Methodology

This is not an academic grade. “Adequate” is not a passing C. This is a fitness-for-the-seat scale measured against a fixed standard. The median is failing by design, the bar is the seat, not the average officeholder. The F-line sits at 5.0, not the conventional 60%.

The fourteen measures (M01–M14)

Each scored 0–10 against documented historical anchors from both parties across two centuries. M15 (Functional Capacity) is a dormant exception-flag, not a routine score.

The Constitutional Weight Schedule

The composite is not a flat average. It is weighted so the duties the oath ranks highest carry the most weight. M03 was moved to the Core domain (“Persons of Equal Worth”) because equal personhood is upstream of the consent of the governed. The three domains with no built measures were dropped and the schedule renormalized, every input is a real, displayed, anchored cell; no estimated or invented inputs.

Same schedule for every Republican, Democrat, and Independent, published, fixed, reproducible by any reviewer. We weight what the oath ranks, not the outcome.

The formula, published, identical for everyone

This is not a curve. No one is graded against other politicians, everyone is measured against the same fixed standard (the seat, the oath), and the bar never moves. The top scorer does not set the grade; the standard does. Two corrupt senators do not make corruption “average.” Only your own documented conduct changes your score, and the computation is the same for every Republican, Democrat, and Independent:

composite = Σ ( measure_score × measure_weight ) ÷ total_weight

  Civic Realism Core   29.4%   M01, M02, M03
  Fiduciary Conduct    23.5%   M06, M10, M11
  Power Dynamics       17.6%   M04, M07
  Honesty & Integrity  11.8%   M09, M13
  Discourse & Rhetoric 11.8%   M05, M12
  Character & Empathy    3.5%   M08
  Cognitive Capacity    2.4%   M14

credit score = composite mapped onto 300–850 (bands inclusive at lower bound)
support       = credit ≥ 700

Hand anyone a politician’s fourteen scores and this formula and they will reproduce the composite exactly. That reproducibility is the point: you do not have to trust the result, you can check it yourself.

The character credit score

Bands are inclusive at their lower bound; the credit number is interpolated within the band. The ceiling is the standard the office demands, not the best officeholder we found. A perfect score is reachable in principle and almost never reached in practice, even our highest historical figures fall short. That headroom is the doctrine made visible: the seat asks more than anyone has fully given.

Evidentiary sufficiency, the burden is on the evidence, and it cuts three ways

The scorecard grades the documented record, not what was asserted. A claim is not a finding. The burden is on the evidence, always, and the evidence cuts in three directions, not two:

1. It clears the falsely accused. Unproven, uncorroborated, or contested allegations are not scoreable conduct. An allegation that surfaces late, lacks contemporaneous documentary support, or arrives at a moment of maximum leverage does not meet the bar, no matter who it would help or harm. We do not ruin people on accusation.
2. It convicts the documented-guilty. When there is documentation, a contemporaneous report, a recording, a sworn finding, a conviction, the conduct is scoreable and named plainly. We do not shield people from the record.
3. It brands the documented fabricator a liar. When the evidence affirmatively establishes that an accusation was false, that is not suspicion, it is a finding of record, as Tier-1 as a conviction, and the fabricator is a liar. The framework that has the spine to call documented corruption corruption has the same spine to call a documented fabrication a lie. Anchor: the 2006 Duke lacrosse case, the North Carolina Attorney General declared the accused players innocent (April 2007) and the prosecutor was disbarred. That is the evidence affirmatively establishing falsehood, not a case merely collapsing.

None of the three is editorializing; each is the record. Real victims with real evidence are scoreable; suspicion is not; and a proven fabrication is its own documented misconduct.

False accusation and the smear, scored harshly, campaign latitude notwithstanding

A documented false accusation is graded harshly, and most harshly when it is lobbed carelessly, the recklessness is itself the breach. It scores low on M13 (lying); and where an officeholder weaponized a knowingly false accusation through office or state power against a target, it crosses into severity-flag territory. Politicians have historically enjoyed wide legal latitude to say things in campaigns, but legal latitude is not score immunity. “It’s allowed” is not a defense the score recognizes; it is the same “everyone does it” the Doctrine of the Seat refuses. If you smear and you lie, your score suffers, the latitude the law grants the campaign trail does not extend to the standard the seat requires.

No imputation forward, no excuse backward

The instrument scores only documented conduct against the documented choices real contemporaries actually made, never opportunity, never imputed character. A figure escapes a flag when there is no conduct to score, not because they are judged a better person; that silence is integrity. A historical figure is judged not by today’s standards but by the best conscience of their own era, the people in their own time who faced the same choice and chose otherwise. The test that separates context from excuse: did the era deny them the capability, or did they choose against an available truth?

What this scorecard refuses

• Rhetorical inflation, no “Hitler,” no “hero,” when documented conduct does not match the claim, in either direction.
• Letting wealth excuse conduct, pre-office wealth is not a penalty; office-driven enrichment is the breach.
• Letting partisan alignment determine score, the same conduct produces the same score regardless of party.
• Causal claims the evidence does not support, conduct is documented on its face, not asserted as cause.
• Grading on a curve, a failing class does not make a failure average. The bar is fixed.

