This page answers:Should I do this myself, or should this be handed to someone else?
P1 Can you spell out the deliverable for these things (what you produce / what standard counts as good / how long it takes)
P2 Can you say for each one “if I do this, I have no time for what?”
P4 Do you have at least 1 channel that turns this relative advantage into money (customers / platforms / employers / partners)
P5 Which kind of gap is this (it sets the typical time to disappear, §4 Step4)
How much longer the gap is expected to last (months)
Time you need to break even (months)
This page answers “which thing should I sell?”, not “am I strong?”. The algorithm comes from 《04-比较优势算法.md》:
list the candidates (§4 Step1) → compute each one's give-up cost and its relative advantage = my give-up cost ÷ the cost the reference party pays to do the same thing (§4 Step2; <1 is your comparative advantage)
→ then, per §4 Step3, see where you rank among those 3 reference parties (1st — you can trade it outward / 2nd — only good as part of a combined play / 3rd — just buy someone else's)
→ the window test (§4 Step4: an information gap lasts weeks–months / a skill gap 1–3 years / a position-and-resource gap several years / a cognition gap, long term)
→ and finally, per §4 Step5, the conclusion: make it your lead / build the channel first / buy someone else's. Do not swap those two columns in the table: the second column is “how much I give up by doing this” (express it as your hourly rate or your per-job income), and the third is “what it costs the reference party to do the same thing”;
a ratio <1 = you give up less = you hold a comparative advantage on that item. Sorting convention: the table is sorted by “my give-up cost” ascending (= lowest opportunity cost first, §4 Step2), not by the row order you typed. The reference set must be specific people: §9 says explicitly that an “industry average” cannot be used as the reference — which is why the third column requires ≥3 real names; with fewer than 3, P3 fails outright.
This page is a conditions-preparation tool, not investment advice. It only computes the ranking and the conclusions from the document's formulas and thresholds; it does not judge whether you are strong or not.
Every threshold and decision (≥3 deliverable things, ≥3 specific reference parties, relative advantage <1, the three treatments for rank 1/2/3, window > payback period)
comes from the internal methodology document 《04-比较优势算法.md v1.0》 (on the algorithm-port machine under /opt/inventory/econ_algorithms/, read-only, unchanged in this round),
and has not been through external empirical testing — these are internal practice rules, not academic findings. Where it is not aligned / has been simplified (stated plainly)
· ① The second half of §4 Step2 in the document gives a ranking method (when there are no exact figures, hand-rank the 3 candidates by “what you would least want to give up”; last place = lowest opportunity cost) — this page does not implement the ranking method and only takes the numeric route; when a figure is missing, that row cannot compute a relative advantage and shows “—”.
· ② The document asks for “relative advantage <1 marked in green”; this page uses the existing table renderer, which has no colour, so it uses the text marker “✅ comparative advantage” instead and keeps the numeric relative advantage.
· ③ “Where I rank” is typed in by you (1–3); the page does not infer the ranking for you — the reference set the document means is those 3 specific people you are going to trade with, and the page has no way to know that.
· ④ If “relative advantage <1” and “rank 1” give inconsistent conclusions, this page shows both and does not adjudicate (§4 Step2 and §4 Step3 are two parallel conventions and the document does not say which wins). Unconfirmed
· Unconfirmed ①: the window in §4 Step4 is qualitative (“weeks–months / 1–3 years / several years / long term”) and gives no number of months; “how many more months can the gap survive” is yours to estimate — the page only hands you the document's qualitative tiers and offers no default.
· Unconfirmed ②: §3 P2 only requires that you “can say what you have no time for if you do it”; the document does not specify whether the second column must be filled in as “¥/hour” or as “income per job”; this page accepts both (as long as there is a number).
· Unconfirmed ③: §3 P1's “≥3 items” and §4 Step1's “3–8 items” are both in the original, but it does not say what to do past 8; this page does not block it — however many rows you fill in is how many it counts.
[Evidence section · where every number can be checked] (1) Source: the internal methodology document 《04-比较优势算法.md》 v1.0 (on the algorithm-port machine at /opt/inventory/econ_algorithms/04-比较优势算法.md, read-only): start conditions P1–P5 = §3; decision rules = §4 Step1–Step5; the preparation list and first step = §5.2 / §5.3; external data = §6. §1 of the document gives its literature sources (Ricardo 1817, DOI 10.1017/cbo9781107589421; the law of one price, Lamont & Thaler 2003, DOI 10.1257/089533003772034952). (2) Three categories
· [Verifiable] The give-up cost, relative advantage = my give-up cost ÷ the reference party's cost, the <1 mark for comparative advantage, the three treatments for rank 1/2/3, the four window types, the “make it your lead / build the channel first / buy someone else's” of §4 Step5, and P1–P5 — all of it lines up with §3 / §4.
· [This page's display convention] The table is sorted ascending by “my give-up cost” (following the sorting convention of §4 Step2, not the row order you typed); “mark <1 in green” is replaced by the text “✅ comparative advantage” because the table renderer has no colour — that substitution is the page's.
· [No source found] The four window tiers “weeks–months / 1–3 years / several years / long term” are qualitative in §4 Step4 with no months given → no quantitative basis was found (this page gives no default); and which one takes priority when “relative advantage <1” and “rank 1” conflict is not written in the document → unconfirmed. (3) How to verify it yourself: list 3 things, with my give-up cost 100 / 300 / 500 and the reference cost 200 / 200 / 200 → relative advantage 0.5 / 1.5 / 2.5 → only the first is <1, so it should be marked ✅ comparative advantage, and it should come first (lowest cost).
Generated by /opt/slices_build/build.py from the spec /_shared/slices/comparative-advantage.json — to change this page, edit the spec and re-run the generator; never hand-edit the HTML.