--- name: asb-voters description: "Distills a person's or company's strengths, experiences, and obsessions down to their one or two VOTERS — the decisive, idiosyncratically extreme traits that win the contest because customers who value them accept every other trade-off. Runs each candidate through a hard gauntlet: extremity earned through obsession (not mere competence), rarity among peers, decisiveness (name the weaknesses it overpowers), and reverberation (it must force decisions in product, pricing, and market). Caps the answer at two, records survivors in VOTERS.md — plus the near-miss 'special strengths' to deploy — and delivers the honest zero-voter verdict when nothing is extreme yet. Takes a self-portrait file (M-numbered), strengths or keystones charts, or a live capture. Load when the user asks what makes them special, what to bet the strategy on, their unfair advantage or superpower, or to 'find our voters.' Do NOT load to build the full self-portrait (the previous step) or to design the whole strategy — this file feeds it." --- # Voters: The One or Two Things That Win the Whole Game In high-school debate, the final speaker gets three minutes against the opponent's seven. Answering every argument is impossible — so the winning tactic, "the Voter," reframes the whole debate: it all comes down to one or two key points that ought to decide the outcome — chosen, of course, where you are strongest. Win those and you win the debate, despite losing everything else. Business works the same way: one or two decisive strengths override a parade of deficiencies, because customers who value those voters accept every other trade-off. This skill distills everything true about you — or your company — down to those one or two, and refuses to stop at a comfortable list of five. ## The mental model ### What a voter is A voter is not a strength. Strengths are common; voters are the one or two traits held at such an idiosyncratic extreme that they decide the contest. Steve Jobs's voter was industrial design — obsession down to the layout of components no customer would ever see — and it made the iPhone dominant *despite* iTunes being famously terrible. The voter didn't fix the weakness; it *overpowered* it. Likewise the founder whose product became the fastest in its market: his voter was caring about performance far beyond what anyone else would invest — a twenty-year obsession that began with optimizing math functions for fun in high school — and the whole strategy (top-end hardware, premium price, banned slow add-ons, segments walked away from) reverberated outward from it. ### The bar — every voter passes all of it 1. **Extremity, earned.** Not competence — excess. "Magic is just someone spending more time on something than anyone else might reasonably expect." The evidence is a history: years of obsession, care beyond what the job required, depth nobody asked for. A trait acquired last quarter or aspired to next year is not a voter. 2. **Rare among peers.** Plenty of people know the platform; plenty of founders like charging more for quality. Test: would a well-informed peer, asked "who is THE person/company for X?", name you? If the honest answer is "you and forty others," it isn't rare. Rarity means the *trait's* actual scarcity, not the market's awareness of it: a buried eleven-year obsession nobody has heard of is still rare — the missing reputation is an investment plan, not a failed test. When reputation IS the claim, ask for the citable artifact (the post peers link, the thing you're known by), not self-report. Confirm the current competitive field with your search tools; training knowledge is stale and may miss the peers who make a trait common — or the gap that makes it rare. 3. **Decisive — name what it overpowers.** A voter proves itself by the trade-offs customers accept for it: the terrible companion software, the banned add-ons, the walked-away-from segments, the premium price. If you can't name a real cost that people who value it happily eat, the voter is unproven — possibly imaginary. "Customers" means anyone who *chose* the user: a buyer, but also an employer, a client, a team, a collaborator, a community. Payment is not required; a real choice that accepted a real trade-off is ("the team put up with my missed deadlines because I was the one who could make the demo work"). At the highest intensity, the ideal customer reads your weaknesses as *proof* of the voter (on a handmade marketplace, an imperfection is evidence a human made it, and the high price reads as supporting the artist). 4. **It reverberates.** A real voter forces second-order decisions — in product, pricing, market, hiring, what you refuse to ship. If naming it changes nothing downstream, it's a slogan, not a voter. 5. **It energizes.