Active duty, not passive abstention, where most fail

The standard is not satisfied by not doing wrong. It grades affirmative conduct, what the officeholder actively did to uphold the standard, not merely what they refrained from. This is where the largest number of officeholders fail, because they mistake a clean passive record for a passing one:

• Calling it out, aggressively, including your own side. Silence in the face of a character violation is not neutral; it is a scored failure (M07). The duty is to name the breach, loudly, especially when it is your own team committing it. “I stayed out of it” is not a defense, it is the violation.
• Disclosing your own conflicts, before you are asked. Proactive disclosure of potential conflicts (over-compensation, a spouse's or staff's financial entanglements, anything that could compromise the trust) is an active fiduciary duty. Waiting to be caught, or technically complying only after exposure, fails it. The honest officeholder volunteers the appearance of impropriety and addresses it; they do not make the public discover it.
• Self-correction over self-protection. When wrong, the active duty is to say so plainly and fix it, not to wait it out, blame-shift, or issue a conditional “sorry if anyone was offended.”

Passive non-wrongdoing earns the middle of the scale at best. The top of the scale is reserved for those who actively defended the standard at cost, and the bottom is not only for those who breached it, but for those who watched a breach and said nothing.

Two systems, on purpose

The fourteen measures ask: did this person’s conduct meet the standard? The Four Pillars ask a different question: is this someone worthy to be followed at all? They can diverge, competent conduct can sit beside hollow character. When they disagree, the disagreement is the finding. Severity flags are a third, independent layer: documented criterion-class conduct disqualifies regardless of any composite.

The Four Pillars are scored against a fixed set of questions, the trinity beneath each pillar. These questions are the grading criteria; the rule is strict: if the answer to any of them is “no,” that person is not worthy of your followership, no matter how impressive they seem on the surface.

Each pillar is scored 0–10 against its trinity; the four sum to a total out of 40 (Strong 32–40 · Moderate 24–31 · Weak 16–23 · Unfit 0–15). The Four Pillars in full, with the meaning behind each question →

The severity-flag layer, three tiers

Severity flags are the third axis, independent of the composite and the Four Pillars. Documented criterion-class conduct registers here whether or not the 14 measures happen to capture it at their weight. Flags come in three tiers, distinguished by how the conduct bites:

• Drag flags lower the composite, because the underlying conduct is weighable in degrees, financial enrichment, office-to-family payments. There is more or less of it, and the composite reflects it; no separate penalty.
• Capping flags do not change the number but foreclose an Author’s Verdict of “supported”, a documented criterion-class breach of this severity cannot be endorsed, however strong the rest of the record. This is the tier for conduct the standard treats as corrosive out of proportion to its measure weight: sustained enemy-making / incitement (Criterion 10), and institutional-norm / process subversion (Criterion 8) in its capping form (see the note below on when Criterion 8 caps and when it is only a drag). The number still computes and is shown; the flag overrides the endorsement, not the math.
• Terminal flags do not lower a number, they suspend it entirely. Conduct that kills or causes the death of persons without due process, true tyranny, genocide, the theft of an election or refusal to cede power, or (in historical scope) slaveholding carries a terminal classification, DISQUALIFIED, and the measures are not computed, because a composite is an average, and an average implies that good conduct can offset the catastrophic. It cannot. The side-constraint principle: certain acts are not quantities to be weighed but lines that, once crossed, end the analysis. The trains running on time is not a number that offsets the trains running to Auschwitz.

Criterion 10, Sustained Enemy-Making / Incitement. The framework holds, with the Tolerance Continuum, that casting fellow citizens as enemies who do not belong is the corrosive force a constitutional oath exists to resist. But on the 14 measures the discourse domain carries only a minority of the composite weight, so a record could floor every discourse measure and still clear the support bar on the strength of the rest. The capping flag closes that gap: a documented pattern of enemy-making or incitement (not a single heated line, and never mere policy heat or contested argument) forecloses support on its own. It is a capping flag, not terminal, it bites the endorsement, not the number, and it is applied symmetrically: the same evidentiary bar, both sides, conduct only.

Criterion 8, when it caps and when it is only a drag. Process subversion covers a wide range, from cynical calendar hardball to schemes that try to undo an election. The standard draws the capping line at a single place: a Criterion-8 flag caps (forecloses support) only when a legal-on-its-face power is used to nullify or overturn a completed or certified democratic outcome, signing an amicus brief to discard another state’s certified electors, a fake-elector scheme, a refusal to certify a result already decided. Procedural hardball that exploits a vacancy or controls a calendar, however cynical or pretextual, is scored as a heavy drag on integrity and institutional stewardship, not a cap, because the underlying power is genuinely the body’s, the bare duty to act on a nominee is constitutionally contested, and no outcome already decided is being undone. A senator who runs out the clock on a court seat has abused a process; a senator who signs a brief to throw out votes already cast has attacked the result itself. Both are graded, but only the second forecloses an endorsement. This is why the 2016 Supreme-Court blockade is weighed as a drag while the December 2020 certified-elector amicus is weighed as a cap.

