NFT Sales Held Flat on 85% Fewer Buyers — the Question This Week Is the Count, Not the Dollars

The week ending Oct 10, 2026 recorded $39.85M in organic NFT sales — up 0.90% on the capture's own base — on 6.92% fewer transactions from 85.33% fewer buyer addresses. The dollars held; the wallet count broke.

This report decomposes the flat print into its concentration layers, extends the breadth series to five captures — where the Oct 10 print acquires the Sep 12 anomaly as a precedent — scores last week's calls, and sets the one checkable test that separates a market that consolidated from a dashboard that broke.

1. Background: The Week the Wallet Count Broke

The week ending Oct 10, 2026 held the value column steady and broke the participation column. CryptoSlam's seven-day dashboard recorded $39.85M in organic sales — up 0.90% against the prior-week base its own comparison implies (about $39.50M, a figure it does not print) and down just 2.52% against the $40.88M in our own Oct 3 capture. But buyer addresses fell 85.33% to 30,332, seller addresses fell 84.45% to 30,598, and transactions fell 6.92% to 778,055. On our base, the market shed $1.03M of value and 176,456 buyer addresses in the same seven days — a 0.5% dip in dollars against a collapse in participants.

My position up front, because this report exists to test a claim, not to narrate a red candle. Last week I wrote that value had reverted while participation had not, and left the question of which column tells the truth to this capture. This week muddies the question instead of settling it. The value column is internally consistent and cross-checked: Courtyard, CryptoPunks and Pudgy Penguins all reconcile against their prior prints, the five largest trades sum to $871,411, and the chain table sums to the headline. The address column is the problem: 30,332 buyers against 206,788 is not a decline, it is a discontinuity — and this series has seen one before. The Sep 12 capture printed 41,959 buyers at 21.87 transactions per address, an anomaly we have carried unexplained for four weeks. The Oct 10 capture prints 30,332 buyers at 25.65 transactions per address. The anomaly has a sibling, and two unexplained prints in five captures is a pattern in the measurement, not in the market.

2. Event Timeline

1. Oct 3 — our previous capture records $40.88M in organic sales, 863,295 transactions and 206,788 buyer addresses, with all five blue-chip floors lower; Panini America's +557.53% week and GONDI's 3M USDC Punk sale carry the tape.
2. Oct 5 — NFT Ltd. (NYSE American: MI), a Hong Kong-listed art and collectibles platform, closes near $7.00, up about 670% on roughly 165.7M shares with no company-specific catalyst — the day's standout among exchange-listed NFT names.
3. Oct 6 — Igloo Inc., Pudgy Penguins' parent, announces that its Abstract Ethereum layer-2 will shut down on Dec 15, 2026, after 18 months and tens of millions of dollars in losses; the same day, MI closes at $1.23, down 82.43%, and Pudgy Penguins' Token2049 Huddle series begins in Singapore.
4. Oct 8 — NFT Ltd. prices a $1.0M registered direct offering — 1,000,000 units at $1.00 with five-year warrants — closing Oct 9, with proceeds for its NFT platform and an AI robotics project.
5. Oct 9 — NeoDrop's briefing, cutoff 09:00 ET, records ether down 6.95% and four of five blue-chip floors lower, with Doodles the only holdout; Pudgy Penguins trades 127.5% more ETH on 304 sales, and every repair test set a week earlier fails.
6. Oct 10 — the capture records $39.85M in organic sales on 778,055 transactions, with 30,332 buyer and 30,598 seller addresses. The week's largest listed trade clears on Bitcoin — $8888 BRC-20 at $327,125.81 — the first Bitcoin-led top sale since the Sep 7–13 BRC-20 sweep.

3. Data Support

3.1 What the flat print is made of

On our own base the week lost only $1.03M of organic sales — but the flat headline hides a rotation, not a steady state. Ethereum added $0.52M (+3.0% on our base) and Polygon $0.41M (+5.0%), with Courtyard's +$0.37M accounting for 90% of Polygon's chain-level gain — the same one-collection arithmetic as last week, now on a smaller canvas. The declines came from the smaller chains: Bitcoin −$0.28M, BNB Chain −$0.36M, Solana −$0.38M. The five largest listed trades total $871,411 — 2.2% of global volume, down from $996,000 and 2.4% — and for the first time since the week of Sep 21–27 the largest print is not on Ethereum.

