Are the signals earning trust?

Performance

SP Strikeouts carries the strongest edge (+4.7 pts over base). SP Hits Allowed is the volume leader (444 graded, -4.4 pts).

What's working

Every market's measured edge against how much of it we actually serve, ranked by lift. The tables further down read served predictions only, so a market that is good but never clears the floor is invisible there — and looks the same as one that does not work.

MarketAll predictionsServed (70+)Share served

SP Strikeouts+4.7 pp n=364+8.1 pp n=25470%

SP Hits Allowed-4.4 pp n=444-2.7 pp n=24856%

“Starved” means the market beats its base rate by 5+ points across everything it predicted, yet under a tenth of those ever clear the floor — the scale cannot reach, so the edge is never offered.

Last 90 days · All qualifying292–21058.2%n=5025 void · 0 pending · avg score 80 · 22 slates

All qualifying292–21058.2%n=502

Featured2–433.3%n=6

Other qualifying290–20658.5%n=496

Deeper diagnostics

Market pulse

Recent slate hit rates compared with each market’s own average. The right edge shows the selected period and the latest three slates versus the prior three.

MarketPeriodAug28Aug27Aug26Aug25Aug24Aug23Aug22Aug21

SP Strikeouts64%n=254↑ +7 pp90%n=1060%n=562%n=1362%n=1362%n=869%n=1344%n=9100%n=8

SP Hits Allowed52%n=248↑ +18 pp69%n=1362%n=838%n=825%n=860%n=544%n=962%n=1327%n=11

Above own averageNear own averageBelow own averageFaded cells have fewer than 5 decisions

Does a higher score perform better?

Higher score bands have not produced higher hit rates in this sample. The 80–84, 85–89, 95–98 bands remain small.

Current scoring engine only — 350 predictions from superseded versions excluded. Pooling engines once made the top band look anti-predictive when it was not; per-version records are in the model table below.

Score bandRecordHit rateSampleVoidsvs base

70–7424–972.7%n=331+22.7 pp base 50%

75–7929–2950.0%n=582-4.3 pp base 54%

80–84Small sample0–20.0%n=20-46.6 pp base 47%

85–89Small sample0–10.0%n=10-50.0 pp base 50%

95–98Small sample16–1061.5%n=260+5.9 pp base 56%

99–10022–1264.7%n=340+13.3 pp base 51%

Over vs Under

Over0–0Not gradedn=0

Under292–21058.2%n=502

MarketOverUnder

SP StrikeoutsNot graded n=063.8% n=254

SP Hits AllowedNot graded n=052.4% n=248

Edge finder

SegmentHit ratevs baseRecord30-day trend

SP Strikeouts63.8% n=254+8.1 pp base 56%162–9264% 30d

SP Hits Allowed52.4% n=248-2.6 pp base 55%130–11852% 30d

Consistency

Last 7

59.3%

80–55 n=135

Last 30

58.2%

292–210 n=502

Prev 30

Not graded

0–0

Season

58.2%

292–210 n=502

All time

58.2%

292–210 n=502

By month

MonthRecordHit ratevs base

2026-08292–21058.2% n=502+2.8 pp base 55%

Model versions

VersionActiveRecordHit ratevs base

MLB SP Strikeouts

sp-v42026-08-21→2026-08-2854–2568.4% n=79+14.1 pp base 54%

3 earlier versions2026-08-03→2026-08-20108–6761.7% n=175+5.4 pp base 56%

MLB SP Hits Allowed

sp-v42026-08-21→2026-08-2837–3849.3% n=75-2.0 pp base 51%

3 earlier versions2026-08-03→2026-08-2093–8053.8% n=173-2.9 pp base 57%

All qualifying includes every published prediction scoring 70+. Featured is the main-page eight. Observed hit rates are not probabilities; prediction counts are correlated within the displayed slate count.