7f69fef14c08dd976a2fd8b13fd1e19444bfc8eb
455 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 builtbykev
|
7f69fef14c |
Add an on-demand endpoint for the lineup-context ingest
The first prod run wrote zero rows while the parser demonstrably works locally (144 rows, 10 games with lineups posted), so the zero was wiring rather than absence -- but diagnosing that required a full snapshot, which now takes about three minutes and 524s at the edge. Same reasoning as the statcast refresh endpoint: a job is proven by running it and reading the result, never by waiting for the slot it rides in. This makes the ingest verifiable in seconds, so 'zero rows' can be told apart from 'no lineups posted yet' immediately -- which is the exact confusion the defence ingest hit when a doubled path 404'd and read as 'Statcast has no fielding data'. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
08276c0880 |
Ingest lineup + baserunner context: the input RBI and runs always needed
RBI is power TIMES opportunity. The same swing drives in one run or three depending on who is on base, and a hitter batting with the bases empty cannot drive anyone in however hard he hits it. Every context-free model of RBI here has failed, and the failure kept being read as 'skill inputs don't work for RBI' when the truth was that we were modelling half the stat. Both halves are free from statsapi.mlb.com, which we already call for game logs, schedules and probable pitchers. No new provider, no key, no quota. RUNG 1, batting order: schedule?hydrate=lineups returns homePlayers and awayPlayers as ORDERED arrays of nine, and the order IS the batting order -- index 0 is the leadoff hitter. That single fact gives CATALYST its identity and supplies lineup-position context for every context-dependent stat. RUNG 2 turned out cheap, which the cheapest-first rule did not expect. It looked like it would need play-by-play reconstruction across a season; statsapi serves situational splits directly, so 'how often does this hitter bat with runners to drive in' is ONE call per player rather than one per game. Measured on a real hitter: 87 plate appearances with runners in scoring position producing 25 RBI, against 302 with the bases empty producing 17. That ratio is the opportunity half of the stat and it is the thing no amount of exit velocity can tell you. Both tables are dated in the primary key. statcast_aggregates was built upsert-in-place and that silently made every backtest leak the games it was predicting; a lineup is worse still, because it is a PRE-GAME fact that changes by the hour, so an in-place table would overwrite what we knew at grade time with what turned out to be true. Absent stays absent throughout: no lineup posted is an empty slate rather than a guessed order, a short lineup records fewer slots rather than padding to nine, and a hitter with no splits is null rather than a zero RISP share -- which would assert he never bats with runners on, a strong claim and usually a false one. Wired into the snapshot best-effort, so a context failure can never break the pipeline it rides in. The three pre-registered theories are now marked input-ready rather than input-blocked: DRIVER's power x runners-on and power x lineup-position, and CATALYST's speed x on-base x power-behind. They are sample-blocked from here, and the proofs run under native cumulative correction as sample accumulates -- ingesting is not proving. Counter and frozen clusters byte-identical. 4,250 tests green (338 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
52a3142c5f |
Split infield defence out, and pre-register the DRIVER/CATALYST/SINKER batch
Every slot in this batch is far below the gate -- DRIVER x hits 39, CATALYST under 22, SINKER not yet gradeable at all -- so per the order's own sample rule these are CANDIDATE-pending-accumulation, not tested-and-failed. Testing them now would produce noise and burn cumulative-correction budget on it. What IS deliverable is the input SINKER's theory needs, and it turned out to be free. The OAA feed already carries each fielder's position, so infield-only defence is derivable from data ingested yesterday: 1B/2B/3B/SS summed separately from the outfield. Team-total OAA is the wrong unit for a ground-ball pitcher -- he lives on the infield converting grounders and his outfielders are close to irrelevant to him, so averaging them in dilutes exactly the signal. On a real team the split shows a +15 infield inside a +2 team total, which is the dilution made visible. Under three measured fielders in a unit is absent rather than zero, same rule as everywhere else. The three theories are now PRE-REGISTERED in the registry with their mechanism and the skill each would validate, marked CANDIDATE. That is the point of writing them down before the sample exists: the claim is on the record with a date and cannot be quietly reshaped into whatever the numbers turn out to support once they arrive. Two of them are input-blocked rather than sample-blocked, and the distinction matters because waiting will not fix them. DRIVER's RBI theory needs baserunner state and CATALYST's runs theory needs both baserunner state and batting order; we ingest neither, and player_role_profiles is empty. So RBI does not unblock on DRIVER -- it unblocks on ingesting lineup context, which is a sourcing question, not an accumulation one. SINKER is the only one of the three whose inputs are now ready. Premise note: no registry re-adjudication demoted anything last session. The proven set was empty, zero features were demoted, nothing was recalibrated, and nothing is published -- node scripts/proven-status.js confirms it in one command. Counter and frozen clusters byte-identical. 4,238 tests green (337 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
ff037e40c2 |
Re-adjudicate: nothing to demote, and close the hole that would have mattered
There is nothing to re-adjudicate. The proven set is empty and always has
been -- verified three ways: proven-status reports EMPTY, validatedSkills()
returns {} for every archetype, and zero conditioning entries have ever
reached PROVEN. The one PROVEN feature is recent_frequency_prior, which is the
incumbent counter itself, proven by the S78 ablation as ~100% of the
champion's resolution. It is the baseline every challenger is measured
against, not a conditioning interaction, and demoting it would leave the model
with nothing to grade from.
A correction to the premise: the cumulative gate did NOT catch a false
positive last session. It caught nothing, because there was nothing in the
proven set to catch. What it did was tighten alpha from 0.0026 to 0.0013
within one session, which demonstrated the mechanism working rather than a
demotion. So steps 3 and 4 -- demote, recalibrate -- are vacuous here, and
readjudicateAll says so plainly rather than glossing a no-op.
But the worry behind the order was well founded, and the audit found the real
exposure: promote() did not require the cumulative denominator. It checked n,
lift and CI, and nothing stopped a future session from testing eight
hypotheses, correcting by eight, and promoting on a p-value that would not
survive the programme's real denominator. That is precisely the hole that
makes a retroactive re-adjudication pass necessary later, so it is closed at
promotion time instead. isSufficient now refuses evidence carrying no
correction, evidence corrected against fewer tests than the cumulative count,
and any p-value that does not clear 0.05 over its own test count. The same
rule guards a PROVEN conditioning entry.
The second audit found two of four analysis scripts still correcting
per-session; pitcher-prove-k and tb-solo-and-interactions now use the
cumulative ledger, so the correction is native on every path.
reAblation.js is the standing second line: pure and injectable, so the
decision rule cannot drift from the gate's, and every verdict records both
p-values and both test counts so a demotion is re-derivable by anyone. A
feature promoted at alpha 0.05/20 can demote on the same p-value once the bar
is 0.05/60 -- correct, because the bar rose only after the programme had more
chances to get lucky. No fresh measurement is PENDING_RETEST and never a
demotion: absence of a re-test is not evidence, and demoting on it would
punish whichever stat happens to be off-season.
Net effect on the proven set is zero. No demotions, no recalibrations, and no
public ledger event -- announcing "recalibrated after re-adjudication" when
nothing changed would itself be a false signal of rigour.
4,238 tests green (337 suites); web build exit 0; counter byte-identical.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9
|
||
|
builtbykev
|
ece2b9f5f9 |
Ingest defence, and make Bonferroni cumulative across the programme
Two things shipped that stand regardless of sample. DEFENCE. Statcast Outs Above Average is free on the host we already pull six feeds from, so there was nothing to decide. 514 fielders, aggregated to team level -- the unit a batter's prop actually needs, the defence behind the pitcher he faces -- and persisted as 31 team rows. Verified in production. Cubs +56 best, Mariners -29 worst. Unknown is not zero, and it bites unusually hard here: an OAA of 0 is a REAL reading meaning exactly average, so coercing absence to 0 would assert that every unmeasured fielder is league-average, which is the commonest defensive profile there is. team_defense also carries as_of_date in its primary key from the first row -- statcast_aggregates was built upsert-in-place and that silently made every backtest leak the games it predicted, so point-in-time is available here before it is needed rather than after a wrong answer. A bug worth recording as a class: BASE already ends in /leaderboard, so the new feed built a doubled path and 404'd. Because a failing feed degrades to an empty index by design -- correct, so one broken source cannot fail the whole pull -- it surfaced as "fielding_oaa: 0 rows", which reads exactly like "Statcast has no fielding data". Graceful degradation makes a wiring bug look like an honest absence. CUMULATIVE CORRECTION. Bonferroni had been applied per session throughout: a run testing eight features corrected by eight. Across a programme's lifetime that is wrong in the dangerous direction, because every order gets a fresh generous alpha and the false-positive rate compounds quietly. Correcting by 8 when sixty have been tried is how a noise result eventually gets recorded as PROVEN with a p-value to point at. The denominator is now distinct hypotheses ever tested, persisted, and it moved 19 -> 38 within this session alone, alpha 0.0026 -> 0.0013. Re-tests deliberately do not inflate it: re-asking the same question on more data is not a new shot on goal, and counting it would punish the discipline of waiting for sample. THE MEASUREMENT. The differential the theory predicted is present: defence correlates with the counter's residual at +0.130 for GHOST, the contact and speed archetype, and -0.018 for BOMBER, the power archetype. A GHOST's hits depend on whether anyone can range to the ball; a BOMBER's barrels clear the defence entirely. So a flat BOMBER result is the theory working rather than the test failing. It is not a result. GHOST is n=104 against a 500 bar, with p=0.188 against a corrected alpha of 0.0013 -- three orders of magnitude short. Both are recorded as CANDIDATE with their measured lift, tagged contact-skill, so the re-run at full sample compares against a recorded baseline. Nothing proved, so nothing was recalibrated and nothing shipped. 4,228 tests green (336 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
25e36c0257 |
Fix the doubled /leaderboard path — the fielding feed 404'd silently
BASE already ends in /leaderboard, so the new feed built .../leaderboard/leaderboard/outs_above_average and 404'd. Because a failing feed degrades to an EMPTY index by design -- correct behaviour, so one broken source cannot fail the whole mechanism pull -- it surfaced as 'fielding_oaa: 0 rows' rather than as an error, which reads exactly like 'Statcast has no fielding data'. Worth noting as a class: graceful degradation makes a wiring bug look like an honest absence. Verified: 514 fielders, 31 teams. Cubs +56 OAA, Mariners -29. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
010a876b3c |
Ingest free Statcast fielding (OAA) as team defence, dated from day one
Defence was the one conditioning category with no derivable proxy: nothing we ingest measures fielding, and a team's pitchers' hits-allowed conflates pitching with defence and would validate the wrong skill. Statcast publishes Outs Above Average free on the same host as the six feeds already pulled -- verified live at 513 fielders -- so there was nothing to decide. Added as a seventh feed, indexed per fielder and aggregated to team level, which is the unit a batter's prop actually needs: the defence behind the pitcher he faces. Summed OAA is the team's outs converted above average; the mean rides along because a team with more measured fielders would otherwise look better merely for being measured more, and a team with under three measured fielders is absent rather than thin. Unknown is not zero, and it bites unusually hard here: an OAA of 0 is a REAL reading meaning exactly average, so coercing absence to 0 would assert that every unmeasured fielder is league-average -- the most common defensive profile there is, and a fabricated fact rather than a neutral default. team_defense carries as_of_date in its primary key from the first row. statcast_aggregates was built upsert-in-place with a single as_of date, which silently made every backtest leak the games it was predicting and cost a session to find; this makes point-in-time available before it is needed instead of after a wrong answer. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
ac1361486e |
Build the conditioning registry, and a probe so "proven" stops drifting
The order opens with "two proven clusters live". They are not proven -- the
proven set is empty -- and this is the fourth consecutive order to start from
a stronger claim than the measurements support. Correcting that in prose four
times has not worked, so this session adds scripts/proven-status.js, which
recomputes the answer from the ledger: hits LOSES (-0.096, CI excluding zero),
total_bases INCONCLUSIVE (+0.004), strikeouts INCONCLUSIVE (+0.259 at n=57).