Flags rest on conduct connected to the office. A severity flag, the axis that can foreclose an endorsement, attaches to conduct undertaken while holding or seeking the office, under the oath, or while using the power the flag concerns. Private-citizen conduct from before a person took office is documented in the record as context, and it may inform the discourse measures, but it is not the basis of a capping or terminal flag. The reason is principled and symmetric: the standard grades whether a person honored the public trust they were given, not who they were before they were given it. The clearest illustration is a member whose most violent pre-office social media is set aside for flag purposes, yet whose in-office record of dehumanizing rhetoric and enemy-framing clears the Criterion-10 bar on its own.

M03, the boundary-ladder template

The rubric stops being “trust my eye” and becomes “apply my test.” Each measure publishes the rule that separates one score level from its neighbor, so the rating is reproducible by a hostile reviewer. M03 (Persons of Equal Worth) ships as the working template. Distinguishing criterion: the function of the language on its face, observable, not subjective intent. Does it treat the target as a fellow person with bad ideas, or strip their personhood?

Boundary tests. 3→2: does the word place the target inside the set of persons (stupid, wrong) or outside it (sub-human, infesting)? Inside = 3–4, outside = 2. 2→0: dehumanization without action-direction = 2; dehumanization that invites or celebrates removal = 0–1.

Symmetry of standard, not symmetry of outcome

The same rule is applied to every officeholder. Symmetry is how the methodology detects bias, anchors at each level should be findable on both sides when the judgment is fair. It is not a quota. We do not place an act at a level it does not belong to manufacture balance. If the standard, honestly applied, lands harder on one side on a given measure, that asymmetry is the finding, not a flaw to correct. Forced balance is itself a thumb on the scale.

Partisan gamesmanship, identified and set aside

A fixed standard has to refuse the partisan narrative as firmly as it refuses the partisan defense. Many officeholders are the target of loud, organized accusations that are debunked, overstated, unadjudicated, or simply policy dressed up as misconduct. The methodology does not quietly ignore these, it names them. Where a record carries notable accusations the standard declined to count, the dossier lists them in a "Partisan gamesmanship, identified & set aside" section, each labeled by why it was excluded (debunked, overstated, dismissed allegation, policy-not-conduct, or an unproven legal conclusion). This is the same discipline, applied to every side: the score rests only on what is actually established, and the reader can see exactly what was refused. A standard that can be weaponized by whoever shouts loudest is not a standard; making the set-aside visible is how the scorecard stays the measure rather than the weapon.

Asymmetric recovery, hard to raise, easy to lose, never a one-way trap

Character behaves like trust: slow to build, fast to destroy. A score is hard to raise because raising it requires sustained demonstrated conduct across multiple cycles; a single good vote does not undo a pattern. It is easy to lose because a single severe breach reveals character no amount of good conduct conceals. But recovery is always possible, a score that could never recover would be a blacklist, not a measure. The standard never moves; the evidence does.

M07, silence is scoreable

Silence in the face of one’s own side’s misconduct is not neutral, it is an M07 (Duty to Call Out) failure. Compliance can itself be scored conduct. The officeholder who chooses not to speak when their party is breaching the standard is making a decision the rubric records, the same as the one who calls it out.

M15, Functional Capacity (dormant exception-flag)

M15 is not a continuous score. The default for every officeholder is “No functional-capacity events on record,” contributing nothing to the composite. It activates only on documented, verifiable functional events. Strict scoping: capacity of record only (never “seemed unsure”), age-neutral (an 82-year-old with a clean record scores fine; a younger officeholder with documented failures does not), style excluded (tangents and insults are rhetoric, scored elsewhere). This is the defense against ageism and armchair diagnosis: it fires only on hard evidence.

The Mechanical Inclusion Rule & scope

A scorecard that hand-picks its subjects can be accused of cherry-picking, so the ranked roster is bound by a stated rule: any entry meeting the rule is in; any that doesn’t goes in a labeled appendix. In scope: sitting Congress, sitting Governors, current + immediately-preceding Cabinet, modern Presidents/VPs (Truman forward), and recent departures (Romney, Manchin, Sinema). Anchor sets (not rankings): State-Power, Historical, and Judicial figures, reference points, completeness never claimed. Article III justices are scored under a separate SC-M01..SC-M06 sub-rubric. “Why X and not Y?” is answered by the rule, not by the author’s judgment.

How to dispute a score

Any reader can file a rebuttal citing one of three valid arguments: (1) the cited evidence is factually wrong, (2) the anchor neighborhood is misidentified, or (3) the shared features between conduct and anchor are misstated. Arguments outside these three, “the other side does worse,” “this is how politics works,” “everyone does this,” “they were not charged”, are not the methodology’s currency. Evidence is. The methodology is the legitimacy: we publish the standard, the sources, the weights, the anchors. A hostile reviewer can take the same library and produce competing scores. The credibility survives.

The methodology is the legitimacy.