Action: Open the Workshop, load an MLB template, and swap in roster-snapshot features for any deadline trades. Do not trust models citing current Vegas MLB lines.
The original MLB trade deadline guide assumed a live odds feed with daily MLB lines, team totals, and win-total futures that lag for 3-7 days after a trade. The August 23 odds feed says that market does not exist.
The feed returns college football, not baseball
26 unique games in the current odds table. 25 are college football. One is NFL. Zero MLB games.
Seahawks at Titans: -4.5 . That is the only NFL line.
Every college game shows one book and zero dispersion. The spreads are blowout territory:
Missouri -54.5 vs Arkansas-Pine Bluff . Minnesota -43.5 vs Eastern Illinois . UCF -42.5 vs Bethune-Cookman . USC -38.5 vs San Jose State . Oklahoma -41.5 vs UTEP . Kansas -38.5 vs LIU . Utah -33.5 vs Idaho . Rutgers -30.5 vs UMass . Florida State -31.5 vs New Mexico State . Wake Forest -23.5 vs Akron . Buffalo -24.5 vs UAlbany . Kennesaw State -22.5 vs West Georgia . Illinois -28.5 vs UAB . Delaware -28.5 vs Merrimack . Georgia State -28.5 vs NC A&T . Purdue -35.5 vs Indiana State . North Carolina -7.5 vs TCU . San Jose State -4.5 vs Eastern Michigan . Sacramento State -9.5 vs Eastern Michigan . Jacksonville State -7 vs North Dakota State . NC State -5.5 vs Virginia . Hawaii -5.5 vs Stanford . Memphis -5.5 vs UNLV . Colorado -7 vs Georgia Tech .
Zero dispersion on every line. One book. The market has not shown up.
What the original guide assumed, and what the feed broke
The old guide had a clear model. Lines move on credible reports 24-72 hours pre-deadline. Confirmed trades trigger sharp repricing in 5-15 minutes. A 7-day reset window follows where daily moneylines and team totals lag while season-long win totals absorb the trade. The cleanest angles: opponent team totals against new aces. F5/full-game splits on bullpen trades. Win-total fades on salary-dump sellers.
All of that requires MLB lines to exist in the feed. They do not.
The guide assumed team totals drive the math. Those totals come from the spread and over/under. If the spread is absent, the team total is absent. Any deadline model from it is a guess.
The EV tool and MLB picks page both require multiple books before they trust a line. The raw feed gives zero books on MLB.
The Builder lets you construct projections from roster snapshots instead of Vegas-implied totals. The Week 1 RB waiver wire shows how usage trends drive decisions when lines are unreliable. The track record page grades every model against closing lines.
What would change our mind
If the feed adds MLB games with 3 or more books and dispersion under 0.5 runs, Vegas-implied team totals become usable again. The number that flips the take is the book count on MLB games, not the spread. Until an MLB line appears with a range across several books, the deadline reset window is a theory with no data to ground it.