** A voter is executed for years at an intensity nobody sustains through discipline alone; one the user dreads wielding will collapse, however real the extremity is today. In company mode, run this test against the human(s) who actually carry the obsession; a company voter carried by nobody in particular is suspect. ### Three tiers below the bar The debate tactic works because a judge can hold one or two points; so can a market. Three voters is a feature list wearing a costume — so at most two survive (**V1, V2**), and everything else lands in one of three tiers: - **Special strength** — a near-miss: genuinely special (top-1%, real extremity, real energy) but missing one decisive test — usually *fairly* rare but not THE-person rare, or no real cost ever eaten for it. Never inflate it into a voter; never bury it as common either. It is a secondary pillar the strategy should deliberately deploy. The *standout investment candidate* (extremity real, decisiveness merely untested) is the classic case. - **Shaping strength** — true and strategy-relevant but common or shallow (real familiarity, ordinary diligence). It points the strategy; it doesn't make you special. - **Cut** — virtue-words, aspirations, and evidence-free claims. ### Special, not best — and the zero-voter verdict A voter does not need to be objectively superior; it needs to be *yours*. A "refuted" chess opening beats prepared opponents at your level precisely because nobody prepares for it — competitors who think your strategy is wrong are evaluating the wrong player, and their disbelief is why they won't copy you. Better than different is *extreme*: not just a minimal interface but so minimal it runs on the command line. So when no candidate passes the bar, say it plainly: **nothing is special-and-extreme yet** — never pad the file with shaping strengths relabeled as voters. The response is twofold: invest the deepest *genuine* obsession on the table into extremity (extremity is a choice), or change games to one where a fact you already hold IS extreme. For anyone early in their career — or new to this kind of work — zero voters is the *expected* result, because extremity is earned over years and proved by people choosing you. "Nothing is extreme; go work and see" is true and useless. So the verdict always comes with **the bet**: one chosen obsession (or change-games target) and a concrete, dated plan to turn it into a voter. The verdict is honest; the bet is what makes it worth having. ### Drains are costs in the plan, not walls around it When the bet or a voter's proving trade-offs require an activity the user finds draining — cold outreach for a founder with no audience, selling for a solo consultant — keep it in the plan with a budget (bounded, scheduled doses) and a training step (practice and a system make most drains cheaper). Route around it only when a real alternative keeps the goal intact. Don't build the strategy on becoming excellent at a drain; don't drop a necessary activity because it costs energy either. ## The distiller's posture ### Be clear, not clever Write to be understood, not admired. The work here wrestles with hard concepts, and clever metaphors, wordplay, or cute turns of phrase make them harder to grasp, not easier. Say plainly what you mean. If a sentence reads more clearly without a flourish, cut the flourish. State the actual point rather than gesturing wittily at it. ### Restate references; never cite a bare token When you mention a numbered or lettered item to the user — K4, W2, O17, H3, and the like — add a few plain words on what it actually is ("K4 — the owner whose career rides on the site"). A bare token is unreadable to a human who saw it defined hours or days ago: the tag is for traceability, the gloss is for comprehension. Keep the tag for accuracy; always add the gloss. ### Generate willingly; gate mercilessly Harvest candidates liberally — from the self-portrait file, the strengths charts, the user's war stories, and your own reading of their material ("this twenty-year pattern in M3 smells like a voter — defend it"; a keystones file is mined indirectly: the strengths its keystones cite, and the extremity behind them). Nominated virtue-words enter the harvest like anything else — they die at the gauntlet, on the record, or yield the specific fact underneath. Then run the gauntlet one candidate at a time, all five tests visible, pressing every soft spot: "everyone in your market claims craftsmanship — what's the evidence a peer would name YOU?"; "what have you actually refused to ship because of this?"; "what did this obsession cost you that a reasonable person wouldn't have paid?" A candidate the user is stuck defending gets your best steelman first, then the honest kill. Two bookkeeping notes: when the evidence pressed out of a dying candidate really belongs to another (the "craftsmanship" facts were all about the tooling stack), cut the word and transfer the evidence — record the transfer in both places; and a candidate may be *cut-merged* into another (a drive that is just the fuel of an existing voter) — cut list plus evidence, no independent entry. ### This skill requires the press Distillation to one or two is where users flinch — everything feels essential, and vague answers ("quality," "our people") feel safe. If a devil's-advocate interrogation skill is installed in the environment (for example *Rude Q&A* / `asb-rude-qa`, from the same author as this method), invoke it twice: per-candidate whenever a verdict is soft — resting on a single piece of evidence, or on self-report with no citable artifact — and against the final file with this brief: *attack these voters — find every one that is a common strength wearing an extreme costume; every one with no named cost it overpowers; every one whose evidence is aspiration rather than history; every one that decides nothing downstream; and attack the count — if two survived, argue one should die; if the file dodged the zero-voter verdict, or wrote it without a concrete, dated bet, say so; if the plan drops a necessary activity only because it drains the user, say so.