What does Step-Down Communication Protocols conclude, in numbers, about “Expertise Reversal: The Best Number of Bridging Steps Depends on the Receiver's Chunk Dictionary”?

Answer: Chapter 3 computes the best number of bridged steps: novices are best served by bridging everything, a practised receiver by two steps, and full bridging raises the latter's failure rate from 0.344 to 0.401; probing first beats the best fixed policy only at a classification accuracy of 0.81 or more.

Scope of the numbers: Its parameters are illustrative; all calculations are generated by its programs; the book offers no psychological assessment or clinical judgement, names no real person or case, and is general, not personalised, advice on business communication. Chapter 8 gives the step-down ledger, a field checker tool with its specifications, and a pre-registered test with 198 records per arm; in the synthetic ledger green records have a rework rate of 0.115 against 0.520 for red ones.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.344, 0.401

expected cost R*P(fail) + T*b, R10.0, T = 0.3
level q b0 b=1 b=2 b=3 b=4 b=5 b=6
novice0.20 9.74 9.76 9.54 8.91 7.66 5.66 2.80
middle0.60 7.66 7.46 7.08 6.56 5.84 4.88 3.60
expert0.95 3.44 3.35 3.18 3.50 3.70 4.78 5.81
best b and failure rates
novice best b 6 fail at best 0.100 at b0 0.974 at b=6 0.100
middle best b 6 fail at best 0.180 at b0 0.766 at b=6 0.180
expert best b 2 fail at best 0.258 at b0 0.344 at b=6 0.401
Show the full program
# Chapter 3. p31: the best number of bridged steps depends on the
# receiver's dictionary (expertise reversal, Kalyuga et al. 2003).
# A compressed message skips N = 6 middle steps; the sender spells out
# b of them without knowing which ones the receiver already holds.
# Each step is in the receiver's dictionary with probability q.
#   unbridged, in dictionary:   rebuilt with P_IN, no extra load
#   unbridged, outside:         rebuilt with P_OUT, +1 chunk of load
#   bridged, outside:           understood with P_BR, no extra load
#   bridged, in dictionary:     redundant, +W chunk of load
# Load = 2 + extra; capacity 3/4/5 (1:2:1); factor 0.5 per excess
# chunk. Cost = R * P(fail) + T * b. All values are illustrative.
import math

N = 6
P_IN, P_OUT, P_BR = 0.95, 0.60, 0.98
W = 0.5
R, T = 10.0, 0.3
CAP_W = ((3, 0.25), (4, 0.50), (5, 0.25))
LEVELS = (("novice", 0.2), ("middle", 0.6), ("expert", 0.95))


def binom(n, k, q):
    return math.comb(n, k) * q ** k * (1.0 - q) ** (n - k)


def wm(load):
    return math.fsum(w * 0.5 ** max(0.0, load - c) for c, w in CAP_W)


def success(b, q):
    terms = []
    for j in range(b + 1):          # bridged steps already held
        for m in range(N - b + 1):  # unbridged steps already held
            pr = binom(b, j, q) * binom(N - b, m, q)
            out = N - b - m
            p = P_BR ** (b - j) * P_IN ** m * P_OUT ** out
            load = 2 + out + W * j
            terms.append(pr * p * wm(load))
    return math.fsum(terms)


def cost(b, q):
    return R * (1.0 - success(b, q)) + T * b


if __name__ == "__main__":
    print("expected cost R*P(fail) + T*b, R = {}, T = {}".format(R, T))
    print("  level    q    " + "  ".join("{:>4}".format("b=%d" % b)
                                          for b in range(N + 1)))
    for name, q in LEVELS:
        row = "  ".join("{:.2f}".format(cost(b, q)) for b in range(N + 1))
        print("  {:<7} {:.2f}  {}".format(name, q, row))
    print("best b and failure rates")
    for name, q in LEVELS:
        best = min(range(N + 1), key=lambda b: cost(b, q))
        print("  {:<7} best b {}  fail at best {:.3f}  at b=0 {:.3f}"
              "  at b=6 {:.3f}".format(name, best, 1.0 - success(best, q),
                                       1.0 - success(0, q),
                                       1.0 - success(N, q)))

Single source and citation

Fan, Y. (2026). Step-Down Communication Protocols for High-Clock-Speed Minds: Cognitive-Bandwidth Matching, Inference-Gap Bridging and Zero-Friction Alignment (Version 1.0.0), Chapter 3: Expertise Reversal: The Best Number of Bridging Steps Depends on the Receiver's Chunk Dictionary. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-431 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026431,
  author    = {Fan, Yafu},
  title     = {Step-Down Communication Protocols for High-Clock-Speed Minds: Cognitive-Bandwidth Matching, Inference-Gap Bridging and Zero-Friction Alignment},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book