It deliberately reports sample readiness separately from recorded verdicts, so
"n>=500" can never again be read as "passed".
A counting error worth recording. The first read of the top-volume archetype
said BOMBER x hits was 641 rows -- gate-ready. It is 287. model_snapshots
holds one row per prop PER SNAPSHOT CYCLE, so joining it to ledger_entries
counts each ledger row once per cycle it appeared in. Deduping on the ledger
row id gives the true figure, and my own status script had the same bug until
it was fixed. That is the difference between running the gate and being short
by 213.
So no archetype x stat combination reaches the gate. BOMBER x hits at 287 is
the closest; pitcher archetypes are untestable at 58 settled strikeout rows
across all of them, so the pitcher half of this order could not be run.
The registry is built: recordConditioning keys archetype x underlying-skill x
interaction x status with measured lift, and the skill tag is MANDATORY and
enforced -- untagged entries are refused, and PROVEN without sufficient
evidence is refused. validatedSkills() returns the coherent profile as it
stands, which is {} for every archetype, by design.
BOMBER x hits conditioning was tested across the order's categories and every
result is underpowered: arsenal (barrel x breaking share) incremental +0.043,
batted-ball (launch x pitcher GB) +0.001, contact quality -0.020 and -0.015,
K x K -0.063. Within BOMBER the counter still leads on hits, 0.218 to 0.160,
consistent with the closed pooled negative.
One bug fixed mid-run: fromStatcastRow maps percentage and raw fields only and
does not carry pitch_mix, so the arsenal category first reported n=0 for every
row -- it was measuring nothing rather than failing. Without catching it,
"arsenal doesn't matter" would have been recorded from a column that was never
populated.
On defense: I looked for a derivable proxy before calling it unsourceable, and
there isn't one. We ingest no fielding data at all, and opposing pitchers'
hits-allowed conflates pitching with defense, so it would validate the wrong
skill. It needs Savant's fielding endpoint -- free, same host as the five
feeds already ingested -- and it is not sourced here, because sourcing it to
test at n=282 would answer nothing.
Nothing proved, so nothing was recalibrated and nothing shipped.
4,221 tests green (335 suites); web build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9
|
||
|
builtbykev
|
9538e11198 |
Derive the lineup K-rate free, and fingerprint the cap fix
Two premise corrections first. Pitcher stuff features have NOT proven solo through the gate -- every one was refused on sample (n=57 against 500). Four exceed the effect-size bar (arm angle -0.250, whiff +0.213, k rate +0.206, chase +0.195), which is why they are worth pursuing, but clearing one of three thresholds is not passing. And the carrier was not blocked only on the lineup input: that input was built and measured last session at 94.7% coverage. What blocks it is n, and n was being throttled by the grading cap. RUNG 1 IS DERIVED AND COSTS NOTHING. Opposing-team K-rate comes from joining the opposing roster to the batter k_pct values already in statcast_aggregates -- no new feed. The improvement this session is that it is PA-WEIGHTED: an unweighted roster mean counts a 12-PA callup the same as an everyday starter, which is not the lineup a pitcher faces. That change alone reversed the term's sign. Unweighted, the lineup term HURT the model (0.1738 -> 0.1285). PA-weighted, it HELPS (0.1738 -> 0.1953). Same hypothesis, same data -- the derivation was the problem, not the signal, which is the entire argument for deriving the best honest version before sourcing anything. Head-to-head is now +0.2592 with a CI of [-0.0167, +0.5645], very nearly excluding zero, at n=57. Within archetype, the two strata come out with OPPOSITE signs -- FLAME incremental -0.152, non-FLAME +0.145 -- and the pooled value (+0.077) sits between them, which is the shape a conditional effect makes and is invisible when pooled. That is what stratifying was for. But n is 20 and 24, the standard error on a correlation there is about 0.22, and the direction contradicts the theory that predicted a stronger effect for finesse arms. It is recorded as a structure to re-test, not as a finding. Rungs 2 and 3 are NOT triggered. A rung fails only once it has been fairly tested, and Rung 1 is n-blocked rather than failed. Sourcing confirmed lineups now would be paying for precision on top of a proxy we have not yet measured. THE RESULT THAT DECIDES THE TIMELINE: yesterday's cap raise is fingerprinted in production at 907 grades per snapshot, up from 334, with strikeouts going 6 to 17. That puts n>=500 for pitcher Ks about a week out instead of three months. Operational note: the manual internal snapshot endpoint now 524s at the Cloudflare edge because grading the full board exceeds 100s -- the run still completes server-side (this very snapshot was written by a 524'd request) and the cron is in-process, so a 524 there is not a failure. Nothing proven, nothing calibrated, nothing shipped. The counter remains anti-predictive on strikeouts at -0.064 and the skill model leads it by 0.26. 4,221 tests green (335 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
843c8c6d4b |
Build the pitcher engine, and find the cap was eating the whole board
Strikeouts are NOT proven -- n=57 against a bar of 500. But the finding that matters is not a correlation. THE CAP. Measured on the live slate via the refusal diagnostic: 1,244 unique gradeable props exist, the 500 cap graded about 334, and because dedupeProps takes first-row-wins in FEED ORDER, what survives is decided by feed position rather than value. Pitchers are 2.6% of a batter-dominated feed, so we were grading SIX strikeout props a slate against 32 available -- putting n>=500 three months away for every pitcher stat. Pitcher props were never being refused (graded 5, refused 0, suppressed 0); it was truncation. Raised 500 -> 1500 on measured cost: 721ms per prop at concurrency 5 is about 179 seconds for the full board, against a cron that runs five times a day and a fire-and-forget caller that never holds an HTTP response. statsapi is free and unlimited. Concurrency stays at 5 -- one variable at a time. This unblocks every n-blocked stat in the programme, not just pitchers. THE ENGINE. pitcherEngine.js is its own engine, not the batter engine pointed at pitchers: the batter model asks whether contact becomes a hit and reads contact quality, the pitcher model asks whether the plate appearance ends without contact at all and reads stuff. Archetypes are FLAME (whiff-led), SCALPEL (chase-led), SINKER (pitches to contact) and DEFAULT, and a test asserts the weight keys are not the batter engine's. The projection is K% by log5 against THIS lineup, times batters faced, through a binomial. An unclassifiable arm gets the balanced map, never a guessed archetype. THE MEASUREMENT, at n=57 and contaminated. Four solo features clear the 0.15 effect bar and fail only on sample: arm angle at -0.250 -- the largest correlation measured anywhere in this programme -- then whiff +0.213, k rate +0.206, chase +0.195. The batter cluster's best was 0.135. Head to head, pitch-v1 resolves 0.1285 against the counter's -0.0639, delta +0.192 with a CI spanning zero. That negative is the interesting number. The counter is ANTI-PREDICTIVE on strikeouts: counting a pitcher's recent Ks is worse than useless, because his recent totals track which lineups he drew and how long he was left in rather than his skill. It is the one stat where the incumbent has no defensible edge. A bug caught on the way. resolveTeam wants an abbreviation and the game log supplies full team names, so the roster join silently resolved nothing and the first run reported 0% lineup coverage -- the theorized stuff x lineup carrier was never being tested, not failing. Fixed; coverage is now 94.7%. The carrier still shows no incremental signal over whiff alone, and adding the lineup term lowered head-to-head resolution, which is recorded rather than dropped. Calibration was not reached: nothing passed the first bar. The batter model and the counter are byte-identical, verified by diff. 4,221 tests green (335 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
c0621e7aa2 |
Measure the batter cluster: the proven set is empty, and hits is closed
PREMISE CORRECTION FIRST, because it defines the bar. total_bases has not
passed BAR 1. Its head-to-head is inconclusive at parity -- delta +0.004 to
+0.007 with a CI spanning zero -- and it is contaminated, and no feature of
its passed the gate. It was described last session as the first challenger
that did not LOSE, which is not the same as proven. If it is installed as the
frozen proven reference and every other stat is held to "the identical bar
total_bases cleared", the bar becomes "be inconclusive at parity" and the
whole cluster passes on a null result. The proven set is EMPTY.
HITS IS NOW A FINAL ANSWER. At n=803 it clears the gate's sample requirement,
so its features were properly TESTED rather than refused: every one fails on
effect size (max marginal |r| 0.053 against a 0.15 bar), every interaction's
incremental contribution collapses to about zero, and the model loses
head-to-head by 0.096 with a CI excluding zero. That is a well-powered
negative and hits should be closed rather than retried.
The rest are n-blocked: total_bases 383, rbi 391, home_runs 228, runs 188,
against a bar of 500. Two leads are worth carrying. home_runs barrel rate has
a marginal r of -0.135, and the sign matters -- higher barrel rate goes with
the counter OVER-predicting, which would be a correction rather than a new
predictor. And runs batterK x pitcherK has the largest incremental in the
cluster at +0.132, with a clean mechanism: strikeouts destroy plate
appearances, and a PA that never happens cannot score.
RBI deserves a caveat rather than a verdict. It is power times OPPORTUNITY,
and we ingest no baserunner state at all, so half its mechanism is missing. A
weak RBI result is evidence that we are modelling half the stat.
total_bases was held frozen: git diff on skillProjection against the prior
commit is empty. The counter is untouched.
Also fixed and verified in production: the point-in-time retention shipped
after yesterday's refresh had already run, so statcast_history was empty, and
its first run then failed on a hand-enumerated schema that had already drifted
from its source ("could not find the 'swing_pct' column"). The refresh itself
still succeeded and wrote all 1,387 aggregate rows, which confirmed the
best-effort guard in prod. The table now mirrors the source via LIKE and the
writer passes rows through whole. Verified live: 1,387 rows retained at as_of
2026-08-03. A usable point-in-time window starts 2026-08-04.