The table's second row is the week. A $1.03M value move against a 176,456-address participation move is a 171:1 ratio, and no plausible market behaviour produces it: the average buyer would have to have increased weekly spending from $198 to $1,314 while transacting six times as often. Bots, treasury wallets and consolidated market-makers can produce high transactions-per-address — but a sixfold jump in one week, in a market where Courtyard's 19,547 buyers cleared 133,068 retail-sized trades, asks the address column to describe a market that the value column says does not exist.

3.2 The breadth test, across five captures

The cleanest way to see the discontinuity is to hold the frame fixed: all five columns below come from our own captures, so no comparison-base problem applies.

Three features stand out. First, the participation column is no longer monotonic: 41,959 → 114,977 → 160,565 → 206,788 → 30,332. The Oct 10 print is the lowest of the five captures and 85% below the one before it. Second, transactions per buyer address printed 25.65 — the highest of the series, above the Sep 12 anomaly's 21.87 — while value per buyer address hit $1,314, four times the Sep 12 print's $1,115 and six times the Oct 3 figure. Third, value per transaction moved the way a normal week moves: $50.98 → $46.44 → $67.24 → $47.35 → $51.22, inside the range the series has traced since September. The value column behaves; the address columns do not. And the one prior print that behaved this way has never been explained by any source we hold.

3.3 Last week's calls, scored

Five claims from the Oct 5 edition were checkable against this capture and briefing. The mean-reversion conditions largely held, the rotation conditions swept, the Base wash field crossed its documented-static threshold, and the Pudgy distribution call confirmed — every score in the constructive direction.

The two confirmed calls deserve plain language. The Pudgy distribution reading is no longer a signature: volume rose 83.5% two weeks ago and 127.5% this week — 945.69 ETH on 304 sales — while the floor fell 12.90% then and 4.88% now, and the listed count grew eleven to 274. A bid that tracks the floor down while 304 tickets clear is a market taking supply off impatient holders. And the Base wash field has now printed $4.80M for seven published weeks while its organic denominator moved every week and its own change column implies $2.67M — the field is documented as static, and every ratio built on it carries an asterisk until the source explains.

3.4 The Courtyard and Pudgy tapes, and where the dollars sat

Courtyard held the top of the collection table at $7.68M across 133,068 transactions — $58 per trade, 19.3% of global organic volume and 89.1% of Polygon's own organic total, both shares higher than last week's 17.9% and 89%. Its +$0.37M increase is 90% of Polygon's chain-level gain, and its buyer count fell only 7.50% — the gentlest participation decline on the board against a market that lost 85%. Pudgy Penguins is the week's demand story: $1.50M, up 125.66%, on 94 buyer addresses and 123 seller addresses — about 16 transactions per buyer at the collection level, a whale-and-workshop shape that arrived in the same week its parent announced Abstract's shutdown and its token fell roughly 20%. CryptoPunks rebounded to $2.45M on 26 trades averaging $94,337 — 1.21× a floor that fell to 31.39 ETH — while Credits, Panini America and $ATMC BRC-20 all left the top seven. Three of last week's five stories did not survive a week.

3.5 Blue chips: four down against a falling unit of account

Against a 6.95% ETH fall, four of five tracked floors slipped in ETH terms and Doodles held flat at 0.2621 ETH — the cohort's first week with a holdout since the week ending Sep 25. In USD every floor is 9% to 14% lower.

The tape splits the cohort three ways again. Pudgy Penguins printed the distribution case for a second week: 945.69 ETH of volume — 2.3× the prior week — into a 4.88% floor fall. BAYC printed the erosion case: volume down 42.3%, offer down from 97.0% to 93.9% of floor — above the 90% line, but the repair is half-erased. CryptoPunks printed the float case: 906.62 ETH of volume on 25 sales — the largest weekly turnover in the cohort — with the floor breaking the 33 ETH line at 31.39 on a listed float of 1,101, and no published top offer to measure the bid against. Azuki's 20 ETH sale against a 0.605 ETH floor — 33× — is again the only exotic print. The cohort traded 2,234.15 ETH, up 34.2%, with the Punks-and-BAYC share of it falling to 54.7% — and every one of the five repair tests the briefing set a week ago failed.

