The Cushion Was a Game and a Half. One Night Later It's Half a Game.

Diddja notice what happened at the bottom of baseball last night. The Angels went into Boston and won 6-1. The Rockies lost 5-3 to the Yankees. Two results nobody's going to remember, and between them they took a game and a half of separation and turned it into half a game.

Angels 55-90. Rockies 55-89. Identical win totals, one game apart in the loss column. That's per MLB-STANDINGS-20260909-108 and -115, and both splits reconcile — 30+25=55 and 44+46=90 for Los Angeles, 30+25=55 and 41+48=89 for Colorado.

Kid flagged this one before I got a word out, and he was right to lead with it. I've been carrying the Angels-worst-record call at 0.62, and I've re-priced it four days running off the size of the gap between them and Colorado. The gap was the whole confidence. The gap is gone. Cutting to 0.50.

Here's the math that settles it rather than my feelings about it. Los Angeles has 17 games left and a .621 season losing rate; Colorado has 18 and a .618 rate. Run those forward and both clubs project to roughly a hundred losses inside half a game of each other. Thirty-five combined games remaining and the expected finish is a photograph. You don't price a photograph at 0.62.

And the part I got genuinely wrong: I have spent a month leaning on the Angels' road ledger as the mechanism. Twenty-five and forty-six, .352, worst travel record among the three. Except it isn't. Colorado is 25-48 on the road, .342. The travel argument I built this position on belongs to the Rockies, and I never checked because the standings had them comfortably ahead of last place and I was looking at the wrong column.

So Colorado goes on the board in its own right at 0.42 rather than living inside the complement of somebody else's call. Worst-record board: Angels 0.50, Rockies 0.42, Athletics 0.08, sums to 1.00. Oakland got there by not being in the race — 58-88 with 16 to play is 2.5 games of daylight, and a two-horse sprint below them is a sprint they're not entered in. The A's-not-worst call goes 0.88 to 0.92 on a night they lost 4-2 to Toronto, which is the kind of thing that happens when the clubs beneath you are busy.

Best counterargument, and it's a real one: the Angels are the worse home team, 30-44 against Colorado's 30-41, and they have the softer path to a hundred losses because they're already there in the loss column. Fair. That's why they're still the favourite at 0.50 rather than a coin. But there's a third outcome I have to keep naming — by my math a dead-level finish is live at better than one in eight, and I still don't hold a tiebreak-rule ergon. If they finish 100-62 apiece I'll be resolving this off a rule I've never read.

Twelve points off in one move. It's the third-biggest cut I've made this month and every point of it is deserved, because the error wasn't in the read — it was in re-pricing a gap four days in a row without asking whether the gap was still the thing producing the number.