Stage B has nothing to calibrate. Everything now waits on a point-in-time
window and on sample -- both waiting problems, not building problems.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9
|
||
|
builtbykev
|
c2d6e8ee7d |
Mirror statcast_history on its source so retention cannot drift
The first production run of the point-in-time retention failed with "Could not find the 'swing_pct' column of 'statcast_history'" -- the hand-enumerated column list had already drifted from the table it was copying. The refresh itself still succeeded and wrote all 1,387 aggregate rows, which verified the best-effort guard in prod: a retention failure does not fail the refresh. The table is now created with LIKE statcast_aggregates, and the writer passes the row through whole instead of hand-stripping columns, so there is no drift surface left. Recreating was safe -- nothing had been retained. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
4aab18096f |
Prove both on total bases -- and find that my own fix destroyed the backtest
Nothing passed. Nothing promoted. Counter byte-identical. THE BLOCKER, which is the real finding. statcast_aggregates is upserted in place and holds exactly one as-of date. Yesterday's skill backtest was honest only by accident: the nightly refresh was unreachable code, so the profiles sat frozen at 2026-07-21 -- before the settled window. Repairing that cron was right for production and it refreshed them to today, destroying every prior version. Scoring a 2026-07-25 game now uses a season aggregate that contains that game. Point-in-time validation is structurally impossible from that table, so every number in this run is contaminated and directional, and none of it is a gate verdict. Fixed forward: statcast_history retains a dated snapshot on every refresh, so point-in-time becomes "as_of_date < game_date, most recent". Retention is best-effort and cannot fail the refresh; both properties are unit-tested. It has one day of data, which is not yet a window. SOLO BASELINE, n=383, Bonferroni across 12 tests (alpha 0.00417): nothing passes. hard_hit_pct is closest at marginal r 0.135 with p 0.0080, failing both the 0.15 effect bar and the corrected alpha. And it drifted DOWN from 0.153 at n=295 -- an estimate regressing as noise averages out, not an effect firming up. I called that number encouraging yesterday; on 88 more rows it is fading, and it should not keep being quoted at its best value. INTERACTIONS, each scored by partial correlation against the counter residual controlling for both of its own components: none pass. Only barrel x power archetype has an incremental exceeding its parts (-0.101 against 0.019) at n=260 -- the shape Discipline 2 predicts, but a lead, not a finding. A methodological catch worth keeping. The archetype conditioner was first built as barrel_pct over league barrel -- a monotone transform of one of its own components -- so the "interaction" was barrel squared, measuring nonlinearity in barrel rate rather than any archetype effect, and it produced this run's only positive result. A Gauss-Jordan pivot test does not catch that, because the two columns differ by a scale factor. Fixed with a scale-free collinearity check plus real archetype labels joined from model_snapshots. Without it this document would have reported a fabricated interaction as the session's finding. COMBINED vs COUNTER on total bases: 0.2718 against 0.2647, delta +0.0071, CI [-0.065, +0.079] -- inconclusive, and the first time a challenger has not lost. The same engine on hits was -0.116 with a CI excluding zero. That contrast is the whole argument for total bases, and it is what the physics said: contact quality governs extra bases, not whether a grounder finds a hole. Also built: the compound TB projection. skillProjection no longer refuses total bases -- a deterministic bases-per-hit multiplier had made P(TB>=2) exactly P(hits>=1), a relabelled hits curve. It is now a convolution over per-PA base outcomes with hit-type shares shifted by skill. Non-degeneracy is locked by test. 4,204 tests green (334 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
c7cc8f5e52 |
Build the gate, run it, and find we were proving things on the wrong stat
PREMISE CORRECTION FIRST. statModel.js and correlateValidator.js do not exist in this repository. The validation spec's only prior form is src/services/python/blueprints/unconventional.py -- a Flask blueprint in the Python service that is offline in production, scoring NBA factors against a warehouse that was never populated -- and tests/unit/supplementSystems.test.js requires only fs and path while defining its own validateFactor inline at line 368. Those tests assert a re-implementation of the thresholds, not an implementation, which is exactly why they passed for months while nothing was connected. The diagnosis behind the order is right -- every challenger was measured without a gate -- but the cause is that there was no gate on the Node side to import. So it is built, to the exact spec. correlateValidator: n>=500, |r|>=0.15, p<0.05, Bonferroni across the sweep. The p-value is exact rather than approximated (t-transform through a regularized incomplete beta) and is verified in the suite against known values, because scipy is not available here. Pairs with an unknown side are dropped, never zero-filled -- a zero-fill inside a correlation does not add noise, it invents a point at the origin. THE RUN, hits, n=570, Bonferroni-8: every skill feature fails, and not narrowly. The strongest marginal correlation against the counter's residual is 0.062 against a 0.15 bar. That is an effect-size failure at a sample that would have found a real effect comfortably -- a clean, well-powered negative. The head-to-head agrees: value engine 0.0499 against the counter's 0.166, delta -0.116 with CI [-0.189, -0.043]. Not promoted. THE RUN, total bases, n=295: cannot be tested, and that is the finding. hard_hit_pct shows a marginal r of 0.153 -- above the threshold -- and exit velo 0.124, refused solely because n is 205 short of 500. It is the most encouraging number this work has produced, and it is what the physics predicts: contact quality governs extra bases, not whether a grounder finds a hole. We have been testing skill inputs on the one stat where they should not matter much. Two things the run forced. Feature verdicts are now PER STAT, because marking these DEAD sport-wide on hits evidence would have killed, for total bases, the features that look most alive there -- per-sport doctrine one level deeper. And the gate now reports r and p even when underpowered, because "not enough data yet" and "nothing here" demand opposite decisions and a bare refusal was hiding the best signal on the board. Next: build the compound TB projection (skillProjection still refuses total bases by design, since a deterministic bases-per-hit made P(TB>=2) identical to P(hits>=1)), accrue to n>=500, re-run this gate. Leave hits alone. 4,200 tests green (334 suites); web build exit 0; counter byte-identical. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
258d8a6655 |
The skill engine: built, gated by construction, and Stage A honestly lost
Built src/services/model/ -- the forward, archetype-selected, skill-based projection, as a challenger. The champion is untouched. featureRegistry makes "earn its place or it's out" structural rather than aspirational: CANDIDATE / PROVEN / DEAD per feature per sport, liveFeatures() returns PROVEN only, promotion requires n>=200 with positive lift and a CI excluding zero, and there is deliberately no override argument. It ships with exactly ONE proven feature -- the incumbent counter, because it is the only one with a measurement. A test asserts that with only PROVEN features allowed the projection returns null, so an unproven model cannot reach a user by accident. The three champion adjustment layers are registered DEAD with their reasons so they cannot be silently rebuilt. skillProjection is a PA outcome tree: K and BB combined by log5 odds-ratio against league (both identities unit-tested), then archetype-weighted contact quality against contact allowed, then Binomial(PA, p_hit) mixed over a PA distribution. Archetype is a FEATURE SELECTOR, not a nudge -- BOMBER reads barrels at 0.50 and ground-ball speed at 0.00, GHOST inverts it -- and a test locks that the same hitter read two ways moves more than 0.15. STAGE A: IT LOSES. Out-of-sample on 570 settled hits props with 91.9% opposing-pitcher coverage, resolution 0.0499 against the champion's 0.166, delta -0.116 with CI [-0.189, -0.043]. It is not selective either: its eight most confident picks hit 50%, a lift of -0.065. Not promoted. The gate did its job on its first real test, which is the point of having built it that way. Two false starts, both recorded because they nearly produced a wrong verdict: statcast_aggregates stores PERCENTAGES, so raw rows made bip = 1-29.6-17.1 and refused 568 of 576 -- the honest-absent guards made a units bug loud instead of silent, and the conversion now lives at one chokepoint. And the first run resolved an opposing pitcher for 1 of 570 rows, because ledger team/opponent are NULL, so it would have reported "skill-v1 loses" while measuring a batter-only model with no matchup in it at all. The verdict above is from the corrected run. The loss is real but partial: park was passed as 1.0, handedness and opportunity_drift never fired, PA is season-PA over a constant, and the skill profiles carry no recency at all while the champion has a last-5 term. Also fixed: the Statcast nightly refresh was unreachable code. It sat inside tick() below "if (!HOURS_UTC.includes(h)) return" while testing h === 11, so it had never run once; the aggregates were 13 days stale and both of its alerts were in the same dead branch. It now runs on its own tick, and the test that passed happily throughout -- it only checked the string existed -- is replaced by one that asserts it is not behind the guard. 4,182 tests green (333 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
c551bf0340 |
Reality assessment: the forward model exists, wired to the wrong side of the pipe
READ-ONLY. src/ and web/ untouched. Inventoried every forward-model component against the real objective -- a forward matchup projection, not an edge number. The finding is that all of it already exists and is already loaded in production, and 100% of it sits DOWNSTREAM of the grade in challenger columns nothing serves. The served p_win reads three features and a game log; it has never seen a pitcher. Inputs are HAVE, not missing: statcast_aggregates carries 1,354 rows (750 pitchers, 604 batters) with exit velo, launch angle, barrel, hard-hit, whiff, chase, pitch mix, GB/FB, arm angle, and handedness complete on every row. Real gaps are team defense and catcher/umpire. So Stage A is a plumbing-and- modelling job, not a data-acquisition job. Found along the way: the Statcast nightly refresh is unreachable code. tick() returns for any hour not in HOURS_UTC (14,19,22,1,3) and the refresh block then tests h === 11, which that guard can never admit. The mechanism data has been frozen at its 2026-07-21 backfill for 13 days, and the block's own failure alert sits in the same dead branch -- the identical silently-guarded- out shape as the settlement outage. Design shows the counter: every factor label the SIGNAL BREAKDOWN renders is a restatement of recent frequency (l5_hot_vs_line, l20_over_line, back_to_back, home_game) plus several structurally-NBA labels (referees, coach pace, starters out) inside a baseball product. Not one names a pitcher, pitch type, handedness or park. The card's forward-read slots already exist and go unfilled -- the surface needs feeding, not redesigning. On what changes: the prior measurements were outcome-accuracy, not edge, so the metric was right and the question was narrow. proj-v1.1 and hits-v1 stay correctly refuted as DISTRIBUTION swaps on thin inputs -- neither tested a matchup-fed projection. arch-v1 is a market-relative nudge by construction and is the one component genuinely measured on the wrong axis. AT CEILING is provisional: measured only against features the champion already reads. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
