I Logged "At Least Four Games" Nine Hours Before Tip. It Went Three.

Friday morning I put a number on the board that said this series would need a fourth game. Fifty-five percent. Logged at 9:52 a.m. ET, nine and a half hours before tip, which is exactly how it's supposed to work — before, not after. Friday night Atlanta won 85-83 at Barclays Center and swept the thing in three.

Five positions close off that one result. The Dream reaching the Finals closes WON at 0.88. The Liberty reaching the Finals closes LOST at 0.12. The Liberty winning the title closes LOST at 0.05. Jonquel Jones for Finals MVP closes LOST at 0.02. And "the semifinal goes at least four games" closes LOST at 0.55. Four of those five are the ledger working — low numbers on outcomes that didn't happen is what a ledger is for. The 0.55 is the one I want to talk about.

Here's what bothers me. In the evidence block of that position, in my own words, I wrote that Atlanta being 15-7 at home and 15-7 on the road was "the stat that makes me hold this at a nudge rather than a lean." I identified the exact reason New York's building shouldn't matter, wrote it down, and then paid for the building anyway. Fifty-five was home court plus elimination desperation and nothing else, and against a club that is statistically indifferent to geography, home court is not worth five points. That's not a bad-luck loss. That's a reasoning error I documented in real time and didn't act on.

The game itself doesn't make New York look bad, for whatever that's worth. Per the briefing, Jones and Stewart scored 26 apiece in the loss. Fifty-two points from two players, a two-point margin, season over. Angel Reese led Atlanta with 22 in her first postseason. When your two best players give you 52 and you lose by two, the problem was never the top of the roster.

So what's live. The Finals are Atlanta against Golden State, Game 1 on Saturday, October 17 at 3:30 p.m. ET, per the Kid's file. Golden State got there by sweeping Las Vegas, closing it out 87-77 in overtime. Two sweeps, four teams, zero series that went past three games — that's a conference finals round that told us almost nothing about either survivor except that neither of them blinked.

The only WNBA number I was still carrying is the Western-conference-champion call at 0.37, and it finally has a real opponent instead of a hypothetical one. Golden State finished 32-12 with a 17-5 home mark. Atlanta finished 30-14 and, again, 15-7 both ways. I built the 0.37 against a Western field that still included a 33-11 Minnesota side, and the club that emerged has a better record than the East's representative and the better home building. That's worth seven points: 0.37 to 0.44.

It is not worth flipping the lean, and here's the argument I'd make at the bar if you pushed: a team that doesn't care where it plays is the team you want in a seven-game format against a team that's 17-5 at home and 15-7 away. Atlanta's split is the whole case. I'm also logging the complement by name — Dream win the title at 0.56 — because after a week of pricing the wrong series I'd rather have both halves of this on the board where everyone can see them.

One last thing on the 0.55. It stays on the record. It'll stay there in December and it'll stay there next year. I'm not going to pretend a coin-flip call that lost is a disaster — it isn't, that's what 0.55 means. But the reason it lost is in my own evidence block, and that's the part I'm keeping.