* If no such skill is available, run exactly that interrogation yourself, visibly, in both places — the method does not work without it. ### The user owns their history; the bar owns the verdict What happened — the years, the episodes, the costs paid — is the user's record, taken at their word after one honest press. But whether a candidate passes the bar is not a vote: however attached the user is, a candidate that fails a test goes to its tier with the reasoning on the record. Tone stays warm: the refusal is of the classification, never of the trait, which remains true and useful. The wielder never invents history — the evidence in the file is the user's, confirmed. Every test runs on what the user actually said; never mark a test passed from your impression of them ("from how you talk about it, the energy is real") — ask. The closing press may probe whether the evidence is representative (one clarifying round), but it re-attacks classifications, not the user's settled facts. ## How to use this skill ### Phase A — Ingest Ask what this distillation is for — a person choosing what to build, or a company anchoring a strategy — and what exists on disk: a self-portrait file (characteristics numbered M1, M2, … with evidence and contexts), a classified strengths chart or keystones file (S/K numbers), interview or customer-quote material. Read whatever exists and cite numbers in the candidates. If nothing exists, say plainly that the distillation is sharper after the self-portrait step (a companion skill from this method's author produces it — `asb-who-me`, if installed; for a company, an observation-and-classification method ending in a strengths chart), then proceed: capture candidates inline, each with at least one concrete episode, caveated as unvetted. If a `VOTERS.md` exists at the target location, read it first: in-progress header = resume where the pointer says; finalized = ask whether to re-distill (a re-run should have a reason; a bet's checkpoint arriving is a good one). In person mode, establish the track record before the harvest: years working, and whether anyone has yet paid for or chosen the user for this kind of work (the self-portrait's preamble may already say). If it's thin, say plainly up front: "voters are earned over years and proved by people choosing you, so at your stage the likely result is zero voters *yet* — that's normal, not a judgment. Either way you leave with a concrete plan: which trait to build, the goals that would prove it, and what to do this week." Then run the gauntlet at full strength — the bar doesn't lower; the framing and the deliverable change. Output: `VOTERS.md`, same directory as the primary input file (the self-portrait if one exists, else the strengths chart); with no input files, ask where the method's files should live before creating it (default: the current directory). ### Phase B — Harvest candidates Collect every plausible candidate into a visible list — the user's nominations plus your own from the files, each tagged with its source ([M3], [S1], war story) AND, for anything captured live, its episode evidence in a line or two — evidence that lives only in chat dies with the session, and a resumed gauntlet run on re-invented evidence can flip a verdict. No gauntlet yet; the harvest is complete when the user has nothing to add and you've mined the files. Create `VOTERS.md` now with the in-progress header carrying the full candidate list (sources, episode lines, statuses) and pointer — the file is the memory, not the chat. If the fuller upstream path was offered and declined, record that in the preamble so a resumed session doesn't re-pitch it. ### Phase C — The gauntlet One candidate at a time, all five tests run visibly — across as many exchanges as the evidence needs. **One ask per message, at the end of the message:** when a test needs the user's evidence, ask for that one thing and stop; never stack tests ("answer 2 and 3 for me") or braid several questions together. Verdict on the record: **voter-candidate** (written into The voters at verdict time with its V-number; if Phase D's ranking later demotes it, it moves to its tier with a "demoted at ranking" note and its number is never reused), **special strength** (with the test it missed), **shaping strength** (one line of reasoning), or **cut** (dies in the header's candidate list with its reasoning, not in the body). Press per the posture; invoke the interrogation on soft verdicts. Update the file and pointer each exchange — the in-flight candidate's status reads "in gauntlet," and the pointer records which tests have run, their outcomes, and the evidence the user gave, so a died session resumes mid-candidate instead of restarting it. (Resuming a file whose pointer lacks that detail: say so plainly and restart the candidate from test 1.) ### Phase D — Distill and close - **Count the survivors.** Three or more: force the ranking — "when they conflict, which wins? which one or two do you bet on?" — demote the rest to special or shaping strengths, per how special each is. One or two: proceed. Zero: write the verdict section — the plain statement plus the twofold response, naming the specific obsession candidates from the harvest. The standout investment candidate appears in BOTH special strengths (so tagged) and, by name, the verdict. Then build **the bet** with the user — mandatory; the session is not done without it: 1. *Choose one* obsession to invest in (usually the standout investment candidate) or one change-games target — not a menu. If the user can't choose, force it: "which one would you still be doing in five years if nobody paid you?" 2. *Goals tied to the failed tests* — two to four, each with a number and a date (default horizon: 90 days). Extremity: a body of work beyond what anyone would reasonably expect ("ship twelve small tools in this niche"). Rarity: a citable artifact peers can point to. Decisiveness: a number of people who *choose* the user — paying, hiring, committing time — despite a named trade-off. Reverberation: one real decision the bet forces. 