← Recap journal

Settled model recap · 2026-07-28

CLE vs. CIN MLB Pick Recap: CLE +145 Win

Why the ATS Killers model selected CLE, how its probability compared with the sportsbook price, the frozen lineup and player-value context, the final score, and what the result does—and does not—teach.

Published after the saved Picks settled. The original prices and inputs stay attached.

Record1-0-0
Net units+1.45u
Game Picks1
Player Props0

One-minute recap

The quick read

There was 1 settled pick on the card. The result was 1-0 for +1.45 units on the flat-risk ledger.

The price, probability, lineup, and player values below are the ones saved before first pitch. The final score grades that read; it does not give us permission to rewrite it.

Saved Picks and reviews

Open any Pick for the why

MLB Game Picks

Game Pick

CLE moneyline

CLE at CIN · CLE 6, CIN 5

BetRivers +145winSee the review ↓

Postgame review

CLE +145: what the model saw

CLE won, and the final was CLE 6, CIN 5. The archived CLE moneyline at +145 from BetRivers graded as a win for +1.45 flat-risk units.

At the lock, BetRivers had CLE at +145. The model put the side at 40.0% against a 40.8% break-even mark. That was 0.8 points below the price's break-even mark. It still became the tracked daily card because it was the first fully observed matchup to clear the separate calculation-quality screen; that choice is a product baseline, not an edge claim. The largest listed model driver was Long-run team strength; on the receipt it pushed 0.38 probability points toward CLE.

The frozen matchup snapshot used a 0-100 rolling scale: the starters were Slade Cecconi (48.6) and Chase Burns (56.4); the lineup values were CLE 50.3 and CIN 50.6; the highest displayed hitter values belonged to Chase DeLauter (54.8) and Elly De La Cruz (55.9); available bullpen values were CLE 53.6 and CIN 45.9. Those are descriptive pregame values, not an after-the-fact claim about what caused the final score.

The win belongs in the 40.0% forecast band, but it does not prove that any single lineup, starter, or bullpen input was the reason. We keep the result, compare it with similar reads, and leave the production weights alone for now.

Saved pregame inputs

Model probability40.0%
Price break-even40.8%
Calculated difference-0.8 pts

Selected side

CLE

Team form50.7
Lineup form50.3
Starting pitcher48.6
Bullpen available53.6
Slade CecconiTop hitter value: Chase DeLauter 54.8

Opponent

CIN

Team form47.9
Lineup form50.6
Starting pitcher56.4
Bullpen available45.9
Chase BurnsTop hitter value: Elly De La Cruz 55.9

Values use the saved 0–100 rolling scale. They are inputs to the selection screen, not direct probability-point weights.

What comes next

What we learn—and what we do not

This game now joins the review queue beside the exact decision, matchup snapshot, and settlement receipt. The recap is the explanation layer; it is not allowed to train the model by itself.

If the same kind of miss or win repeats, we can test that pattern by pitcher role, lineup shape, price range, and other frozen segments. Only a repeatable result that survives a prospective challenger test can change the live model.

Technical receipt

Fact pack: 2211dd3f60ce64a2894628121a188991ae8e1d39a77a86541265a742e39fafdb

Decision receipts: mlb-current-decision:617df57a1a5a4adb3035b8a9d037a6d5

Settlement receipts: mlb_live_settlement_898a091640a7bef4b7d5e43a