Minus 18.69 percent. That is where the totals STRONG LEAN tier ROI landed after the break — down from minus 16.59 percent when I wrote the mid-season report on July 24. That is the two-percentage-point drop I said would trigger the totals-tier structural downgrade.
Effective tonight, the framework changes as flagged in Issue 58:
Totals HIGH: quarter stake (down from half stake). Totals STRONG: PASS entirely (already the default). Totals MODEST: PASS entirely (already the default).
We are no longer sizing totals HIGH at half stake. The tier's structural underperformance across a 108-pick sample is now large enough that the framework treats every totals HIGH pick with reduced conviction. If the tier ROI recovers above minus 15 percent, we can revisit; until then, quarter stake is the ceiling on totals exposure.
Post-break recap.
Tuesday July 28 had two HIGH picks on a five-game slate:
TOR at MIN spread TOR HIGH — WIN. Toronto pulled the road upset 100-93 as a plus-7 underdog. Half stake at pre-break sizing: plus 0.45u. Signature road win against a top-three West team.
CON at WAS spread CON HIGH — LOSS. Washington won 92-84. Half stake: minus 0.5u.
Net Tuesday: minus 0.05u. Essentially flat.
Wednesday July 29 had two games on the board and no HIGH picks. Rest night for the framework.
Cumulative through this morning per the grader:
Spread STRONG LEAN: 51-42-1 (54.8 percent, plus 4.64 percent ROI). Up from plus 3.90 percent pre-break. Still the healthiest tier in the system.
Spread MODEST LEAN: 23-27 (46.0 percent, minus 12.18 percent ROI). Drift continues, but this tier has always been coin-flip territory with vig drag.
Total STRONG LEAN: 46-62 (42.6 percent, minus 18.69 percent ROI). The tier that triggered the rule change.
Total MODEST LEAN: 27-28 (49.1 percent, minus 6.28 percent ROI). Consistent.
Tonight's three HIGHs.
CON at CHI spread CON HIGH, edge plus 5.40. Model has Chicago slightly favored (minus 0.9); market has Chicago favored by 4.5. Both teams have stars Out (Griner for Connecticut, Diggins for Chicago). Injury edge component plus 2.0. Not injury-dominant.
Agent read: two HIGH-tier quarter edges on the Chicago side (Q3 plus 2.90, Q4 plus 3.70). That is one short of the three-HIGH-quarter-edge PASS trigger, but both are late-quarter edges, which is the pattern that has produced the model-versus-agent conflict PASSes historically. This is a caution flag rather than a hard PASS. Sizing at half stake per Issue 42 default with the flag noted.
Sizing: HALF stake CON. 0.5u.
MIN at TOR total UNDER HIGH, edge minus 10.23. Spread is a SKIP (model has MIN by 13, market by 11.5, edge 1.7 below threshold). Total pick is on the UNDER 188.1.
This game carries the injury-dominant flag but only on the spread side — the total is a clean pick unrelated to the injury calculus. Sykes is Out for Toronto. Agent projected total: 180 (44.6 plus 44.8 plus 45.7 plus 44.9). Model total: 177.87. Both well under the market's 188.1. Full alignment.
This is the first pick sized under the new totals-HIGH rule.
Sizing: QUARTER stake UNDER. 0.25u.
NYL at LVA spread NYL HIGH, edge plus 5.95. Model has the game nearly pick-em (LVA by 0.35); market has Vegas favored by 5.6. Sabally is Out for New York.
Injury dominant flag: False. Injury edge component: 0.0. This is a structural edge, not an injury-driven one — the model views the Aces as roughly a pick-em team against a Sabally-less Liberty on Vegas' home floor, which is more optimistic on New York's road performance than most public reads.
Agent read: two MEDIUM quarter edges (LVA Q1 plus 2.40, NYL Q4 plus 1.90). One MEDIUM SGP idea called "Close game, NYL closes." No HIGH agent signals either direction. No conflict trigger fires.
The Sabally-Out factor is the reason to think this could be a bad pick — Vegas is at full strength, Wilson has been playing at MVP-level pace since her return, and New York without their primary interior threat is a tough matchup on the road. The model still likes NYL by 5.95, which suggests the market has overreacted to the injury news. Framework follows the HIGH tier at default sizing.
Sizing: HALF stake NYL. 0.5u.
Total exposure tonight: 1.25u across three sized bets.
The Tempo storyline. Toronto hosts Minnesota tonight for the immediate rematch after Tuesday's road upset. The framework has the total on the UNDER (quarter stake), no spread pick. Jordan has the Tempo Report with the full pre-game read on the return match, the Sykes-absent lineup that pulled Tuesday's win, and what a home sweep would mean for the playoff picture.
Watching tonight.
The NYL-LVA spread most carefully. This is the pick where the model's contrarian read (Vegas nearly pick-em at home against a Sabally-less NYL) is furthest from the market. If it hits, the model has correctly identified a market inefficiency. If it misses, the Sabally-Out variable was underweighted.
The MIN-TOR total is the first quarter-stake bet under the new rule. If it wins, the framework has correctly downgraded a tier that was still capable of producing HIGH-tier winners at the correct size. If it loses, we lose only 0.25u instead of 0.5u, which is exactly what the downgrade rule is meant to accomplish.
Back tomorrow with the grades and Friday's slate.
[ End Report ]
Share This