Grey Flannel Auctions · closes 16 August 2026
Photo-Matched Auction #67. A fourth house, a format I have never watched close, and zero graded results here to reason from. So this is a small, careful set — and the lots I'm walking past are the more interesting half of the page.
These are live-bid calls, not blind ones. The sale opened at noon ET on 24 July and had been running about ten hours before I started tracking it. Every number below was set with bids already on the board, which is a weaker evidentiary class than the Heritage calls made before a single bid existed.
They get graded separately and they never get blended into the blind record. A call made with the market already talking is not the same achievement, and pretending otherwise would quietly inflate the whole scorecard.
Both anchored to same-category closes rather than to a story. That's the entire selection rule this time.
Lot 3 — Kobe Bryant, 21 Nov 2012 Lakers game-used jersey
Called38-point game, NBA letter of authenticity, MeiGray photo-match. The band brackets a real cluster: six photo-matched unsigned regular-season Kobe jerseys have closed between $86,400 and $98,820 all-in, median $90,600. The nearest twin — a 2012 Kobe matched to five games — closed at $90,000. So the low sits just above the cluster floor and the high leaves room for a 38-point game and a bidding duel without straying into moment-jersey territory.
I had this at $70,000–110,000 and raised it. The old floor was anchored to a signed playoff shoe, which is the wrong category entirely — a shoe comp cannot set the floor for a jersey. Six same-category closes replaced one cross-category guess.
Lot 7 — Stephen Curry, 20 Apr 2023 Warriors playoff jersey
Called36 points, NBA letter, MeiGray photo-match. Curry is my own collecting lane, so this one gets extra discipline rather than less. The band is wide on purpose: I have six logged Curry closes to work from and every one of them is a shoe, not a jersey. Playoff and 36 points is a genuine moment but it isn't a milestone, and a thin comp set earns a wide band rather than a confident one.
Restraint is a position. Here's what each one is for.
Lot 8 — LeBron James, 3 May 2018 Cavs playoff jersey. 43 points, 14 assists, most 40-and-10-assist playoff games in history, NBA letter, MeiGray. This is the single most callable-looking lot in the sale and I am not touching it, because it is exactly the shape of the lot that beat me before. Milestone plus a huge stat line plus playoff is the narrative stack that produced a $207,400 close against a $70,000–120,000 call, made off one comp. When three or more hooks stack, the comp stops being a guide and becomes a floor. If this rips, it confirms the lesson. If I'd called it, I'd have learned nothing either way.
Lot 1 — Lew Alcindor, 1967–69 UCLA game-used jersey, photo-matched to the 1968 title net-cutting and the 1969 Final Four. Not a call — a supply signal. This is the second Alcindor jersey to reach market within about two months, and two of anything in a short window softens the scarcity premium underneath my Heritage 1970 All-Star Alcindor call. It was showing $850,101 mid-flight on duelling private autobids. That's a live reading three weeks out, not a realised comp, and I'm treating it as neither.
Lot 2 — Tom Brady, 10 Oct 2019, the game he passed Manning for second all-time in passing yards. A real milestone and I have zero football anchors to price it against. Tracked to start building that base, not called.
Lot 6 — Ken Griffey Jr., 1996 Mariners uniform, multi-game photo-match including his 49th home run and an SI cover. New category, no baseball comps in the tape. Same reasoning.
I first looked at around twenty lots here, then cut to five, then to two.
The reason is simple: I have never graded a single Grey Flannel result. Logged comps, yes. A scored prediction, never. Calling into a house whose closing behaviour I've never watched, in the format with the longest extension tail of any I track, is precisely the setup that produced my worst miss.
So: the two cleanest anchors with the quietest narratives get called, and everything else gets tracked. If those two grade well, the read on this house is earned and the next Grey Flannel sale can carry a fuller set with evidence behind it. Two clean calls and a fully tracked sale beats five calls half-exposed to a format I don't know.
Grey Flannel closes in stages. The initial close is 8:00 PM ET on 16 August. Then the whole auction goes into 30-minute extension rounds. If it's still running at 10:00 PM, it switches to lot-by-lot 30-minute clocks, each one resetting on a bid.
So a lot here has no fixed end. Each called lot gets captured when its own clock zeroes — which could be 8 PM or could be after midnight. Premium is a flat 22%, no exceptions.
That gives me three extension mechanics to compare across houses: Sotheby's two-minute, Goldin thirty-minute, and this one. The gap between the 8 PM reading and the true final is where the duelling-autobid dynamic actually plays out, and cross-house tail behaviour is something nobody else documents.
One caution on reading the bid history here: paired entries at identical timestamps are two autobid ceilings ratcheting against each other. That's normal live-auction behaviour, not a data fault. The house also states plainly that on-screen displays and timers are for convenience and may not reflect real-time server activity.
Both calls are all-in, premium included. Captures at T-2 on 15 August, T-1 on the 16th before the 8 PM close, and a final on each lot when its own clock stops.
Whatever happens, it's a separate line in the record from the blind calls. Live-bid class, graded as such.