d8bf7765db |
Decompose the champion: its whole edge is a hit-rate counter
READ-ONLY. src/ and web/ untouched; 4,159 tests still green. WHAT THE CHAMPION IS. probabilityEstimator is five lines of arithmetic: the empirical frequency of (stat > THIS line) over the game log, blended 0.6/0.4 with the last-5 frequency, then +/-0.03 opponent, +/-0.015 home/away, a cv>0.40 pull toward 0.50, and a clamp to [0.10, 0.95]. It reads three features. featureCache retains a dozen more that p_win never touches. THE ABLATION IS EXACT, NOT A REFIT. Every adjustment is closed-form from stored features and the consistency step is linear, so each layer subtracts algebraically out of the stored p_win -- no re-estimation, no re-fetch, no lookahead possible. Per stat, paired bootstrap: removing ALL THREE adjustments changes resolution by NOTHING on every stat hits -0.0059 total_bases -0.0015 rbi +0.0106 runs +0.0130 walks +0.0008 and rbi's home/away is mildly HARMFUL (+0.0053, CI excludes zero). So ~100% of the champion's resolution is base+recency: how often this player has cleared this number lately. Everything else is decoration. A CORRECTION. Pooled, the champion resolves 0.46; per stat it is 0.196 (hits) to 0.499 (rbi). Pooling stats with different base rates inflates correlation, so 0.46 should not be quoted as the champion's resolution. Last session's paired differences remain valid; only the absolute level was inflated. THE BIGGEST LOSS IS NOT A MISSING FEATURE -- IT IS THE CLAMP. 358 of 1,741 settled rows (20.6%) sit on the boundary, so the model emits a constant there and cannot rank a fifth of the book at all. And that constant hides two opposite failures: 0.900 covers home_runs-under truly winning 99.5% (9.5pts under-confident) next to hits-under truly winning 51.9% (38.1pts over- confident). PROB_CEIL=0.95 makes the 99.5% case inexpressible. Global over-prediction is +3.5pts, +7.6 on total_bases. None of this needs new data. ONE REAL MISSING-WEIGHTING LEAD: opportunity_drift, residual corr +0.156 on hits and +0.145 on total_bases -- it REPEATS across independent stats, unlike the weather hits on TB which sit inside the expected false-positive count (70 tests at alpha .05 expects 3-4). And we already compute it: arch-v1's opportunity axis uses it and extracts nothing (delta +0.0001). Wrong implementation, not a missing feature -- opportunity must scale the rate, not nudge the probability. ARCHETYPE IS UNMEASURABLE, NOT REFUTED. Only 2 of 41 labels (BOMBER, GHOST) reach n>=40 settled rows and every mean residual straddles zero. That is "we have not measured it", and it does not license acting in either direction. Why every challenger has failed is now legible: the ladder and hits-v1 REPLACE the frequency question with a fitted distribution; the environment axis adds inputs the champion ignores. Asking the frequency question at the traded line is the thing that works. Flagged, not fixed: model_snapshots.outcome is NULL on all 22,032 rows -- the retention table built for exactly this replay was never settled, so labels had to be joined from ledger_entries. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
f897c7ec06 |
hits-v1 fingerprint PASSED: 72/72 written in prod, 45 outside the band modelled anyway
The prod-write fingerprint that was blocked by the odds outage has landed on the first snapshot after deploy. hits-v1 records exactly as the live-board verification predicted, and the takeable axis behaves as specified -- scope is book identity, never price shape. The verdict is unchanged: hits-v1 is REFUTED and stays unpromoted. This confirms only that it is recording, so the forward accrual can judge the backtest. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
3ba3dd28f3 |
Scoreboard every challenger; diagnose the 429 as odds-api, not PropLine
PROMOTE-THE-EARNED. Nothing was promoted, because nothing earned it -- not because the bar was held high. Measured on the same bar that refuted hits-v1: own rows only, direction-aligned, paired bootstrap, promote only on a CI excluding zero. arch-v1 n=1741 delta 0.0000 CI[-0.0050,+0.0054] inconclusive contact-v1 n=1055 delta +0.0008 CI[-0.0052,+0.0069] inconclusive proj-v1.1 n=1664 delta -0.0301 CI[-0.0543,-0.0060] reliably WORSE matchup/tb-v1/hits-v1 n=0 genuinely pending (rows dated 08-02+) arch-v1 is the interesting one: it MOVED 76% of rows by 2.5 points on average and resolution is identical to the champion to four decimals, on the moved rows too. That is active movement carrying no information -- a finding, not a pending verdict. These are true prospective holdouts: arch-v1 and contact-v1 wrote p_win at grade time into their own columns before the game. Nothing recomputed. THE 429, read-only. The premise was that we re-pull the full picture every slot and blow the quota. Measured: PropLine is at 5 calls of 3,000/day -- 0.17%. One snapshot is ONE PropLine call per sport, all markets comma-joined. There is no request-pattern problem, so a change-based pull cannot fix it and no tier upgrade is needed. The 429 is odds-api: 478/500 MONTHLY, blocked at 95%. oddsService falls through silently when PropLine returns empty, and the backup's quota gate throws the error -- so an empty slate is indistinguishable from an outage and the message names the wrong provider. Flagged for its own order. Could NOT verify PropLine movement endpoints: docs are auth-gated and the keys are production-only. Not asserted either way. The movement-as-data argument stands on its own merits and should be justified that way, not as a quota fix it isn't. Book-breadth invariant written down: we never discard books. All are kept and shown (DISPLAY_BOOKS = MODEL + REFERENCE + DFS); DFS pick'em is excluded from PRICING only, because a fixed-payout shaded number is not a market price. Verified this is already what bookRoles.js does. Champion byte-identical; every challenger stays wired. 4,159 tests green (332 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
b06a84af80 |
Settlement has been dead since 2026-08-01: a 500-id filter overflowed the URL
The self-learning loop stopped two days ago and reported success the whole
time. 1,444 ledger rows from 2026-08-01 sit unsettled with settle_attempts=0
-- never even attempted -- and every accruing challenger has been starved of
settled sample as a result.
ROOT CAUSE. settleLedger fetched open ids, then REFETCHED the full rows with
.in('id', ids). PostgREST puts filters in the URL, so 500 UUIDs became an
18,499-character request that the fetch layer rejects with "TypeError: fetch
failed". The result was destructured as `const { data: rows } = ...` with NO
error binding, so rows came back null, the loop body never executed, and the
function returned {settled:0, voided:0, unrecoverable:0, pending:0} --
byte-identical to a clean "nothing to settle". Reproduced against prod before
changing anything.
WHY IT HID FOR TWO DAYS. It is volume-triggered. Daily volume ran 20-260 rows
and settled perfectly for weeks; 2026-08-01 was the first day past the 500-row
fetch limit. And the zero-settle ops alarm reads these very return values, so
pending:0 told the watchdog the backlog was empty -- the alarm built to catch
exactly this could not see it.
THE FIX. The refetch existed only to add game_date/settle_attempts/
dclv_computed_at. Selecting them in the first query removes the id list
entirely, so there is no URL to overflow at any volume. A failed fetch now
surfaces its error instead of being reported as an empty backlog.
captureClosing carried the same shape one level down -- .in('id', g.ids) on an
UPDATE, which fails identically once a single line|odds group gets large on a
big slate. Its id filters are now chunked at 100 (~3.7 KB).
Tests: the regression is locked by asserting settlement issues NO id-list
filter at 500 rows, and that a failed fetch is never reported as an empty
backlog -- the two properties that would have caught this. Two existing
suites asserted the old two-query shape and were updated to the real one.
4,159 tests green (332 suites); web build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9
|
||
|
builtbykev
|
2394fb04a1 |
Record the hits-v1 fingerprint as PENDING, and why
The prod-write fingerprint did not land: the odds provider is returning 429 (quota exhausted), so the snapshot refuses with gradeCount 0 and the MLB board has been frozen since 07:30 UTC. The 14/19/22 UTC cron slots failed the same way, all before this change deployed -- hits-v1 sits inside the snapshot's existing try/catch, is purely additive, and had zero grades to attach to. Firing is already verified against the real production snapshot through the real attachProjection path (158/159). What is pending is only confirmation that the deployed process writes the columns, which needs a slate the pipeline can fetch. The exact fingerprint query is recorded in the spec. The odds quota exhaustion is a live outage of the whole grading pipeline and is flagged for its own order, not folded into this one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
07626de3de |
hits-v1: built on the right structure, measured honestly, REFUTED
Hits was diagnosed as a family mismatch: 84% of hits rows trade at 0.5, so the stat rides on P(0), and a negative binomial has unbounded support and no notion of opportunity at all. hits-v1 models it as the bounded conversion it is -- N ~ the player's empirical at-bat distribution, hits|N ~ Binomial(N,q), with the multiplier scaling q (conversion) and never N (opportunity). STEP 0 confirmed the inputs before the model existed: 30/30 real ledger players, 100% combined-input coverage. Every read goes through knownRate -- a row with no atBats is dropped, never counted as a 0-at-bat game. It FIRES: 158/159 hits props (99.4%) on the live production snapshot, through the real attachProjection path. Scoping by book IDENTITY rather than price shape kept 94 out-of-promotion-band props on the board, 93 of them modelled -- 59% that a price rule would have deleted. And it LOST. Point-in-time replay (game log truncated strictly before each row's game_date, real grade-time multiplier), hits-only, direction-aligned, n=242: resolution champion 0.195 / ladder 0.048 / hits-v1 0.026. Paired bootstrap on the same rows: hits-v1 - ladder = -0.022, CI95 excluding zero. Not promoted. The value is in what it eliminates. The family was wrong AND the mean was not the constraint -- hits-v1 moved the line-0.5 mean 0.554 -> 0.581 toward a 0.598 base rate while resolution fell. What is left is per-prop discrimination: the ladder's inputs, not its distribution. The pre-registered fallback is recorded as WRONG rather than deleted. It said hits might be genuinely low-resolution for anyone; the champion scores 0.276 on the identical 189 rows, so there is real signal and the ceiling claim was the comfortable reading, not the honest one. Its own control refuted it, and that control was already in hand when the branch was written. hits-v1 stays wired as a challenger writing its own ledger columns so the forward accrual can confirm the backtest. Champion, ladder, ranking, calibration, reference ruler and the four accruing verdicts are byte-identical -- the diff has zero deleted lines. Tests 4,156 green (332 suites); web build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01W1sivYNqY2TS5ftykmHBU9 |
