5D Library › Algorithmic Marketing, Multi-Touch Shapley Attribution and Stochastic Control of Acquisition Funnels · 中文
Chapter 7 Reading the Classics: Shi and Jie in the Sunzi
This chapter reads the earlier discussion of timing, budget allocation and marginal effects beside two passages on shi (势, often rendered “energy”, “momentum” or “strategic disposition”) in the Sunzi: the sentence “势者,因利而制权也” in “Initial Estimates” (计篇), and in “Strategic Dispositions” (势篇) the two similes of “rushing water” and “the hawk” together with “his shi is steep, his jie short”. The aim is not to show that Sunzi understood optimal control or advertising response functions, but to see how an ancient military treatise describes “storing” and “releasing”, “disposition” and “timing”, and to mark clearly the limits between ancient military experience and modern mathematical models. Quotations follow Yang Bing’an’s Shiyijia zhu Sunzi jiaoli (Zhonghua Shuju, 1999) (曹操 等 注,杨丙安 校理 1999) and are cited only by chapter; the wording was also checked against common online transcriptions (the Siku edition texts of the Sunzi and the Sunzi jizhu on Daizhige, the bilingual text on the YellowBridge website, search results from the Chinese Text Project and others), with punctuation following common practice. English renderings are the author’s own; Lionel Giles’s annotated translation of 1910 was consulted (Giles 1910).
“Initial Estimates”: shi is shaping plans according to advantage
In plain terms: if (the general) heeds the calculation of advantage, he should create a favourable disposition for it, to support (the plan) in its execution in the field. What is called shi is adapting one’s measures flexibly according to what is advantageous.
Among the Eleven Commentaries, Cao Cao notes “control comes from flexibility, and flexibility is shaped by circumstances” (制由权也,权因事制也), and Zhang Yu “one must follow the advantage of the situation and shape it into stratagems to defeat the enemy” (须因事之利,制为权谋,以胜敌耳). Shi here is not a fixed force but an arrangement that changes with “advantage”; quan (权) is weighing and adapting. Book 26 read the “temple calculation” of “Initial Estimates”, which concerns probability judgement before the event; here only this sense of shi is taken: the disposition is shaped by the calculated advantage, rather than a disposition coming first and being adapted to.
“Strategic Dispositions”: rushing water, the hawk and the drawn crossbow
Source text
激水之疾,至于漂石者,势也;鸷鸟之疾,至于毁折者,节也。是故善战者,其势险,其节短。势如彍弩,节如发机。(曹操 等 注,杨丙安 校理 1999)
(Sunzi, “Strategic Dispositions”)
In plain terms: rushing water flows so swiftly that it carries stones along: this is shi. A bird of prey strikes so swiftly that it maims and breaks its quarry: this is jie (节, “timing”, “measured release”). So the good commander’s shi is steep and his jie short. Shi is like a fully drawn crossbow; jie is like pulling the trigger.
Source text
故善战者,求之于势,不责于人,故能择人而任势。任势者,其战人也,如转木石。木石之性,安则静,危则动,方则止,圆则行。故善战人之势,如转圆石于千仞之山者,势也。(曹操 等 注,杨丙安 校理 1999)
(Sunzi, “Strategic Dispositions”)
In plain terms: so the good commander seeks victory in shi and does not demand it of his men, and therefore he can choose men and make use of shi. One who makes use of shi sends his men into battle as one rolls logs and stones. It is the nature of logs and stones to stay still on level ground and to move on a slope, to stop if square and to roll if round. So the shi of those good at battle is like rolling a round stone down a mountain a thousand fathoms high: this is shi.
Wording and variants. The author’s framework notes “鸷鸟之疾 (also written 击)”. In this round of checking, the Siku-edition texts of the Sunzi and the Sunzi jizhu on Daizhige both read 鸷鸟之疾, and the YellowBridge bilingual text also reads 疾, while the Chinese Text Project transcription reads 鷙鳥之擊. Among the Eleven Commentaries, Zhang Yu writes “when hawks seize sparrows, they must measure the distance and wait closely before they strike (后击), and so can break their prey”, and Du Mu “jie means measuring the distance before seizing, and so it can break things”; the commentaries contain the idea of “striking”, but the main text reads 疾. The reading 击 can be traced to an encyclopaedia: in his note on section 13 of chapter V, Giles writes “For 疾 the Yü Lan reads 擊”, that is, the quotation in the Song Taiping yulan reads 击 (Giles 1910). The main text follows the Eleven Commentaries in reading 疾, and the author is asked to confirm against the collation notes of Yang Bing’an’s edition. Other variants: the author’s framework has “故善战者”, while every transcription checked (including the Chinese text accompanying Giles) reads “是故善战者”, and 是 is restored; the text accompanying Giles writes 势 as 埶; 彍 (“to draw a crossbow fully”) in “势如彍弩” appears in several online transcriptions as the faulty simplification 扩, and some read 张弩 (Du Mu’s gloss “彍,张也” could not be checked online this round and awaits the printed edition); the Chinese Text Project transcription reads “节如机发” and “故能择人任势” (without 而); in “求之于势,不责于人”, the Siku Sunzi transcription reads “不责之于人”, and the book follows the common text. All of this has been checked only against online transcriptions and awaits the printed edition.
