk cost (T8.19) and status-quo/default (T8.17): inertia biases behind commitment traps and opt-out defaults — opt-out organ-donation defaults produce near-universal consent (Johnson & Goldstein, 2003) [C4]. Confirmation bias (T8.1) seals the others in, corrupting the checking process itself (Nickerson, 1998) [C4].

Table 13.1 — The exploitable-bias reference (Category 8)

BiasHow it’s exploited (ID)Detection indicatorDefensive correction
AnchoringReference/first-offer pricing (T8.2, T20.13, T21.10)A number supplied before you form your ownSet your own independent reference first
Framing / loss aversionGain/loss packaging; “don’t miss” (T8.8, T8.6)Choice flips when you restate the frameRestate both ways, in absolute terms
EndowmentFree-trial-to-paid; “keep what’s yours” (T8.7, T20.17)Value inflated because it’s now “yours”Value it as if you didn’t already own it
AvailabilityVivid anecdote in place of a rate (T8.3)A story where a statistic belongsAsk for base rate and denominator
Base-rate neglectDetailed scenario that “fits” (T8.14)Plausible specifics, no prevalence“How common is this actually?”
SurvivorshipWinners-only testimonials (T8.11)Successes shown, failures invisibleAsk for the average and the failures
DecoyEngineered pricing tiers (T8.9, T20.3)An option no one would rationally pickJudge each option in isolation
Sunk-costCommitment/loyalty lock-in (T8.19, T18.16)“Can’t quit after investing so much”Decide on future value only; past is gone
Status-quo / defaultOpt-out traps, auto-renewal (T8.17, T20.19)Pre-selected option you didn’t chooseActively choose; assume defaults serve the setter
Confirmation biasBelief-consistent feeding (T8.1, T13.19)Only supporting evidence appearsSeek the strongest disconfirming case

3. Enemy Forces · The Stack Is the Real Threat

Individual biases are exploitable; stacked biases are how real manipulations work. Read a high-pressure sale as a bias stack: an inflated “regular price” anchors (T8.2); the discount is loss-framed as expiring (T8.8, T8.6); a middle tier acts as a decoy (T8.9); vivid testimonials feed availability while hiding the survivorship denominator (T8.3, T8.11); a manufactured deadline keeps System 2 disengaged so none of the corrections run; and after a deposit, sunk cost (T8.19) and the auto-renewal default (T8.17) keep you in. No single bias carries the sale — the stack does.

The honest-influence boundary

Working with biases is not inherently manipulative — all communication is framed and all choices have some default. The distinguishing question tracks the governing diagnostic: does the framing help the recipient decide in line with their own goals, and does it survive disclosure? A default that protects inattentive people (auto-enrolled retirement savings, opt-out organ donation) uses status-quo bias for the user and remains defensible when explained. The line is crossed when the anchor, decoy, or default serves the architect against the chooser’s interest and depends on the chooser not noticing — the fake “original” price, the pre-ticked add-on, the extractive default. Ethical use still works when fully explained; manipulative use collapses the moment it is pointed out.

Table 13.2 — Choice architecture: ethical nudge vs. manipulative exploit

Bias engagedEthical use (survives disclosure)Manipulative use (needs your inattention)
Status-quo / defaultAuto-enroll in savings; easy opt-outPre-ticked add-on; hidden auto-renewal (T20.19)
AnchoringGenuine reference price; real MSRPInflated fake “original” price (T20.13)
FramingClearest true framing for comprehensionLoss frame hiding the neutral baseline (T8.8)
Decoy / contrast— (no legitimate use as a trap)Tier that exists only to mislead (T8.9, T20.3)
Scarcity / lossHonest low-stock or real deadlineFabricated countdown, false “last chance” (T4.2)
Indicators · Warning OrderDetection · S-C-A-R-E-D
  • A number arrives first. A "regular price," first offer, or suggested amount reaches you before you form your own estimate of value (T8.2). [S-C-A-R-E-D: C · verification gap]
  • The choice flips on rewording. An offer built on what you'll lose; a decision that reverses when you restate it in the opposite frame (T8.6, T8.8). [A · affect / loss lever]
  • A vivid story where a rate belongs. A testimonial or worst-case scenario offered in place of a base rate or denominator (T8.3, T8.14, T8.11). [C · consistency / verification gap]
  • An option no one would pick. A menu containing a pointless or strictly-worse tier, present only to reshape the comparison (T8.9, T20.3). [R · scarcity/anchor lever]
  • "Can't quit after all this." A decision justified mainly by prior investment; a pre-selected option or auto-renewal you didn't choose (T8.19, T8.17). [E · exit friction / lock-in]
  • Deciding while rushed or loaded. Pressure to decide fast, tired, or overloaded so no correction can run; only confirming evidence appears (T8.1). [S · speed · D · deviation]
Rules of Engagement · Defense DoctrineP-A-U-S-E-D