||
|
builtbykev
|
d103ecf4c3 |
Disambiguate takeable: THREE questions shared one word, now three names
BYTE-IDENTICAL. The audit found no consumer getting the wrong axis, so this is a disambiguation, not a bug fix. 4,131 tests / 331 suites green. STEP 1 AUDIT -- and the order's premise was wrong in a useful way: the four accruing challengers read the flag ZERO times (not four) the ranking gate wants PROMOTION, gets promotion [correct] the ledger column holds the LEDGER band, consumed as such the UI (LiveHeroProp) TYPES a `takeable` field it never renders THERE ARE THREE DEFINITIONS, NOT TWO -- and I only found the third by tracing the ranking gate: 1. IDENTITY can it be bet? book identity (takeability) 2. LEDGER BAND worth recording? odds >= -160, UNCAPPED plus 3. PROMOTION worth crowning? -160..+200, i.e. band PLUS a ceiling (2) and (3) genuinely disagree, and I measured it rather than asserting it: 439 rows -- 28.2% of all takeable=true ledger rows -- carry prices above +200, up to +1300. A +1300 longshot is a real bet worth RECORDING and not one worth CROWNING. Both are correct for their own purpose. THE DANGER WAS NEVER THE LOGIC. It was that three questions shared one word, so a reader could not tell which answer they held -- and hits, which must model thin/juiced/one-sided REAL markets, would have been the next reader to guess wrong. RESOLUTION: all three now have distinct names in config/takeability.js; gradeRanking calls isWithinPromotionBand so its intent is self-evident (a test pins it byte-identical to the old valueEngine call across the whole price range); the ledger dual-writes within_price_band with `takeable` kept as a documented DEPRECATED MIRROR so nothing breaks. Column comments in the database now say what each column actually holds. I did NOT redefine `takeable` in place. Four readers and a ranking gate sit on it, and silently changing its meaning under cover of a naming change is exactly the class of move this session keeps removing. Gates: 4,131 tests / 331 suites green; next build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
8c764c22a4 |
Structural hardening: unknown-is-not-zero + takeability-is-book-identity
Both guards are ADDITIVE. The full suite (4,111 -> 4,126 tests, 331 suites) passes unchanged through the migration, which is the evidence that no currently-correct output moved: served path, champion, reference ruler and the four accruing challengers are byte-identical. GUARD 1 -- src/utils/known.js. Number(null)===0 has produced at least SIX separate defects here, including one in a module written the same week its author documented the trap. Per-module vigilance has demonstrably failed, so the rule lives in one place and SEVEN sites now delegate: platoonSplits, projectionChallenger, challengerProjection, contactChallenger, statcastAggregateService, consensusRuler, gradeRanking -- plus compoundTotalBases moved onto knownRate. Two functions, deliberately: knownNumber (any finite number -- a REAL 0 is a fact and must survive) and knownRate (non-negative, rejects booleans -- for counts/rates where `true` or -1 is broken, not thin). Collapsing them is how the next variant gets in. firstKnown() exists because `a || b` discards a measured 0 and `a ?? b` does not. MY OWN GUARD HAD THE BUG IT EXISTS TO PREVENT, and its own test caught it: Number([]) === 0, so an empty array coerced to a measured ZERO. Same trap wearing a different type. Both helpers now reject objects outright. GUARD 2 -- src/config/takeability.js. Takeability is BOOK IDENTITY and never price shape. Baseball prop markets are genuinely thin, juiced and one-sided, and all three are NORMAL structure: betrivers and hardrockbet legitimately quote one side only (5 such rows surfaced in yesterday's re-stamp), and a hits-over at -300 is a real placeable bet. A rule that inferred un-takeability from price extremity or one-sidedness would throw those away while still admitting a DFS book at an ordinary -119 -- exactly backwards, because the -119 is the fake one. THE DISTINCTION THAT MUST NOT COLLAPSE, now enforced by test: isTakeableMarket(book) -- CAN it be bet? (identity) isWithinPriceBand(odds) -- SHOULD we promote? (policy band, floor -160) A -300 DraftKings prop is takeable AND out of band; a PrizePicks -119 is in band AND not takeable. Independent axes. FLAGGED, NOT SILENTLY CHANGED: the ledger's `takeable` column is the PRICE-BAND answer, and its name predates this distinction. Four challengers and the ranking gate read it, so renaming or redefining it is its own order -- doing it here would have changed correct current behaviour under cover of a hardening change. Fixtures are REAL prod rows from the 2026-08-02 re-stamp, not invented. Gates: 4,126 tests / 331 suites green; next build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
f67245e1e5 |
Re-stamp A: 862 rows recovered by honest join (not 936 -- see deviation)
Database only; no application code changed, so the served path, champion and reference ruler are byte-identical. RESULT: 862 rows re-stamped from the takeable LOCK-TIME price in lock_lines, 862/862 now anchored to takeable books, tagged price_source='archive_restamp', quarantine lifted. 812 pending clean rows recovered into the accruing verdicts. Holdout verification: 2,792 rows, 862 re-stamped included, 0 re-stamped rows non-takeable, 144 still excluded, 0 quarantined rows leaked, and 0 NON-TAKEABLE rows remain in the holdout population since 2026-08-01. DEVIATION, stated rather than buried: the order authorised 936. That figure came from a takeable book posting the same LINE. Requiring what a re-stamp actually needs -- that book's price for the GRADED SIDE at LOCK TIME -- resolves 862. Of the other 74, 73 have a takeable side-price only OUTSIDE the lock window and 5 are genuinely one-sided markets. I did not widen the window to reach 936. A takeable price captured hours after the grade is a later market moment, not a lock price; substituting it is precisely the reconstruct-vs-join line this order was fenced against, and it would have been invisible in the totals -- showing only as a cleaner-looking 936. Those 74 were also RE-TAGGED, because their old label had become a lie: recoverable_same_line -> no_takeable_lock_price_for_side. A future attempt reading the old tag would have been invited to widen the window and call it recovery. takeable was RECOMPUTED from the recovered price rather than carried over -- the old flag was computed FROM the contaminated price and was wrong on its own terms. 101 rows had their flag change, which is the direct measure of how wrong it was. Provenance travels with the data (price_source), on the same principle as is_proxy: a value recovered by a later join is not identical in kind to one captured natively at grade time, even when it is the same number. EVIDENCE FOR THE NEXT ORDER'S INVARIANT: 5 of the excluded rows are one-sided TAKEABLE markets, and betrivers/hardrockbet legitimately quote one side only. A guard that inferred takeability from price shape would throw away real markets while still admitting a DFS book at -119 -- takeability is book IDENTITY, never price extremity or one-sidedness. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
e29ab6fd6a |
Takeable enforcement: verified on real rows, 1,006 tagged, re-stamp call ready
PART 1 verified by inducing the REAL rowsFromSnapshot over REAL lock_lines
rows from prod. Three cases, 0 non-takeable anchors:
Narvaez (dabble/kalshi/prizepicks/smarkets, NO takeable book)
-> book=null, price=null, takeable=null [honest absent]
Schwarber(bovada/dabble/novig/PINNACLE before draftkings)
-> draftkings +102 [pinnacle SKIPPED, proving TAKEABLE not MODEL]
Ohtani (dabble/onexbet before draftkings) -> draftkings -266
Narvaez is the case that matters: pre-fix he was stamped dabble +104
takeable=true; he is now honestly absent.
A HARNESS BUG RECORDED: my first verification pulled live /api/odds/mlb,
which returned {"error":"Odds data temporarily unavailable"}. The script
read that as 0 props and printed "all from takeable books? true" -- a
VACUOUSLY TRUE pass. I caught it only because I also printed the book list
and it was empty. Same family as the silent-false traps: a probe that finds
nothing looks identical to a probe that finds nothing wrong.
PART 2: 1,006 rows tagged via the purpose-built quarantine_reason at ROW
level with three sub-cases (recoverable_same_line 936, no_takeable_quote
49, takeable_line_differs 21). getModelAggregate ALREADY excluded
quarantined rows, so the public record and the n>=20 gate were clean
automatically; all five committed holdout scripts now carry the exclusion
explicitly.
PART 3 -- the re-stamp call is now fact-based. The takeable LOCK-TIME price
is recoverable for 936/1,006 (93.0%) from lock_lines, the correct
instrument. Only 431 appear in closing_captures, which is the wrong timing
for a lock price anyway.
LINE CONTAMINATION ANSWERED (previously unverified): the stored line
MATCHES a takeable book's line on 936 (93.0%), DIFFERS on 21 (2.1%), and is
unverifiable on 49 (4.9%) where no takeable book quoted the prop at all.
That makes it cleanly row-level: re-stamp the 936 as an honest JOIN and
recover 886 pending rows for the holdouts, or leave all 1,006 excluded.
Either way the 21 + 49 stay out -- re-stamping those would invent a lock
price, or a line, we never captured. Nothing re-stamped; Kev's call.
Gates: 4,111 tests / 330 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
5de464330c |
URGENT: anchor the ledger price/book/takeable to TAKEABLE books
Ships before tonight's settle. Served path, champion, ranking and the
reference ruler are untouched.
TWO leaks, not one. The audit found ledgerService.indexProps; tracing the
lock price found that snapshotService.indexOdds has the SAME defect -- it
also indexed the full props list, so gradedAt.odds (the price a grade is
locked at) could itself be a DFS or exchange price. Fixing only the ledger
would have left the contamination flowing in through the lock.
Both now gate on TAKEABLE_BOOKS -- deliberately NOT MODEL_BOOKS. pinnacle
is model-eligible and correctly not takeable, so a MODEL gate would
re-break this the moment pinnacle's feed recovers. A test asserts pinnacle
cannot anchor a price.
TWO INDEXES, TWO ROLES, because the row needs two different things from a
prop and they have different correctness rules:
PRICE / BOOK / TAKEABLE -- takeable books only.
GAME FACTS (game_time, game_date, team/opponent) -- book-INDEPENDENT.
First pitch is first pitch whichever book listed it, so these still
come from any book. Gating them too would drop otherwise-valid rows
for no gain.
Collapsing those roles into one index is precisely the bug.
No takeable quote leaves the key ABSENT and the price null. An honest
missing price beats a price from a book you cannot bet -- and it keeps the
takeable flag from being computed off a DFS number, which is what made it
wrong on its own terms rather than merely mislabelled.