How the commentators read jie. According to the Daizhige transcription of the Sunzi jizhu, Cao Cao glosses “险犹疾也”, reading the “steep” of “his shi is steep” as “swift” (to be checked against the printed edition). Du Mu explains jie as “measuring the distance”, striking at the right range; Giles, in his notes, leans towards reading it as a kind of restraint, holding back until the right moment (Giles 1910). Both readings point to timing: the bird of prey is not in constant high-speed motion, but strikes with full force only in the last instant before reaching its quarry.
Rereading in the book’s language: an analogy
What follows is an analogy, not an interpretation of what Sunzi meant. Think of “storing force” in marketing as a depreciating goodwill stock: Nerlove and Arrow treated advertising as investment in goodwill that decays by a fixed fraction each period (Nerlove and Arrow 1962). Let goodwill on day be , the square root expressing diminishing returns to a single day’s spend; a 30-day budget is fixed, and on day 30 a promotional event (a big live-stream sale, say) converts a fixed fraction of that day’s goodwill into sales. How should the budget be spread over the 30 days?
This is the author’s model. Lagrange multipliers give the solution directly: spending on day is proportional to , rising towards the event day; the faster the depreciation, the more of the budget should be pushed into the last few days.
Calculation
# THE AUTHOR'S MODEL, an analogy for Sunzi's "shi" (stored force) and "jie" (short,
# well-timed release). A goodwill stock (Nerlove & Arrow 1962) decays by a fraction
# delta a day: G_{t+1} = (1 - delta) G_t + sqrt(A_t) (diminishing daily returns). A
# 30-day budget B is spent to maximise the stock on day 30, when a one-day sales event
# converts a fixed share of it. Lagrange: A_t proportional to (1 - delta)^(2 (30 - t)),
# checked by a numerical search. Output: share of the budget spent in the last 7 days
# and the stock relative to spending evenly, for slow and fast decay.
import math
DAYS, B = 30, 30_000.0
def stock(A, delta):
G = 0.0
for a in A:
G = (1 - delta) * G + math.sqrt(a)
return G
def optimal(delta):
w = [(1 - delta) ** (2 * (DAYS - 1 - t)) for t in range(DAYS)]
return [B * x / sum(w) for x in w]
def search(delta, sweeps=200):
A = [B / DAYS] * DAYS # pairwise transfers keep the budget fixed
step = B / DAYS / 2
while step > 1e-3:
improved = False
for i in range(DAYS):
for j in range(DAYS):
if i != j and A[i] >= step:
C = A[:]; C[i] -= step; C[j] += step
if stock(C, delta) > stock(A, delta) + 1e-12:
A, improved = C, True
if not improved:
step /= 2
return A
if __name__ == "__main__":
print("daily decay last 7 days' share of budget stock: optimal / even"
" search check")
for delta in (0.01, 0.05, 0.15, 0.40):
A = optimal(delta); even = [B / DAYS] * DAYS
share = sum(A[-7:]) / B
chk = None
if delta in (0.05, 0.40):
chk = stock(search(delta), delta) / stock(A, delta)
print(f" {delta:5.2f} {share:6.1%} "
f"{stock(A, delta) / stock(even, delta):5.3f}"
+ (f" {chk:.4f}" if chk else ""))
print("half-life of goodwill (days): " + ", ".join(
f"{d}: {math.log(0.5) / math.log(1 - d):.1f}" for d in (0.01, 0.05, 0.15, 0.40)))
# daily decay last 7 days' share of budget stock: optimal / even search check
# 0.01 29.0% 1.004
# 0.05 53.7% 1.091 1.0000
# 0.15 89.7% 1.572
# 0.40 99.9% 2.739 1.0000
# half-life of goodwill (days): 0.01: 69.0, 0.05: 13.5, 0.15: 4.3, 0.4: 1.4With goodwill depreciating 1 % a day (half-life 69 days), the last 7 days take only 29.0 % of the budget, and the result hardly differs from even spending (goodwill on the event day is only 0.4 % higher); at 5 % a day (half-life 13.5 days), the last 7 days take 53.7 % and goodwill is 9.1 % higher than with even spending; at 15 %, the last 7 days take 89.7 % and goodwill is 57.2 % higher; at 40 % (half-life 1.4 days), almost the whole budget falls in the last week and goodwill is 2.739 times that of even spending. A numerical search agrees with the analytic solution at both depreciation rates checked (ratio 1.0000).
