// Methodology

19 metrics, 35 versions, all of them dated.

Every chart carries the version of the metric that produced it. This is what those versions mean — what each one measures, how it is defined, what changed at every bump and whether values that had already been published moved. It is generated from the source of each module on every build, so it cannot describe a definition the site is not using.

// Why a version on a chart matters

Retuning a metric silently rewrites every historical chart drawn with it. Change the diminishing factor in the risk model and every value the site has ever shown is a different number, with nothing on the page to say so. That is the failure this catalogue exists to make impossible, and it is why each entry below states not just what changed but whether the change reached backwards.

3 of the 19 metrics have restated a published value at least once — flags, resrisk, tech. A chart drawn before one of those bumps does not match the same chart drawn after it, which is a fact about the numbers rather than about the code, and belongs where a reader can find it.

Initial: The first version of this metric. There is no earlier definition to restate.
Additive: Output was added. Nothing already published changed, so every chart drawn under the previous version still reads the same.
Restated: At least one value that had already been published is different now. Charts drawn before this version do not match charts drawn after it.
Not stated: The module does not say whether values already published moved. We would rather show this gap than close it with a guess.

// Market metrics — 11

Metric	Version	Versions	Charts	Historical values	What it measures
align	align-v2	2	2	Never moved	Putting two sources on the same calendar day.
alt	alt-v1	1	1	Never moved	Alt season index — share of large alts outperforming BTC over a rolling window.
breadth	breadth-v2	2	4	Never moved	Market breadth and cross-asset correlation.
cycle	cycle-v5	5	115	Never moved	Cycle and ROI comparison.
dom	dom-v1	1	3	Never moved	Dominance — share of a defined basket, not of "the total crypto market cap".
implied	implied-v1	1	2	Never moved	Market-implied probabilities from Kalshi candles.
logreg	logreg-v2	2	44	Never moved	Logarithmic regression — fit log(price) against log(days since genesis).
risk	risk-v1	2	46	Never moved	Risk metric — reconstruction, not a clone.
riskstat	riskstat-v1	1	15	Never moved	Statistics about the risk metric.
tech	tech-v4	4	137	Restated	Technical indicators.
val	val-v1	1	44	Never moved	Valuation multiples and the asymmetric regression fan.

// Settlement and resolution — 8

Metric	Version	Versions	Charts	Historical values	What it measures
cat	cat-v1	1	3	Never moved	Market categories — the denominator for "dispute rate by category".
flags	flags-v2	2	3	Restated	Phase 4 — the flag taxonomy. This is the product, not the plumbing.
latency	latency-v1	1	—	Never moved	Settlement latency — how long a venue takes to pay out after the question is answered.
match	match-v3	3	—	Not stated	Phase 4 step 3 — candidate generation for cross-venue matching.
resdb	resdb-v2	2	3	Not stated	Phase 4 — the resolution database. Four tables, one per RESOLUTION_DB's design.
resrisk	resrisk-v2	2	3	Restated	Resolution risk as a metric module — the locked contract, plus the evidence.
uma	uma-v1	1	—	Never moved	UMA's DVM: who settles disputes, with what stake, which way, and when slashed.
verify	verify-v1	1	—	Never moved	Phase 4 step 3, stage two — LLM verification of candidate links.

// Gaps — 3

These versions changed something, and their modules do not say whether values already published moved with it. The catalogue is generated, so it can either report that or invent an answer. It reports it. Until the module says, treat a chart stamped with one of these as possibly not matching the same chart drawn before the bump.

match-v2Words and numbers separated. v1 tokenised on [a-z0-9']+, so "$4.07" became {4, 07} and "2026-07-07" became {2026, 07}. Rare digit fragments earn high IDF, so a question scored against an unrelated one purely on a shared "07" — "Will average gas prices be above $4.07?" matched "Will Argentina win on 2026-07-07?" at 0.71, above genuine same-event pairs. Numbers now carry no similarity weight; instead, two questions that both quote thresholds but share none are penalised, because "above $4.07" and "above $4.04" are different markets.
match-v3Dates stripped before numbers are read. v2 still let a date's parts become thresholds: "2026-07-05" gave 7 and 5, which then AGREED with an inflation question about 5.00% and cancelled the very penalty meant to separate them — "inflation in Brazil above 5.00%" scored 0.709 against "Will Brazil win on 2026-07-05?". Years are still read from the original text, because a year inside a date is exactly the deadline worth comparing.
resdb-v2settlement_time added — when a contract ACTUALLY settled, which on Kalshi is not expiry_time. See below.

// Machine-readable

The same data, generated in the same pass, with no login and no key: catalog.json and llms.txt. Both carry every version, every verdict above and the sentence from the source that produced it.