Gates: 4,111 tests / 330 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
08e5c908e6 |
Takeable audit: the ledger is contaminated, and I caused it
READ-ONLY. Nothing enforced or fixed; the five challengers untouched. VERDICT: gaps exist, and one is LIVE CONTAMINATION of the ledger -- the exact table every accruing holdout resolves against. book, locked_odds and the takeable flag ITSELF are being stamped from books you cannot bet: DFS dabble (707 rows, 24% of all rows), offshore bovada (214), onexbet (42), exchange kalshi (7, mean |odds| 1120). 0% before 2026-08-01. 47.9% on 08-01. 42.5% on 08-02. It began the day I widened the books for display. LEAK LOCATED, not inferred: recordPipelineGrades indexes byKey over the FULL display-widened props list, then prefers that prop -- book: (prop && prop.book) || g.book, and locked_odds/takeable both fall back to oddsForSide(prop). The grade is computed on a MODEL book and the ledger row is then re-stamped from whatever book indexed first. The takeable flag is therefore not merely mislabelled: it is computed FROM the contaminated price, so it is wrong on its own terms. The served grade path is clean TODAY (428 grades, 100% MODEL books), so dedupeProps' gate works. But MODEL_BOOKS is NOT a subset of TAKEABLE_BOOKS -- pinnacle is model-eligible and correctly not takeable -- so the projection may anchor to a reference line by design. Harmless while pinnacle returns nothing; live again when it recovers. BLAST RADIUS bounded but growing: 47 contaminated rows have already settled (21% of settled rows since 08-01) and ~700 are still pending and will settle into the holdouts. The damage is mostly ahead of us, which is what makes this urgent rather than historical. NOT VERIFIED and not claimed either way: whether the stored `line` is also contaminated. It traces to the graded prop, but I did not check it end-to-end; the enforcement order should. The prediction-vs-reference distinction HOLDS and must not be collapsed: the prediction target must be takeable, while fair_prob / consensus / edge stay reference. The bug is not the three-way split -- it is that one write path ignores it. Stack sequenced in the plan: (a) takeable enforcement, (b) structural Number(null)===0 guard (hits will re-trigger it -- its 0.5 lines make P(0) the whole game), (c) hits. Carry-forward: tb-v1 verdict, the third pre-registered branch, and the 100s Cloudflare timeout vs a ~115s snapshot. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
aa1228ec42 |
tb-v1 report + plan: diagnosis on trial, branch pre-registered
Firing verified on a real prod snapshot: 10/10 total_bases props carry proj_tb_p_over. The snapshot HTTP call returned 524 (Cloudflare's 100s origin timeout vs a ~115s snapshot) but the work completed server-side -- confirmed from the ledger rather than assumed. Face validity is good and diagnostic: means agree almost exactly with the ladder (1.813 vs 1.833), so this is a SHAPE-ONLY intervention, which is what was intended. Component rates are plausible, and Carroll's triples rate (0.112, far above his peers) is a clean check -- he is a speed player and the model sees it. AN OBSERVATION I AM NOT RESOLVING BY EYE: tb-v1 reads systematically LOWER than the ladder (0.424 vs 0.540 at the same mean). That is the expected DIRECTION, since the NB overstates P(>=2) by treating a home run as four accumulating events -- but whether 0.424 is right or an overcorrection is not knowable from face validity. A ~1.8-TB hitter clearing 1.5 empirically sits nearer 45-50%, between the two. I am not claiming tb-v1 is better; the holdout decides. BRANCH PRE-REGISTERED, before the result, so the verdict cannot be reinterpreted afterward: improves -> family-mismatch HOLDS, similarity stays off the critical path, hits is next; does not improve -> hypothesis WRONG and the mean-weakness/similarity branch REOPENS. Also recorded: I hit Number(null)===0 in my own new module -- a null component rate treated as a measured zero, the difference between "never triples" and "we don't know his triple rate". A test caught it. Sixth appearance of this trap in this codebase, and it caught the person writing the warnings about it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
eabf3b5bcf |
tb-v1: model total_bases as a compound outcome (challenger)
Current ladder (proj_p_over_line) and champion p_win are BYTE-IDENTICAL. tb-v1 writes alongside them, on total_bases props only. STEP 0 -- components confirmed on real data, not assumed. statsapi has no singles field, but hits - doubles - triples - homeRuns reproduces stored totalBases EXACTLY on a real 10-game log. So the decomposition is exact, not an approximation. THE MODEL. Each component gets its own per-game Poisson rate; TB is their weighted sum, and the PMF is built by exact convolution rather than simulated (TB support is small). It inherits the SAME combined multiplier proj-v1.1 computes, so the two models differ only in STRUCTURE. Why this is the fix: with identical mean TB of 1.0, a pure-HR hitter and a pure-singles hitter get P(TB>=4) of 0.221 vs 0.019 -- a 12x difference an NB on TB alone cannot express, because it treats one home run as four events. A test asserts that separation, and asserts P(TB>=4) for a pure-HR hitter equals P(at least one HR) exactly. INDEPENDENCE IS AN APPROXIMATION AND IS LABELLED AS ONE: a plate appearance that becomes a double cannot also become a single, so the components are weakly negatively correlated and independent Poissons slightly overstate the tail. Closer to the truth than what it replaces; not a solved problem. HONEST-ABSENT throughout: fewer than 3 usable games, or no derivable component, returns null and the prop keeps the current ladder value. An inconsistent row (hits < extra-base hits) is SKIPPED rather than clamped to zero -- clamping would invent a plausible line out of a broken one. I HIT THE Number(null)===0 TRAP IN MY OWN CODE and a test caught it: a null rate passed a naive finite check and was treated as a measured zero, which is the difference between "this player never triples" and "we do not know his triple rate". Both tbPmf and tbMean now reject null/''/boolean strictly. Holdout committed: TB ROWS ONLY (49 of 437 settled -- averaging into other stats would hide the effect) and DIRECTION-ALIGNED, since the unaligned comparison is the artifact that accounted for 41% of the ladder's apparent loss. If tb-v1 does NOT improve, the family-mismatch hypothesis is wrong and the mean/similarity branch reopens -- recorded in the query header. Migration applied: proj_tb_p_over + proj_tb_meta, NULL-meaningful. Gates: 4,104 tests / 329 suites green; next build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
48706210fe |
Diagnose proj-v1.1: concentrated mean failure, NOT a similarity problem
READ-ONLY. Nothing built or fixed; the four challengers untouched. 41% OF THE REPORTED GAP WAS A MEASUREMENT ARTIFACT. p_win is P(graded side); proj_p_over_line is P(over); 31.4% of settled rows are UNDER-graded, so comparing them raw measures the ladder backwards on a third of the sample. Matched + direction-aligned (n=437): 0.252 vs champion 0.352, not 0.108 vs 0.331. The PRODUCT is not making this mistake -- I checked; projectionChallenger normalises both to the over basis deliberately. The error was in the measurement. THE LOSS IS CONCENTRATED. hits (n=245, res 0.060) and total_bases (n=49, res 0.009) are 67% of rows and carry essentially no signal. Everything else is fine or better: walks 0.519 vs champion 0.544, runs mean 0.345 vs 0.392, and on DOUBLES the ladder's mean BEATS the champion's (0.207 vs -0.062). IT IS THE MEAN, NOT THE SHAPE. On the two failing families the mean itself carries no signal (0.052, -0.019) against the champion's 0.158 and 0.085. Where the mean is good the probability is good -- shape follows mean. A HYPOTHESIS I TESTED AND DISPROVED: prediction compression. I expected P(>=1 hit) to sit in a narrow band and fail to rank. It does not -- spread ratio 0.94 overall, 0.80 for hits, 0.94 for total_bases. The ladder has comparable spread; it is spread in a direction uncorrelated with outcomes. Recorded because it was a plausible story the data refused. PRIORS AND PLUMBING CLEAN. proj_factors carries form_rate, combined_multiplier and breakdown on every row; proj_point 100% populated with sane centres (hits 0.830 vs line 0.578). Not the environment-style silent-null failure. NAMED CAUSE (structural, flagged as hypothesis not finding): the count model mismatches those two stats. total_bases is a WEIGHTED SUM (1B..HR = 1..4), so an NB treats one home run as four events and mis-states variance -- and TB has the worst result in the table. hits is BOUNDED BY AT-BATS and mostly traded at 0.5, so almost everything rides on P(0), the region where the wrong family hurts most. walks/runs/doubles ARE genuine low-rate counts and are exactly the ones that work. FIX BRANCH: targeted per-stat fix for hits and total_bases. THIS REMOVES THE MLB SIMILARITY BUILD FROM THE CRITICAL PATH -- that branch assumed a GLOBAL mean weakness, and the mean is fine or better on three of six stat families. Similarity may be worth building later, on evidence, not on this. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
f5997778a2 |
Dormant-layer audit: nothing to connect; proj-v1.1 is live and losing
READ-ONLY. Nothing connected, built or wired; the accruing challengers were not touched. "Dormant" meant three different things and in no case is the answer "connect it". DISTRIBUTION LADDER IS NOT DORMANT. projection/distribution.js is consumed by projectionChallenger (proj-v1.1), live on every snapshot at 94.2% coverage (276/293) with 437 settled rows since 2026-07-23. It is a FOURTH accruing challenger, and it is LOSING: resolution 0.108 vs the champion's 0.331. That verdict is no longer thin. It is also PER-STAT and doctrine-correct -- nine distinct league priors (hits 0.90, total_bases 1.45, home_runs 0.15, ...) each feeding a gamma-Poisson posterior into a negative binomial. Correcting the plan: §10.3's "single additive index across hits/Ks/TB" is engine1's GRADE, not this ladder, which made a solved problem look open. SIMILARITY IS WRONG-SPORT. Zero callers, and its weights are NBA vocabulary: pace 0.15, referee_tendency 0.06, lineup_context 0.12, score_state_context 0.05, travel_fatigue 0.08. MLB has no pace and no referees. Connecting it would be the sport-stubbed-in-on-another-sport's- template breach, and it would fail QUIETLY -- missing factors are skipped, so the score would silently collapse onto whatever few dimensions happened to exist. CONSTRUCT, not connect. BAYESIAN WOULD REGRESS THE MODEL. Zero callers, and DISTRIBUTION_SHAPES keys on rbis / runs_scored / strikeouts_batter / outs_recorded / pitcher_strikeouts / walks_allowed / pitches_thrown -- NONE of which are live stat keys (S41: they are rbi / runs / outs / strikeouts). getDistributionShape defaults to 'normal' on an unknown key, so wiring it as-is would model COUNT stats as Gaussian, silently, on most MLB props. It is also superseded by distribution.js. Do not connect; retire or rewrite. DEPENDENCY, inverted: a better mean would help the ladder, but the ladder is already connected and both would-be foundations are unusable -- so this is not "connect similarity first", it is "the ladder is live and underperforming, and strengthening its mean requires BUILDING an MLB similarity layer that does not exist". Next-order pointer moved to diagnosing proj-v1.1: the only candidate already carrying settled evidence, and its diagnosis decides whether the similarity build is worth doing at all. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
ec815b0e37 |
Matchup axis report + plan reconciled: three challengers now accruing
Records the verification that matters: firing measured on a real prod snapshot rather than inferred. environment 248/293 (84.6%) -- also its FIRST confirmed ledger write, which the previous session could only infer -- and matchup 243/293 (82.9%) on tier batter_own_split. Both were 0/634. Collinearity guard passed at n=243: r = -0.003 vs the projection, +0.074 vs p_win, +0.003 vs line, -0.068 vs environment, -0.150 vs opportunity. The axis is not re-encoding recent form. The nudge distribution is also the right SHAPE -- mean +0.0007, 123 positive / 120 negative -- a balanced two-sided signal; a one-sided distribution would have suggested a sign or baseline error. Plan reconciled in place: arch-v1 condition axes marked firing, three challengers listed with coverage and their own holdout queries, and the next-order pointer moved to connecting the still-dormant layers (similarity, Bayesian, distribution ladder) with archetype_x_archetype as the named alternative. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
9ebd77b68e |
Build the matchup/platoon axis: three joins fixed, axis now FIRES
The axis was already wired and firing on 0/634 prod rows. Three separate
absences kept it silent, and all three are now joined:
1. oppPitcherByTeam 0 -> the self-origin /api/schedule/mlb/pitchers route
returned nothing in prod. Added the statsapi probable-pitcher hydrate as
a fallback, mirroring the one the schedule step already uses. 29/30
team-sides, one free request.
2. handById 0 -> follows from (1); the batched people call now has ids.
3. bats 0/120 -> batter hand rode ONLY on statcast aggregate rows, which do
not cover the slate. The season player list we ALREADY fetch and cache
carries batSide on 1342/1342, so this is a join, not a fetch.
Switch-hitters ('S') are preserved as-is; platoonSplits decides what to
do with them, not the map.
Verified end-to-end against the live API: opp_declared 29,
pitchers_with_hand 29, batters_with_hand 1342, and a real read --
multiplier 0.966, L vs R, 287 observed PA, weight 0.324 -- composing
alongside environment in one challenger.
FALLBACK LADDER, and a deliberate deviation from the order. Shipped tier:
`batter_own_split` (the hitter's OWN vs-L/vs-R line, regressed toward HIS
OWN overall rate), labelled on every adjustment.
`league_generic` is deliberately NOT implemented. platoonSplits already
handles thin evidence by regressing toward the hitter's own rate, which
covers the thin case per-player; its own doc-comment argues a hitter with
no split evidence should get NO adjustment. A league split applied to such
a hitter models the LEAGUE, not the player -- the doctrine breach the order
itself names in the same step. Adding it would have produced more firing
rows and a weaker signal.