3.6 The wash overlay, and the field that stayed frozen

Polygon's ratio eased to 2.21× — $19.03M of wash against $8.62M organic — falling back under the 2.5× line after last week's 2.80× series high, with seven of eight published weeks still at or above 2.21×. Base's wash exceeded its organic volume for a seventh published print at 2.00×, and the numerator is the same story: $4.80M, identical for seven weeks, against a change column that implies $2.67M and matches neither print. Ethereum's wash share rose to 6.5% ($1.15M on $17.60M), Bitcoin printed $80,297 (2.3% of organic), Immutable entered the table at $47 — a rounding-level 0.002% — and BNB Chain, which returned last week at $1,687, went silent again. One more divergence to carry: CryptoSlam's own daily table for Oct 4–10 sums to $33.42M against the $39.85M weekly dashboard — a $6.4M internal gap we disclose rather than reconcile.

4. Competing Readings

Where I land: the measurement reading owns the address columns and the concentration reading owns the value column, and the honest position is that this capture cannot separate wallet collapse from dashboard collapse — the checkable test is the Oct 17 capture, which decides whether 30,332 is a level or an artefact. What this week does settle: the value column's flat print is real on either reading, and it is narrower than the headline — Courtyard at 19.3% of global volume, one Bitcoin print at 9.2% of a chain's week, and a top-five trade list worth less than one GONDI settlement from the week before. Whether participation re-based or the counter broke, the market that remains is measurably more concentrated than the one the series has been describing since August.

5. Implication Projection

Conditionals, not forecasts. Each is checkable against the Oct 17 capture.

1. If the measurement reading holds: buyer addresses snap back above 150,000 and transactions per buyer return under 10, pairing the Sep 12 and Oct 10 prints as capture artefacts in the same column. The series would then need a rule: address-column discontinuities get flagged, not narrated.
2. If the concentration reading holds: buyer addresses stay five-figure, value per buyer stays above $500, and Courtyard plus one whale print carry the tape. A second five-figure week would make 30,332 a re-based participation level — and the series' address history would need a structural break marked, not smoothed.
3. If the Pudgy distribution pattern extends: volume eases toward the 59.31 ETH daily pace with the listed count falling under 263 — the briefing's own done-distribution markers — while the Dec 15 Abstract deadline approaches. Listings still climbing on 135-plus ETH a day would say the pattern is not done.
4. If the floor slide extends: CryptoPunks below 30 ETH with more than 1,100 items listed would say the float caps any bounce, and BAYC's offer under 90% of floor would mark last week's 97% print as one bidder who has left. Doodles' flat floor meeting a rising listed count is the quiet test of whether a level held on no volume is real.
5. For the Base wash field: an eighth identical $4.80M print extends the documented-static record to two months; any change in the level while the change column stays contradictory would confirm two divergent series inside one source. Either outcome resolves the asterisk.
6. For anyone quoting the headline: "NFT sales held flat" is true and incomplete on its own — the flat print rests on one capture whose address columns printed a discontinuity, whose daily and weekly tables disagree by $6.4M, and whose value concentration rose on every measure this series tracks. The honest sentence carries the composition, not just the change.

6. Open Questions

• What happened to 176,456 buyer addresses in seven days? Neither the capture nor any source we hold offers a mechanism, and the Sep 12 precedent suggests these columns have produced unexplained discontinuities before — the Oct 17 print is the discriminator.
• Why do CryptoSlam's daily and weekly tables disagree by $6.4M in the same capture? The daily sum for Oct 4–10 is $33.42M against the weekly $39.85M — either a window difference or a revision cycle, and the source publishes no note that resolves it.
• Is Pudgy Penguins' week brand demand or deadline rotation? Sales rose 126%, the token fell roughly 20%, the floor fell 4.88% on 127.5% more volume, and the parent's chain shuts Dec 15 — three signals in three directions, one week apart.
• Was the $8888 BRC-20 print a market or a moment? The collection's whole week was $327,134 across two transactions, and the larger of them was 9.2% of Bitcoin's weekly total — a whale print wearing a leaderboard crown.
• Does Courtyard pass 20% of global organic volume? It took 19.3% of a shrunken tape at $58 per trade; its buyer count is the only five-figure participation base left on the board.
• Will the transaction column reconcile at all? The capture's implied prior has now run 3.2% below our print after 3.6% last week and 3.6% the week before — a fourth miss makes the offset systematic rather than a window artefact.