3. *This week* — two or three first actions, small enough to start today. 4. *Drains in the plan* — budgeted and trained, per *Drains are costs in the plan*. 5. *The checkpoint* — the date to re-run this gauntlet, and what result would mean switching branches (invest ↔ change games). Designing the full company strategy around a bet is downstream; the bet's plan is this skill's job. - **Walk the weaknesses** (for each settled voter): from the self-portrait's hindrances, the strengths chart's weaknesses, or the user's list — which does this voter overpower, and which would the ideal customer read as *service to the voter*? Record them under the voter. A voter that overpowers nothing gets one more press before it stands. - **The closing press** (delegated or self-run, per the posture — including the count attack). - **Finalize:** remove the in-progress header (cuts and their reasoning move to a brief "Cut at the gauntlet" list), and close with the handoff. In a zero-voter file the handoff is the bet itself — start this week's actions, return at the checkpoint; there is no strategy to audit against voters that don't exist yet. Otherwise: audit the current strategy against the voters — walk pricing, positioning, product, and channels asking where each voter expresses and where it's contradicted; the mismatches are where to pivot. Expect others to call the strategy wrong — their disbelief is why they won't copy it. If a positioning skill from this method's author is installed (for example *Positioning* / `asb-positioning`), note that its statements should lead with the voters and deploy the special strengths as secondary claims. ### The file structure Format for a human reading it in a markdown viewer, not as a wall of text. Every voter is an H3 with a short name, then a **bold one-sentence headline** (the voter itself), then a one-line supporting note, then the bar rendered as **labeled bullets** — never a single multi-line indented paragraph, which collapses into an unreadable blob (and 4-space indents render as code blocks in many viewers). The same labeled-bullet discipline applies to the strengths sections. ```markdown # Voters — > ⚠️ IN PROGRESS — candidates harvested: a one-line episode note for live-captured evidence, and status: > pending / in gauntlet / voter-candidate / special / shaping / > cut (reason)>; > currently at the gauntlet: outcomes so far, with the evidence given; note any soft verdict a > resumed session should re-press>. complete; ranking / weakness walk / closing press / finalization > remaining."> If you are resuming, continue there. (Removed at > finalization; episode evidence worth keeping folds into the > surviving entries first.) ## The voters ### V1 — **** ** **Draws on:** - **Extremity, earned** — . - **Rarity** — . - **Overpowers** — the trade-offs its fans happily eat: - ** — . - ** — <…>. - **Reverberates into** — . - **Energizes** — . ### V2 — (only if a second genuinely survives) ## Special strengths (near-miss voters — deploy as secondary pillars) - — . [M/S] ## Shaping strengths (true, useful, not voters) - — . [S1] ## The verdict (only when zero voters survive) ## The bet (required whenever the verdict is written) **** - **Goals by ** — - *Extremity:* . - *Rarity:* . - *Decisiveness:* >. - *Reverberation:* . (Only the tests the bet actually failed; two to four goals.) - **This week** — . - **Drains to budget** — . - **Checkpoint** — . ## Cut at the gauntlet - — . ## Next steps ``` V-numbers are stable once written; wording may sharpen under a frozen number with the user's confirmation while in progress. ## Refusal conditions - **A third voter, however insisted.** The third demotes to a special or shaping strength with the ranking reasoning on the record. - **Misfiling a near-miss — either way.** Inflating a special strength into a voter, or burying it among the common shaping strengths, both lose its value. - **Virtue-words as voters.** "Quality," "hard work," "customer focus," "our people" — nothing every peer would also claim can decide a contest. Press to the specific fact underneath or classify it out. - **Aspiration as extremity.** "We're becoming known for…" is strategy, not history. A voter's evidence is what already happened. - **A voter with no overpowered cost.** One more press, then special or shaping strength. - **Dodging the zero-voter verdict — or giving it without a bet.** Padding the file with relabeled shaping strengths poisons every decision downstream; "go work and see" leaves the user with nothing to do. A zero-voter file always has the bet: one choice, dated goals, this week's actions, a checkpoint. - **Drains as walls.** Never steer the bet or strategy away from a necessary activity only because it drains the user. - **Building the full self-portrait.** That's the previous step; capture-inline is a fallback, not the exercise. - **Designing the whole strategy.** This file anchors the strategy audit; running it — resolving conflicts, redefining boxes, changing target markets — is downstream work this skill names but does not perform.