`archetype_x_archetype` is scoped, not built: it needs the opposing
starter classified per game, which is real work and a separate order. The
tier vocabulary is in place for it.
Honest-absent on every join: no starter, no pitcher hand, or no batter hand
-> NO matchup adjustment, never a fabricated neutral. A neutral multiplier
produces no adjustment row at all.
Holdout committed (scripts/matchup-axis-holdout.sql), filtered to
matchup-carrying rows, and it keeps MATCHUP'S OWN nudge visible rather than
only the combined challenger -- arch-v1 composes four axes into one
p_win_challenger, so a combined-only view could not tell which axis earned
the movement, or which one is dragging.
Champion p_win, ranking, calibration, the armed invariant and the two
accruing verdicts are untouched.
Gates: 4,093 tests / 328 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
9fc17a4689 |
Fix the team resolve properly: backfill the name BEFORE confirmation
My first attempt did not work in prod -- team stayed 0/323 after deploy.
I resolved the team name AFTER the hint-confirmation check, but the check
itself reads hit.currentTeam.name, which is undefined because
/sports/1/players returns { id, link }. With a FULL-NAME hint (what
snapshotService passes) neither branch of teamRecordMatchesHint could
match: the name branch had no name, and the abbr branch cannot resolve a
full name to an abbr. Confirmation failed, the team was nulled, and my
later backfill ran on an already-null value.
withTeamName() now backfills the name from the cached /teams list BEFORE
any comparison, and is used at all three confirmation sites plus the
return. Verified against the live API on all four cases: no hint, FULL-NAME
hint, abbr hint -> "Philadelphia Phillies"; WRONG hint -> null.
That last case matters most: a wrong hint must still REFUSE. The
confirmation exists so a namesake collision cannot tag a player to a team
he is not on, which would fabricate opponents downstream. Making the match
succeed must not make it succeed wrongly, and a test locks it.
Gates: 4,087 tests / 327 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
cfda597fb5 |
Reconcile MASTER-PLAN to true state; next order = matchup axis
Reconciled in place, not regenerated. Next-order pointer now MATCHUP AXIS with its verified sourcing table, and an explicit note that SOURCE-LINEUPS-first is NOT needed. Marked DONE with their evidence: p_win ranking + edge retirement, calibration DECIDED, MLB isotonic DECIDED (provisional label retracted), grade cap 25->500 (board 7->365+), book widening, S59 invariant armed, environment axis repaired. Records the honest shape of Phase 1: it is further along than the phase table implied, but mostly because the work turned out to be CONNECTION AND REPAIR rather than construction -- the ladder question dissolved, the cap was discarding 95.7% of the slate, and two condition axes were wired but firing on zero rows. Carried forward without softening: WNBA is NOT BUILT rather than failed, and the ruler is MARKET-not-SHARP with PENDING-RECOVERY status until PropLine answers the Pinnacle question -- not to be enshrined as permanent. Remaining ~19 orders, ~9 unblocked. The two accruing verdicts are time, not code. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
03efdda33c |
Arm the S59 invariant by fixing its input; matchup sourcing = BUILDABLE
PART 1 -- PREMISE CORRECTION, then the real fix.
The order said the invariant's blocker was removed because "team is now
populated 416/416". It is not: what became 416/416 is home_team/away_team.
`team` (the PLAYER'S roster team) is still 0/416. Arming the guard off
home_team would compare the prop's game to itself -- always a match, a
permanent no-op that LOOKS armed. That would be worse than leaving it
disarmed, because it would read as a working guard.
The guard is also ALREADY fail-safe by construction (`if (knownTeam &&
gameTeams && ...)`), so Part 1's requirement was met in code all along.
What was missing was the data.
ROOT CAUSE: /sports/1/players returns currentTeam as { id, link } with NO
name, so searchPlayer's `hit.currentTeam?.name` was ALWAYS undefined and
every resolve returned team: null. The id is present on 1342/1342 and the
/teams list (already cached 24h) maps id -> name, so resolving it costs no
new request. Verified: Schwarber -> Philadelphia Phillies, Ohtani -> Los
Angeles Dodgers, Judge -> New York Yankees.
Five tests lock the fail-safe: drops only on a positive not-in-game;
abstains on unknown player team; abstains on unknown game participants;
and a row carrying only home_team/away_team does NOT satisfy the guard --
so the tautology can never be reintroduced.
PART 2 -- MATCHUP SOURCING: BUILDABLE. Measured on tonight's real board
against the free feeds, by VALUE not endpoint presence (the environment
trap: wired and null 634/634):
opposing starter 29/30 team-sides (home 14/15, away 15/15)
pitcher hand 1342/1342 (pitchHand.code)
batter hand 1342/1342 (batSide.code; L 416 / R 848 / S 78)
SHARED DEPENDENCY, and it is the finding: /sports/1/players -- a list we
ALREADY fetch and cache -- carries currentTeam.id, batSide AND pitchHand.
One join unlocks the invariant's input and two of the three matchup inputs
at once. The third (probable starter) comes from the schedule hydrate that
already exists.
So matchup is BUILDABLE and is the next order; SOURCE-LINEUPS-first is NOT
needed. Archetype-level reach on the opposing starter is available too
(the SP resolves to a player id, so the existing classifier applies) --
noted, not built.
Champion p_win, ranking, calibration and both accruing verdicts untouched.
Gates: 4,082 tests / 327 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
0d43fb7db8 |
arch-v1 axis audit: env/matchup were dead; environment fixed
Report for the audit + fix already committed. Records the two things worth carrying forward: 1. The environment axis has NOT yet been observed writing to the ledger, and I am not claiming it has. recordPipelineGrades upserts with ignoreDuplicates and dedupes on (user_id, player_key, stat, line, side, game_id) -- correctly, so a re-run never overwrites the original lock. Today's 429 rows predate the fix, so the axis cannot backfill onto them; first ledger observation is tomorrow's slate. What IS directly verified is the resolver (105/120) and the join key (416/416) -- the two things that were actually broken. 2. Matchup is not fixed and is not claimed as fixed. It needs the opposing starter and BOTH hands, and the audit shows three separate absences: oppPitcherByTeam 0, handById 0, bats 0/120. Fixing the pitcher feed without the hands, or the hands without the feed, still produces an axis that fires on zero rows. Also noted: the S59 slate JOIN INVARIANT keys off the same null `team` field, so it is currently inert too. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
4435856f46 |
Audit finds env/matchup axes DEAD in prod; fix the environment join
STEP 0 AUDIT -- the "already partly live" premise was half true: the CODE is wired, the axes are NOT firing. Across 634 graded prod rows the environment and matchup axes fired on ZERO rows, while 13 archetype axes fired normally (power 80, swing_miss 69, contact 56, launch 51, line_drive 43, ...) plus opportunity 142. Ledger confirms it from the other side: env_multiplier, env_park_base, env_weather_mod, wx_forecast and env_weather_state are ALL null on 634/634. ROOT CAUSE, located rather than inferred. A drop-off audit against the live snapshot: with_team_field 0/120, with_bats 0/120, with_playerId 120/120, oppPitcherByTeam 0, handById 0. `team` is a KEY on every stored grade and NULL on 416/416 -- so an environment resolver keyed off the player's roster team could never find a venue, while buildContext sat there with all 30 teams mapped and 14 weather forecasts resolved and unused. Coors composes to 1.241 the moment it gets a key. FIX -- and it is the more correct join, not just a workaround. The park and the weather belong to the GAME, not to the player's roster team, and the game rides on the prop from the odds feed. gradeBestSide now carries home_team/away_team onto the graded row (the legacy grade shape dropped them), and contextFor joins on the game first, keeping the roster team as a fallback. This no longer depends on a stats-resolve that can legitimately fail. MATCHUP/PLATOON IS NOT FIXED HERE and is not claimed as fixed: it needs the opposing starter and both hands, and the audit shows oppPitcherByTeam=0, handById=0 and bats=0 on the slate -- three separate absences. Per "one axis at a time" that is its own order with its own diagnosis, not a second fix smuggled into this one. Champion p_win, ranking, calibration and opportunity_drift's accruing verdict are all untouched. Gates: 4,077 tests / 326 suites green; next build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
c9d56d5668 |
Audit endpoint: where the env/matchup context chain drops off
The arch-v1 environment and matchup axes fired on ZERO prod rows across 634 graded props while archetype axes fired normally, and buildContext works locally (15 games, 14 with weather, Coors composing to 1.241). So the failure is downstream of buildContext and has to be located, not inferred from an absence. Replays buildContext + contextFor against the CURRENT cached snapshot grades and counts the drop-off at each hop: team field present -> resolves to an abbr -> abbr matches a game -> environment produced; and bats / playerId / opposing-pitcher known -> matchup produced. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
48a2f764ac |
opportunity_drift: coverage 94%, collinearity PASSES, holdout n-blocked
STEP 1 -- input mapped and measured. opportunity_drift 94% coverage on 100 real props: 100% for batters (total_bases, hits, home_runs), 40-67% for pitchers, which is correct -- pitchers accumulate few at-bats so the ratio is genuinely undefined and ABSTAINS rather than being invented. STEP 2 -- THE COLLINEARITY GUARD PASSES DECISIVELY. Pearson r on n=94: drift vs l20_avg -0.020, vs l5_avg +0.027, vs ab_per_game -0.029. All essentially zero, so the axis is orthogonal to every existing projection input and carries information the projection does not already contain. That also validates the ratio-over-level decision EMPIRICALLY: ab_per_game is the same quantity over the same denominator as l20_avg, so the level would have been redundant. Dividing by the player's own baseline removed the collinearity -- r = -0.029 against the very quantity it is built from. STEP 3 -- live as a challenger, verified on prod over an induced 416-grade snapshot: 142 of 276 rows (51.4%) carry the opportunity axis, the challenger moved on 190 rows, mean |delta| 0.034, range -0.089..+0.108. Champion p_win and the live grade path are unchanged. STEP 4 -- HOLDOUT IS n-BLOCKED BY CONSTRUCTION and I am not manufacturing one. Settled rows carrying the axis: 0. Its first rows carry game_date 2026-08-01 -- games that have not been played. Running the test on rows the axis never touched would dilute the comparison with rows where challenger === champion by construction, making a null result look like a small positive one. Query committed for when n arrives; it filters to axis-carrying rows for exactly that reason, buckets before measuring reliability, and splits time-forward. BOTH metrics must improve or the axis is shelved. A MEASUREMENT TRAP RECORDED: the first prod run showed drift at 0% while ab_per_game read 94% -- indistinguishable from "the feature does not compute". It was the 120-second feature-vector cache serving payloads written by the previous image. A new feature field is invisible for one cache generation after deploy. I nearly reported it absent, having already confirmed atBats is present in the live statsapi payload and that the code produced drift = 1.05 locally on that exact data; the contradiction between those two facts is what saved it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
092f8f09cd |
Build opportunity_drift axis on challengerProjection (arch-v1)
Champion p_win and the live grade path are BYTE-IDENTICAL: the axis writes only to p_win_challenger / challenger_adjustments in the ledger. STEP 1 -- MAP THE INPUT. MLB_LOG_FIELD now maps at_bats -> 'atBats'. Deliberately NOT added to outcomeService's map or liveTracking's LIVE_BOX_FIELD: those exist to SETTLE and TRACK graded props, and nothing grades at-bats, so adding it there would imply a settlement path for a market we do not carry. A test asserts the settle map still lacks it. STEP 2 -- DRIFT, NOT LEVEL. opportunity_drift = mean(last-5 atBats) / (season atBats / games). The LEVEL is collinear with l20_avg (same games denominator; hits/game ~= (hits/AB) x (AB/game)), so the projection already embeds it multiplicatively and adding it would double-count. A deviation from the player's own baseline is the part the projection does not contain. HONEST ABSENCE throughout: fewer than 3 at-bat rows, no at-bats in the logs, or no season baseline all leave drift UNDEFINED -- never 1.0 by default and never 0. Number(null) === 0 here would read as "zero at-bats", the strongest possible fade, invented from missing data. Four tests cover the absent paths. STEP 3 -- THE AXIS. opportunityNudge composes in the same log-odds space as park and platoon (log of a ratio), with two guards the measured axes do not need: a +/-10% DEADBAND (a rest day or a blowout can move a 5-game window without any role change) and a tighter cap (0.15 vs the environment's 0.30) so a noisy PROXY cannot outvote measured signals. Every adjustment carries is_proxy: true and proxy_for: 'confirmed_batting_order' so nothing downstream can mistake it for a lineup feed. The axis can stand ALONE -- without it the early return would gate opportunity off on exactly the thin-classification rows it is most likely to help. Zero extra I/O: analyzeViaEngine1 attaches drift from the feature vector it has already built, and attachChallenger reads it off the grade. Nothing re-fetches in a loop that runs over hundreds of props. COLLINEARITY GUARD added to the coverage probe: Pearson r of drift against l20_avg / l5_avg / ab_per_game, returning null under n=8 rather than reporting a correlation on a handful of rows. If drift just re-encodes the projection, the axis is dead signal and gets shelved. Gates: 4,073 tests / 326 suites green; next build exit 0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
8a02c75aec |
Step 0 input check: stop before wiring opportunity, and why
READ-ONLY. Live grade path byte-identical -- no layer wired, no threshold
moved, no challenger added, no holdout run.