Because each bias is a shortcut that skips a specific input, each defense is the same shape — restore the skipped input and let deliberation run. The doctrine is five moves.

ROE // Import the Missing Input
  • Slow the decision so System 2 engages. Every bias passes because the fast system went unchecked; speed, load, arousal, and fatigue are the enabling conditions, not incidental. Interposing time is the master defense (P-A-U-S-E-D: P — Pause the clock).
  • Generate your own reference before you see theirs. Decide independently what a thing is worth and how likely an event is, then look at the anchor or offer. A reference you brought is information; a reference they supplied first is a tactic.
  • Restate every framed choice in neutral, absolute, both-direction terms. Convert "90% survive" and "10% die" into the same figure; demand the base rate behind the vivid story and the denominator behind the testimonials (A — Ask & verify out-of-band).
  • Decide forward, not backward. For any "can't quit now" pull, weigh only future cost against future benefit; name the sunk cost aloud and set it aside. Treat every pre-selection as a claim to check; assume the default serves whoever set it.
  • Seek the disconfirming case. Against confirmation and motivated reasoning, deliberately look for the strongest evidence against the favored conclusion; run it past a second person outside your frame and deadline (S — Seek a second opinion).
Field ChecklistLayer

Check the Numbers & Frames

  • Did a number reach me before I formed my own estimate? (Anchoring, T8.2) → set an independent reference.
  • Is this built on what I'll lose? Does my choice flip if I restate the frame? (T8.6, T8.8) → restate in neutral, absolute terms.

Check the Evidence

  • A vivid story where a base rate belongs? (Availability, T8.3) → ask "out of how many?"
  • Only winners shown? (Survivorship, T8.11) → ask for the average and the failures. Only confirming evidence? (T8.1) → seek the strongest disconfirming case.

Check the Choice Set & Inertia

  • An option no one would rationally pick? (Decoy, T8.9, T20.3) → judge each in isolation.
  • Continuing mainly because of what I've already put in, or a pre-selected auto-renewal? (T8.19, T8.17) → decide forward; actively choose.

Check the Conditions

  • Am I kept fast, loaded, or rushed so I can't check the above? (System 2 offline) → pause; decide later.
  • Have I run this past a second person who didn't get my anchor, frame, or sunk cost? (P-A-U-S-E-D: S)

Two or more unresolved boxes: a bias stack is doing the deciding — slow down and import the missing input before acting. A deal that depends on you deciding now depends on your biases going unchecked.