In this analogy, “steep shi” corresponds to concentrating investment just before the event when depreciation is fast, and “short jie” to the narrow window of release; “shi is like a fully drawn crossbow”: once drawn, it must be released soon, or the tension (goodwill) drains away. But the limits of the analogy are just as clear. First, when depreciation is slow, the optimum is close to even spending, and “storing force for the strike” brings no benefit; the author’s framework treats “storing force, then pulsed conversion” as a general principle, and the model does not support that. Second, the whole-life model of Chapter 6 (with word-of-mouth diffusion) points the opposite way: front-load, invest as early as possible. The two conclusions come from different assumptions (a one-off event against continuing diffusion), and which applies depends on the actual goodwill depreciation rate and word-of-mouth coefficient, which have to be measured. Third, Sunzi discusses force and timing between two armies, where the enemy reacts; marketing’s “opponents” are competitors and platform algorithms, which also react, and this chapter’s model ignores that entirely.
Structural comparison and its limits
| Classical text | Modern concept | Limit of the comparison |
|---|---|---|
| “Initial Estimates”: “势者,因利而制权也” | Adjusting spend to measured marginal returns | Sunzi means a general’s judgement of battlefield advantage; today marginal effects must be measured by experiment |
| “激水之疾,至于漂石者,势也” | Accumulation of a stock (goodwill, word of mouth, brand search volume) | Rushing water is a physical metaphor; the goodwill stock is an abstract variable in the author’s model |
| “鸷鸟之疾,至于毁折者,节也”, “其节短” | Conversion concentrated in a short window (big sales, live streams) | Commentators read jie as “measuring distance” or as “restraint”; the model takes only “concentration in time” |
| “势如彍弩,节如发机” | Timing of investment and of release | The model shows clear gains from concentration only when depreciation is fast |
| “求之于势,不责于人” | Relying on channels and diffusion structure rather than on the conversion of each single touch | In antiquity an organisational principle of war; today the role of structure must be tested by randomised experiment |
These comparisons are structural resemblances only. First, the Sunzi is a distillation of experience and principles about war, not a treatise on commercial competition; applying it to marketing is itself a metaphor. Second, the ancient texts contain no stock–flow equations, depreciation rates or optimisation, and it cannot be said that Sunzi “already discovered” optimal control or the dynamics of advertising. Third, the framework’s “dynamic control of storing force in the funnel and pulsed conversion” receives only conditional support in this book: it holds in a model with fast depreciation and a one-off event, and not in a model with word-of-mouth diffusion.
Summary
“势者,因利而制权也” in “Initial Estimates” and, in “Strategic Dispositions”, “激水之疾”, “鸷鸟之疾” and “是故善战者,其势险,其节短。势如彍弩,节如发机” agree with the common online transcriptions (the framework omits 是, now restored) and await confirmation against the printed edition. The framework’s note that 疾 “is also written 击”: the Eleven Commentaries transcriptions read 疾, and according to Giles the Taiping yulan quotation reads 击; the collation notes of Yang Bing’an’s edition should be checked. The author’s goodwill-stock model shows that “storing force, then a short release” beats even spending clearly only when goodwill depreciates fast; when it depreciates slowly the two hardly differ, and in markets with word-of-mouth diffusion the optimal direction is even reversed. These are analogies, not ancient foresight of modern marketing theory.
References
Fan, Y. (2026). Algorithmic Marketing, Multi-Touch Shapley Attribution and Stochastic Control of Acquisition Funnels: From Thompson Sampling Bandits and Absorbing Markov Chains to BG/NBD Customer Lifetime Value, Read Beside the Strategic Dispositions of the Sunzi [算法营销、多触点沙普利归因与获客漏斗的随机控制] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23836264 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).