INPUTS ARE 100% POPULATED (n=80 real MLB props, through the grader's own
path): ab_per_game, rest_days, l5_avg, l20_avg, l10_stddev and
game_count_in_7d all 100%; opp_rank_stat 65% overall and 0% on
stolen_bases. So there is no honest-degradation problem to solve.
FOUR FINDINGS THAT STOP THE WIRING, three of which would have made the
work unmeasurable or wrong:
1. THE PREMISE IS WRONG. There is no built opportunity layer to connect.
ab_per_game is consumed in exactly one place -- analyzeViaEngine1:379,
which renders "4.3 AB/G" on the grade card. engine1 has NO opportunity
or usage factor at all. A projected opportunity was never built;
building one is construction, not connection.
2. THE INPUT IS THE WRONG SHAPE. ab_per_game = season atBats/games. It is
a per-player CONSTANT (measured: varies for 3 of 20 players, and those
cannot be legitimate since the value can't depend on stat_type), so it
can only move all of a player's props together, never separate them.
And it is collinear with the projection: l20_avg = seasonTotal/games,
the SAME denominator, so l20_avg already embeds opportunity
multiplicatively. Adding it additively double-counts.
3. THE REAL INPUT DOES NOT EXIST. depthChartService returns battingOrder:
null for MLB ("the one lineup slot the free schedule feed exposes") and
PropLine /context carries lineup_confirmed as a BOOLEAN, not the order.
4. ARCHITECTURE: wiring it into engine1 would be unmeasurable BY THIS
ORDER'S OWN TEST. Step 2 proves reliability and resolution, both
measured on p_win. engine1 factors move the grade LETTER and never
touch p_win. The layer belongs in probabilityEstimator, which already
adjusts on opp_rank_stat, home_away and a consistency pull.
SEQUENCING IS ALSO STALE: challengerProjection (arch-v1) is already live
with archetype, matchup (platoon) and environment (park) axes, writing
p_win_challenger to the ledger. Step 2 of the order's sequence is partly
done -- and the harness this order needed already exists.
RECOMMENDED INSTEAD, as its own order: an `opportunity` axis on that
harness driven by DRIFT, not level -- recent AB/G (last 5) over season
AB/G. A deviation is not collinear the way the level is. Per-game atBats
is present in the statsapi log rows but MLB_LOG_FIELD never maps it, so it
is a small contained BUILD, which is why it gets its own order. Honest
caveat carried forward: it is still a proxy, not tonight's opportunity.
PROBE BUG RECORDED: the first run reported 0% for every feature including
l5_avg, on a pipeline that had just graded 365 props -- impossible, so the
probe was wrong. getFeatures takes camelCase and returns { features: {} };
I passed snake_case and read the top level. Fixed to call
computeFeaturesForProp. Same class as the earlier silent-false harness: a
measurement that makes working code look broken invites you to "fix"
something that was never broken.
Gates: 4,059 tests / 325 suites green; next build exit 0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
212c08b11f |
Fix the Step 0 probe: it was measuring itself, not the pipeline
The first run reported 0% coverage for EVERY feature including l5_avg --
which projectionFor requires, on a pipeline that had just graded 365
props. That is impossible, so the probe was wrong, not the pipeline.
Two bugs, both in my probe: featureCache.getFeatures takes camelCase
(playerName/statType) and I passed the prop's snake_case shape, and it
returns { features: {...} } while I read the top level. Either alone
yields all-zeros.
Now calls computeFeaturesForProp -- the grader's own entry point -- so it
measures what the grade path actually sees. Same class as the earlier
harness that returned a silent false: a measurement that makes working
code look broken is more dangerous than no measurement.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
3e78217678 |
Step 0 input check: read-only feature-coverage probe
Before wiring any layer into the grade, measure whether its inputs are actually populated on real props. A layer wired onto sparse inputs does not degrade gracefully by default -- Number(null) === 0 turns a missing opportunity into 'zero opportunity', a fabricated input rather than an absent one. Reports population per feature, SPLIT BY stat_type, because a feature can be 100% present for batters and 0% for pitchers and a pooled number would hide exactly that. Also reports whether ab_per_game varies across a player's own props -- a per-player constant can only move all of a player's props together, which is a very different thing from a per-prop opportunity signal. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
ecdc644621 |
Fix superseded assertion after the ?limit= bisect hook
runSnapshot now takes an opts object, so the route call is ('mlb', {}).
Asserted as EMPTY rather than loosened to any-object: a stray limit
reaching production would silently cap every run, which is the exact bug
the hook exists to diagnose.
I pushed the previous commit without reading the suite result -- the
failure was already on screen. Caught and fixed immediately after.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
|
||
|
builtbykev
|
11b0139481 |
Verify the cap raise on prod: 7 -> 365 graded props
Induced, not projected. DEFAULT_LIMIT=500 produced 365 graded props in 114s (was 7 in 16s) -- 52x the board. All 365 carry a unique forecast_rank and ZERO leak p_win to anonymous callers, so the tier gating holds at 50x the volume. Anon payload 220KB in 0.44s. Stat mix went from three stats to ten. Health green. Measured cost curve via the ?limit= bisect hook: 1->42s, 25->58s, 60->42s, 120->66s, 500->114s. About 42s of that is FIXED overhead (odds fetch, roster logs, archetype classify, retention), paid whether we grade 1 prop or 500 -- grading is the cheap part. MY PRE-FLIGHT ESTIMATE WAS WRONG. I predicted ~72s from per-prop latency measured in isolation, which ignored the fixed cost. Real figure 114s. A FALSE ALARM RECORDED because acting on it would have meant reverting a fix that works: the first induced run 502'd at 13.4s and I hypothesised load -- memory or a proxy timeout under 20x the work. Wrong. A limit=25 run then 502'd in 2 seconds, which no amount of load explains, and both recovered on retry. The 502s were the deploy rolling, not the cap. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
8d052131c5 |
Add a bisect hook (?limit=) to the internal snapshot trigger
The cap raise 25 -> 500 made an induced snapshot 502 at 13.4s and the run did not complete in background either, while a 25-prop run had completed in 16.3s. That rules out a simple duration timeout and means the cause has to be measured, not guessed. ?limit= bounds one run so the regression can be bisected without a prod env change; omitted, the real DEFAULT_LIMIT applies. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |
||
|
builtbykev
|
a7d6cf8e36 |
Raise the grade cap 25 -> 500 on measured cost; refusals are correct
PART 1 (read-only, measured on a live prod slate, n=80) OVERTURNS THE
PREMISE. The refusal rate is not a data problem -- it is 98% correct
behaviour. The cap is the entire problem, and it is worse than "25 of 546".
Composition: GRADED 44 (55.0%) | POLICY-SUPPRESSION 35 (43.8%) |
FETCHABLE-GAP 1 (1.3%) | FALSE-THRESHOLD 0 | ARCHETYPE-GAP 0 |
GENUINE-ABSENCE 0.
THE FIFTH BUCKET the order did not anticipate: all 35 "refusals" are
rare_event_over_below_line -- the 2026-07-19 betting-logic audit
deliberately refusing 0.5-line rare events, setting the SAME
insufficient_data flag as a real data gap, which is why they read as one.
They are entirely doubles (18) and stolen_bases (17), while hits (19/19),
rbi (19/19) and total_bases (5/5) grade at ~100%. Had we "fixed" this we
would have re-introduced exactly the bets a previous audit removed, and the
count would have looked like progress.
THE CAP: 585 unique gradeable props, cap 25 -> 560 discarded (95.7%).
Traced to Session 32 (
|
||
|
builtbykev
|
d18a19f6aa |
Part 1 diagnostic: read-only refusal categoriser (25-cap + 72% refusal)
READ-ONLY. Runs the REAL grade path over a REAL slate and categorises every refusal; writes nothing. Reproduces gradeSlateService.dedupeProps exactly (MODEL_BOOKS, first-row-wins) and calls analyzeViaEngine1 the same way, so it measures what the pipeline does rather than a re-implementation. Adds a FIFTH bucket the order did not anticipate, and it is likely to change how the 72% is read: (e) POLICY-SUPPRESSION. The 2026-07-19 betting-logic audit deliberately refuses rare-event 0.5 markets (doubles/ triples/HR/SB) on the juiced under, plus any over-juiced price -- and it sets the SAME insufficient_data flag as a genuine data gap. Counting those as a data problem would send us hunting for data that is not missing, and "fixing" them would re-introduce bets we removed on purpose. Separates (b) FETCHABLE-GAP from (d) GENUINE-ABSENCE by asking the stats layer directly whether the player has ANY game log, rather than assuming: no log -> genuine absence, keep refusing; a log that exists while the grade path found no projection -> a wiring gap with something to fix. Also measures per-grade latency (mean/median/p90/max, serial and at concurrency) so Part 2 can decide the cap on cost rather than on taste. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc |