lectr
What is lectrNo. 20260912
lectr

Every lot arrives with a guess. We score it against 1,102,760 results.

lectr is the intelligence desk for the auction market. It has read every settled lot across 18 houses, and prices what comes next against the sales that resemble it.

See tonight’s calls How it works

Public and free. No account needed.

1,129,691
Lots under tracking
live and settled, one graph
18
Auction houses
named, with their coverage, below
38,793
Replayed calls
scored against what happened next
The corpus

Depth is the moat. 35 years of results, one schema.

Comparable-sales pricing is only as good as the pool behind it, and auction results are scattered across houses that publish in different shapes, purge lots when a sale closes, and rarely keep a machine-readable archive. lectr holds all of it in one schema — every lot normalised to a dated FX rate, a declared price basis, and a structured identity, so a 1994 Geneva watch sale and a lot closing tonight are the same kind of object.

Christie's198984,265
Lelands200023,231
Sotheby's200058,575
Wright20007,660
RR Auction1996241,076
Bonhams200316,202
REA2004179,803
Huggins & Scott200576,095
Memory Lane200711,235
LAMA20011,070
Love of the Game201311,791
Phillips20139,627
MLB Auctions2014995
Rago20161,756
SCP20218,714
Goldin2022369,384
NFL Auction20251,281
Sports657,78262.0%
Pop Culture240,42922.7%
Watches65,8246.2%
Art59,9515.6%
Science27,7872.6%
Design9,4480.9%

Settled records per house by sale year, then by market. The faint segment is coverage that exists but is thin — Christie’s has a single lot dated 1989 against 1,331 in 1991 — so early depth is shown as early depth, not rounded up. Nothing here is licensed: every row was crawled, normalised and reconciled lot by lot. What that reconciliation looks like

The value engine

What a lot is worth, argued from the sales that resemble it.

The engine does not forecast taste. It answers a narrower question with evidence: given everything that has actually sold, where should this lot clear — and does the house’s estimate agree?

1
Resolve the object
Every lot is parsed into a structured identity — maker, form, model line, reference, edition, dimensions, year, materials — not just a title string. Two lots match on what they are, never on how a cataloguer chose to describe them.
2
Build the comparable pool
Candidates are scored on title similarity and structural agreement, then put through hard gates: a sofa never comps a chair, a Daytona never comps a Datejust. Pools that don’t clear the bar are discarded rather than loosened.
3
Price it, with dispersion
A recency-weighted median over the surviving pool, with a dispersion guard that widens or withdraws the read when comparable sales disagree with each other.
4
Call the direction, and rate the confidence
The output is a directional read — trading below or above where comparables clear — carried with a confidence tier calibrated against 87,487 scored observations, and a band showing how tightly reads at that tier have historically landed.

Validated by temporal holdout: every call is made using only sales dated strictly before the lot in question, so the record below is what the engine would have said at the time — not what it can explain in hindsight. The method in full

The record

The whole thesis, replayed against history.

38,793 flagged calls and 28,975 unflagged controls, each scored on what the lot actually did next. Two bases are published side by side, because they answer different questions: all-in is what a buyer paid, including the house’s premium; at hammer strips the premium out for a like-for-like comparison against an estimate that never included it.

Unflagged control25.5% below estimateFlagged by lectr16.9% below estimate
Where every replayed lot landed against its estimate, as a share of its own arm — the two arms differ in size, so raw counts would draw flagged larger in every bin including the losing ones. Flagged sits further right, and is below the estimate less often than the control: the edge is not bought by taking more risk.
Median result against estimate by sale year, 2000–2026. The flagged line clears the control in every one of the 27 years on record, through two crashes and a boom — the edge is not an artifact of one regime.
The record: unflagged control versus lectr-flagged, same replay, same period.
MeasureControlFlaggedEdge
Median result vs estimate · all-inwhat the buyer paid, premium included+16%+41%+25 pts
Median result vs estimate · at hammerlike-for-like against a hammer-basis estimate−7%+13%+20 pts
Cleared the high estimateshare of lots that beat the top of the range46%65%+19 pts
Failed to selllower is better — the flag does not chase no-sales4.1%3.4%+0.7 pts

Bought-in lots are scored as outcomes rather than dropped, so a flag on something that then failed to sell counts against the record. See the full record

+25points

is what separates a lot lectr flagged from one it didn’t, replayed against every sale that came after.

Flagged lots also failed to sell less often than unflagged ones. The edge is not bought with risk.

The proof

We said what it was worth. The room paid less.

