Twenty-two points. That is the margin Dallas won by Sunday against Seattle. The picks file had Seattle as a HIGH pick with a plus 7.05 edge. The framework sized at half stake per Issue 42 default. The agent had two HIGH-tier quarter edges on the Dallas side — DAL Q1 plus 2.80 and DAL Q3 plus 3.10 — which I flagged in the Sunday preview as a caution flag rather than a formal PASS because the two quarters were not both in the second half (Q3 and Q4), which is the current formal trigger from Issue 61.
Dallas won by 22. Twenty-two points. Blowout. Third time this pattern has fired this season.
The three-instance sample:
CON at IND on July 22 — three HIGH IND quarter edges + "IND dominant" SGP. Formal 3-HIGH PASS rule fired. IND won by 35 (final: 123-88). Framework saved 0.5u by PASSing.
CON at DAL on August 2 — two HIGH DAL quarter edges (Q3 plus 5.00, Q4 plus 4.70), both late-quarter. Q3+Q4 pattern. Sized at half stake before rule formalized. DAL won by 20 (final: 83-63). Loss of 0.5u. That was the game that led me to formalize the Q3+Q4 pattern as a PASS rule in Issue 61.
SEA at DAL on August 23 — two HIGH DAL quarter edges (Q1 plus 2.80, Q3 plus 3.10), Q1+Q3 split. Did not fire the formal Q3+Q4 rule. Sized at half stake. DAL won by 22 (final: 92-70). Loss of 0.5u.
Three-for-three on the underlying pattern producing blowout losses when the framework sizes anyway. The Q1+Q3 case is not structurally different from the Q3+Q4 case — both represent the agent identifying that the opposing team dominates multiple quarters of the game, and both correlate with 20-plus point margins going against the picks-file pick.
The rule tune is straightforward. Effective tonight and forward:
**Two HIGH-tier agent quarter edges on the opposite side of a picks-file HIGH pick (any two quarters) equals PASS.**
That is a broadening of the Issue 61 rule (which required both edges in Q3 and Q4) to include any 2-HIGH combination. The original 3-HIGH conflict rule stays in place as a stronger trigger. This new 2-HIGH rule is the tighter filter.
The tradeoff: 2-HIGH is more common than 3-HIGH. Formalizing this at 2-HIGH means the framework will PASS more picks per week — probably 4-5 per week instead of 1-2 with the older rules. That reduces edge-capture on HIGH picks that would have won, but the three-instance sample says the pass-through picks lose by an average of 25-plus points. Sizing 0 instead of half stake on those is worth the missed edge on the picks that would have won.
If the rule turns out to be overly conservative and starts PASSing too many HIGH-tier winners across the September stretch, we can revisit. But the current sample says it is the right call.
Framework recap: August 23 to 26.
Sunday August 23: SEA at DAL SEA HIGH — LOSS. minus 0.5u. (Rule fired retroactively.)
Monday August 24: GSV at MIN GSV HIGH edge plus 8.95 — WIN. Golden State won outright at Minnesota 80-66 as an underdog. Big edge, clean cover. plus 0.45u.
Tuesday August 25: No HIGH picks. Rest.
Wednesday August 26: GSV at CON total OVER HIGH quarter stake — WIN. Game landed at 153 (89 plus 64) against the OVER line, cleared by whatever margin the low market number was set at. Quarter stake sizing: plus 0.23u.
Four-day net: plus 0.18u.
That is a small positive stretch after the Sunday loss ate most of the Monday gains. Two of the three sized bets won, the third would have PASSed under the new rule going forward.
Cumulative check per grader (limited grades since last update — Odds API window seems to have shifted and older results are backfilling slowly).
Spread STRONG LEAN: 66-58-1 (53.2 percent, plus 1.60 percent ROI). Unchanged from the Aug 23 report.
Spread MODEST LEAN: 30-41 (42.3 percent, minus 19.33 percent ROI). One new win.
Total STRONG LEAN: 71-84 (45.8 percent, minus 12.55 percent ROI). One new win, ROI improved 0.67 percentage points. The tier continues its recovery trend.
Total MODEST LEAN: 35-29 (54.7 percent, plus 4.40 percent ROI). Unchanged.
The totals tier recovery watch continues. Trigger for reversing the quarter-stake rule remains totals STRONG above minus 10 percent ROI. Currently at minus 12.55 percent, closing in.
Tonight (Thursday August 27). Two games on the board, one HIGH pick.
GSV at NYL spread GSV HIGH edge plus 6.35. Model has GSV by 3.25; market has NYL as a 3.1-point home favorite. Sabally is Out for New York — she has been Out on the injuries index consistently for weeks. Injury edge component is not flagged as dominant.
Agent read: one HIGH-tier quarter edge on the GSV side (GSV Q2 plus 4.80). Same direction as the pick. No conflict fires. No caution flag. Clean HIGH.
Sizing: HALF stake GSV per Issue 42 default. 0.5u.
Total pick is OVER MEDIUM — PASS by default MEDIUM total rule.
Total exposure tonight: 0.5u.
The Tempo. The picks-file injuries index has listed both Sykes and Mabrey as Out every day from August 24 through today. Toronto has not appeared in a picks-file game recording during the interval, which means either the games are not being captured in the picks file (a bot-side issue) or the Tempo have not played this stretch (unlikely). The August 23 TOR vs LVA home game also does not have a graded result in the local card. What I can confirm from data: the Tempo's injury situation has not improved. What I cannot confirm: any Tempo game results this week.
Jordan will pick up the Tempo Report when there is verifiable game data to work with.
Back tomorrow with the GSV at NYL grade and Friday's slate.
[ End Report ]
Share This