Key Intel · TakeawaysSummary
  • Bias is systematic, not random. Heuristics are efficient but misfire in predictable, direction-specific ways — and predictability is exactly what makes a bias both exploitable and checkable (Tversky & Kahneman, 1974; Simon, 1955) [C4–C5].
  • Most biases are unchecked fast-system outputs. System 1 produces the impression; System 2 rubber-stamps it. Speed, load, arousal, and fatigue keep the checker offline (Kahneman, 2011) [C3].
  • Anchoring, framing, and loss aversion are the workhorses. A first number pulls estimates (T8.2); identical facts flip choices by frame (T8.8); losses loom ~twice as large as gains (T8.6) [C4–C5].
  • Vividness beats statistics unless you demand the base rate. Availability and representativeness make a story outweigh the rate (T8.3, T8.14); the fix is "out of how many, how often?" [C4].
  • Inertia is exploited by sunk cost and defaults. Past investment (T8.19) and pre-selected options (T8.17, T20.19) keep people in against their interest; decide on forward value and actively choose [C4].
  • The threat is the stack. Real manipulations layer several biases so catching one leaves others working; slowing the decision and importing each skipped input collapses the stack.
Training Op · Red-Team & Operational PracticeAuthorized · Consented
Demonstration ObjectiveLet a class or decision team watch a bias move their own judgment in a controlled setting, so that "the error is systematic, predictable, and therefore exploitable" becomes a measured firsthand result. Target "aha": an arbitrary number I knew was irrelevant still pulled my estimate; the same facts flipped my choice when re-worded.
Exercise Design (consented)Run a consented, debriefed anchoring demonstration (after Tversky & Kahneman): participants write down a manifestly irrelevant number, then estimate an unknown quantity; split the room by high vs. low anchor and compare aggregate estimates. A complementary framing exercise uses the "Asian disease" structure with fabricated, non-emotive figures — half see a gain frame, half a mathematically identical loss frame. All figures notional and non-topical; nothing is sold.
What Participants CatchThat an irrelevant first number anchored their estimate even when they insisted it hadn't; that identical facts produced opposite choices under gain vs. loss wording; that neither correction ran because the fast system was never checked. (S-C-A-R-E-D: C — verification gaps)
Debrief-to-DefenseReveal the split-group design and install the doctrine — import the missing input: generate your own reference before you see theirs; restate every framed choice in neutral, absolute terms; slow the decision so System 2 engages; run consequential judgments past a second person. (P-A-U-S-E-D: P — Pause · S — Seek a second opinion · A — Ask & verify)
Guardrail & Failure ModesConsent and immediate debrief mandatory; notional, non-topical stimuli only — no real pricing, no charged scenario that steers a live purchase; results reported in aggregate. It backfires if a real offer or live negotiation is used as the "demonstration" (that is deployment, not teaching), if biases are stacked to push a real commitment, or if it is framed as "you're irrational" rather than "the error is systematic and checkable." As an effect-type demonstration the deliverable is a decision-hygiene habit — there is no reusable manipulation artifact to distribute.
Comms Net · Linked SectorsCross-references
  • FM 1-1
    Back — bounded rationality & evolutionary mismatch
  • Mechanisms §2.18 · §2.25 · §2.26
    Fwd — overload, motivated reasoning, confirmation
  • Category 8
    Fwd — Cognitive Bias Exploitation, in full
  • T20.13 · T21.10 · T20.3
    Fwd — anchor pricing, negotiation, decoy tiers
  • T20.19 · T20.20
    Fwd — default/opt-out & subscription retention
  • Ch 14 · Ch 17
    Adjacent — automaticity & urgency keep the checker offline
  • Ch 2 · Ch 5
    Back — attention narrowing & arousal disable deliberation
§ Doctrinal Sources · SelectedFurther Reading
  • Tversky, A., & Kahneman, D. (1974). "Judgment under uncertainty: heuristics and biases." Science, 185, 1124–1131.
  • Kahneman, D., & Tversky, A. (1979). "Prospect theory." Econometrica, 47, 263–291.
  • Tversky, A., & Kahneman, D. (1981). "The framing of decisions." Science, 211, 453–458.
  • Simon, H. A. (1955). "A behavioral model of rational choice." Quarterly Journal of Economics, 69, 99–118.
  • Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
  • Northcraft, G. B., & Neale, M. A. (1987). "Experts, amateurs, and real estate." Org. Behavior & Human Decision Processes, 39, 84–97.
  • Huber, J., Payne, J. W., & Puto, C. (1982). "Adding asymmetrically dominated alternatives." Journal of Consumer Research, 9, 90–98.
  • Johnson, E. J., & Goldstein, D. (2003). "Do defaults save lives?" Science, 302, 1338–1339.
  • Arkes, H. R., & Blumer, C. (1985). "The psychology of sunk cost." Org. Behavior & Human Decision Processes, 35, 124–140.
  • Nickerson, R. S. (1998). "Confirmation bias." Review of General Psychology, 2, 175–220.
  • Klein, R. A., et al. (2014). "Investigating variation in replicability (Many Labs)." Social Psychology, 45, 142–152.
  • Open Science Collaboration. (2015). "Estimating the reproducibility of psychological science." Science, 349, aac4716.