The record above is an aggregate. This is what it looks like as individual lots: the engine priced each object from comparable sold evidence, and the hammer came in under that number. Two of the six below carried no house estimate at all — lectr’s figure was the only valuation in existence when the hammer fell.

sports · Sotheby's · 2026-02-10high confidence
Chet holmgren oklahoma city thunder 2025-2026 game worn city edition jersey
sold for$4,826lectr value$10,795
55.3% under
from 18 comparable sales · 8 shown
SoldLotPrice
2026-01-27Chet holmgren ‘christmas day’ oklahoma city thunder 2025-2026 game worn association ed…Sotheby's$10,160
2025-08-20Chet holmgren oklahoma city thunder 2024-2025 game worn association edition jersey | d…Sotheby's$10,795
2025-07-22Chet holmgren oklahoma city thunder 2025 nba finals game worn icon edition jersey | ga…Sotheby's$40,640
2025-03-25Chet holmgren oklahoma city thunder 2024-2025 game worn statement edition jersey | dou…Sotheby's$5,334
2025-02-07Chet holmgren oklahoma city thunder 2024-2025 game worn statement edition jerseySotheby's$11,400
2025-01-22Chet holmgren oklahoma city thunder 2024-2025 game worn statement edition jerseySotheby's$6,000
2024-09-20Chet holmgren oklahoma city thunder 2023-2024 game worn icon edition jerseySotheby's$14,400
2024-09-20Chet holmgren oklahoma city thunder 2023-2024 game worn statement edition jerseySotheby's$14,400

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 18, not over the 8 shown.

The house published no estimate — ours was the only valuation.
design · Sotheby's · 2026-06-15high confidence
"standard" chair
sold for$10,880lectr value$24,320
55.3% under
from 46 comparable sales · 8 shown
SoldLotPrice
2026-06-10“standard” chairSotheby's$24,320
2026-06-10"standard" chairSotheby's$20,480
2026-06-10"standard" chairSotheby's$20,480
2026-06-10"standard" chairSotheby's$35,840
2026-06-10"standard" chairSotheby's$32,000
2026-06-10"standard" chairSotheby's$28,160
2026-06-10"standard" chairSotheby's$23,040
2026-06-10"standard" chairSotheby's$25,600

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 46, not over the 8 shown.

House est $8,000–$12,000.
design · Wright · 2026-03-26high confidence
Conoid dining table
sold for$28,160lectr value$47,880
41.2% under
from 80 comparable sales · 8 shown
SoldLotPrice
2026-01-14Conoid dining tableRago$48,260
2025-05-15Conoid dining tableRago$40,640
2025-01-16Conoid dining tableRago$38,100
2024-11-21Conoid dining tableWright$47,880
2024-09-19Conoid dining tableRago$37,800
2024-05-16Conoid dining tableRago$52,920
2023-09-21Conoid dining tableRago$35,280
2023-06-07"conoid" dining tableSotheby's$53,340

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 80, not over the 8 shown.

House est $25,000–$35,000.
art · Bonhams · 2026-03-04high confidence
After the Party
sold for$21,760lectr value$35,280
38.3% under
from 64 comparable sales · 8 shown
SoldLotPrice
2025-12-09After the PartyBonhams$28,288
2025-10-07After the PartyBonhams$28,160
2025-06-25After the PartyBonhams$38,272
2024-06-04After the PartyRago$35,280
2023-10-12After the PartyRago$40,320
2023-09-12After the PartySotheby's$29,261
2023-09-07After the PartyWright$44,100
2023-05-09After the PartyBonhams$38,175

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 64, not over the 8 shown.

House est $20,000–$30,000.
sports · Goldin · 2026-07-26medium confidence
1958 Alifabolaget #635 Pele Rookie Card - PSA EX-MT 6
sold for$23,180lectr value$43,924
47.2% under
from 36 comparable sales · 8 shown
SoldLotPrice
2026-06-141958 Alifabolaget #635 Pele Rookie Card - PSA EX-MT 6Goldin$40,260
2026-06-141958 Alifabolaget #635 Pele Rookie Card - PSA MINT 9 - Pop 6Goldin$976,000
2026-06-141958 Alifabolaget #635 Pele Rookie Card - PSA VG-EX 4Goldin$21,350
2025-11-301958 Alifabolaget #635 Pele Rookie Card - PSA VG-EX 4Goldin$20,768
2025-09-071958 Alifabolaget #635 Pele Rookie Card - PSA EX 5Goldin$45,140
2025-08-101958 Alifabolaget #635 Pele Rookie Card - PSA EX-MT 6Goldin$48,800
2025-06-291958 Alifabolaget #635 Pele Rookie Card - PSA EX-MT 6Goldin$48,190
2025-04-271958 Alifabolaget #635 Pele Rookie Card - PSA EX 5Goldin$43,924

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 36, not over the 8 shown.

The house published no estimate — ours was the only valuation.
watches · Sotheby's · 2025-03-31high confidence
Reference 5402st jumbo royal oak 'a-series' | a stainless steel automatic …
sold for$40,640lectr value$72,000
43.6% under
from 18 comparable sales · 8 shown
SoldLotPrice
2024-06-05Reference 5402st jumbo royal oak 'a-series' | a stainless steel automatic wristwatch w…Sotheby's$72,000
2024-06-05Reference 5402st jumbo royal oak 'non-series' | a stainless steel automatic wristwatch…Sotheby's$60,000
2023-11-07Royal Oak, Reference 5402ST | A stainless steel wristwatch with date and bracelet | Ci…Sotheby's$42,405
2023-06-24Royal Oak 'A-Series Jumbo', Reference 5402ST | A stainless steel wristwatch with date…Sotheby's$54,966
2022-11-09Royal oak 'a-series jumbo', reference 5402st a stainless steel wristwatch with date an…Sotheby's$85,749
2022-05-10'b-series jumbo' royal oak, ref. 5402st stainless steel wristwatch with date and brace…Sotheby's$112,134
2022-05-10'd-series jumbo' royal oak, ref. 5402st stainless steel wristwatch with date and brace…Sotheby's$85,749
2021-12-09Reference 5402st royal oak jumbo 'non-series', a stainless steel automatic wristwatch…Sotheby's$113,400

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 18, not over the 8 shown.

House est $30,000–$60,000.

How these were chosen. Of 16,628 lots sold since 2024 above a $3,000 estimate, the engine priced 4,151 from at least eight comparable sold lots at non-low confidence — and 738 of those cleared 30% or more below the number. That sweep is exhaustive, not a sample; four of the six above are drawn from it, as is the lot below.

The no-estimate pair come from Goldin and Sotheby’s NBA auctions, which publish none. That population has not been swept exhaustively, so no rate is claimed for it.

International Youth Year · Wright · 2026-05-19 — lectr priced it at $12,065 from 22 comparable sales. It sold for $7,680.

The 22 sales it was priced against · 8 shown
SoldLotPrice
2026-04-13International Youth YearBonhams$15,360
2025-04-03International Youth YearWright$12,065
2025-01-08International Youth YearWright$9,525
2024-09-04International Youth YearWright$9,450
2023-10-03International Youth YearSotheby's$12,065
2023-03-29International Youth YearRago$15,120
2023-03-28International Youth YearBonhams$10,838
2022-10-12International Youth YearRago$11,875

The pool this lot was priced against, most recent first — every one settled before it sold. The value is a recency-weighted median over the full 22, not over the 8 shown.

Restraint

The number we are proudest of is how often it says nothing.

lectr runs an explicit ladder: a confidence-interval index where the data resolves the sign, a measured demand read where coverage allows, and below that, no movement number at all.

4 of 54
Makers clear the 95% bar
the rest publish no index — the interval doesn't resolve the sign
38 of 100
Sub-markets abstain
tracked and searchable, but carrying no movement claim

Every published figure names its method and its sample size — and when one drifts, it is restated in public rather than quietly patched. The corrections register

What this does not cover

lectr reads auction results only — public sales with a published price, at the 18 houses named above. Private treaty, dealer inventory, fixed-price and buy-now listings are deliberately out: a price nobody bid against is not a comparable, and mixing them in would quietly inflate every figure on this page. Fairs, regional and online-only houses are not yet read, so a maker whose market lives largely outside these rooms will show thinner coverage here than it has in the world. Where that is true, the engine abstains rather than extrapolating.

The graph

One lot, linked to everything that explains it.

A price is not an answer on its own. Every lot resolves into a chain you can walk — and every step of it is a page, not a footnote. The links below open real ones.

  1. The lotwhat it was expected to make, what it made, and the call that preceded it
  2. Its comparablesthe exact settled sales the call was argued from, each one clickable
  3. Its sub-marketcards by era and sport, watch families, art kinds, design materials
  4. Its makera quality-controlled index, published only where the interval resolves
  5. Its referenceone of 614 dossiers, each with the full yearly series for that model line

Tracked lots keep permanent addresses. A lot the house purges the moment its sale ends stays resolvable here — with its estimate, its result, and the call that preceded it — which is what makes the archive usable as evidence rather than nostalgia.

What it changes

For the person holding the paddle.

A price is only useful if it changes what someone does. Four people ask the same question — what does this actually clear at, and how sure can you be — and get four different days out of the answer.

The specialistsetting a reserve
todaymemory, three catalogues, and a feel for the roomwith lectrthe comparable sales on the page — dated, with their spread — before the consignment conversation
The collectordeciding what to bid
todaythe estimate, the atmosphere, and how badly you want itwith lectrwhere comparable examples actually cleared, which is the difference between conviction and nerve
The lender or insurermarking a book
todaya valuation letter that cannot be re-derivedwith lectra number defensible line by line — method named, sample size shown, and an abstention where the data will not carry a figure
The sellerchoosing where to consign
todaywhichever house asked firstwith lectrwhat the object is worth before someone else says, and which house has historically cleared that kind of lot highest
Tonight’s book

8,524 lots are on the book. lectr has something to say about 575.

7,949 no call — the comparable pool doesn’t clear the bar575 called · 6.7% of the book · as of Sep 12, 2026
326trading below where comparables clear
249trading above it
85at the high-confidence tier

Both directions are published: a lot trading above where its comparables clear is as much of a call as one trading below.

Stop bidding against a guess.

The whole market read is public and free to inspect. Nothing on this page is a projection — it is what 1,102,760 settled results already say.

See today’s calls Open the research deskRead the notes

Figures on this page are read from the live corpus at build time, dated 2026-09-12. They change when the market does. Each quarter is written up market by market — watches, art, design, sports and science.