Launch Tape
Research example / 01

One recorded sell.
Two measures
of impact.

How much did Odyssey’s pool change around its largest recorded sell? We checked 3,001 trade rows and measured both the SOL reserves and the pool’s price.

● Selected historical caseOdyssey · October 10, 2026Public sample · reproducible calculations
Seller received
159.6 SOL
Largest recorded sell in this sample
Pool SOL change
−64.9%
248.0 → 87.1 SOL in adjacent recorded states
Pool price change
−84.5%
Adjacent recorded marks, not a strategy return
Odyssey’s recorded pool SOL rises toward 248 SOL, falls to 87 SOL around the largest recorded sell at 178.735 seconds, and is below 1 SOL at the final recorded trade. The chart ends at 11 minutes 57.6 seconds.
The blue line connects the pool state after each recorded trade. The marked event is the largest sell by SOL received. On a phone, scroll the chart sideways for the full detail.
The answer

The pool got smaller.
The price moved more.

At +178.735 seconds in the sample, a sell recorded 159.584707613 SOL received by the seller. The adjacent pool states were 248.042363686 SOL before and 87.086791985 SOL after: a fall of 64.9%.

The pool’s price mark fell 84.5% across those same two rows. These percentages answer different questions. The first measures the change in SOL reserves; the second measures the change in the pool’s reported price per token.

The reserve change is 160.955571701 SOL, which differs from the seller’s proceeds. This reduced sample does not include every fee and accounting field needed to reconcile the difference.

Odyssey’s pool price mark per one million tokens falls from 4.006 SOL to 0.620 SOL across the adjacent states surrounding the largest recorded sell, a change of minus 84.5 percent.
Orange shows the recorded pool price, multiplied by one million for readability. This is a pool mark. It does not include the costs, size impact or timing of a hypothetical trader’s buy and sell.
Check the arithmetic

Two rows. No hidden model.

Adjacent recorded states surrounding the largest sell
MeasurePrevious rowSell row
Seconds after graduation178.542178.735
Solana slot455293232455293233
Pool SOL248.04236368687.086791985
Pool price, SOL per token0.000004006290.000000620131

Change (%) = (after ÷ before − 1) × 100. Rounded to one decimal in the charts.

What this sample can tell us

We observed 3,001 rows: 682 buys and 2,319 sells. Their recorder timestamps span 14:42:23.320–14:54:20.686 UTC. The first row is at +0.206 seconds and the last at +717.572 seconds.

The source is a 30-minute selection, but it contains no trade row after +11 minutes 57.6 seconds. The charts stop at the last recorded row. We do not infer the later pool state or fill empty minutes with invented trades.

The companion graduates CSV marks this coin as followed for 1,800 seconds with data_gap=0. That metadata reports no flagged gap; the reduced trade sample by itself does not prove complete chain coverage.

This is one selected historical coin. It does not establish a typical outcome, a profitable strategy, wallet ownership, coordination or anyone’s motive.

Reproduce it

The data and the method are yours to check.

Source trade rowsJSONL · 3,001 recorded trades Minute summaryCSV · empty buckets keep missing states blank Reproduce the calculationsPython · standard library for summary and CSV Exact resultsJSON · measurements, source hash and limitations

Download the source trade rows and Python method into the same folder, then run:

python3 research-example-method.py one-coin-odyssey-2026-10-10-1442-1512utc.jsonl --csv minute-summary.csv

The script prints the exact measurements and writes a 30-row minute summary. It makes no network requests. The summary and CSV modes need only Python 3; the optional repository chart build also uses Pillow.

Ordering, source identity and calculation details
  1. Parse every nonblank JSONL line. Sort by Solana slot, retaining the original row order within each slot. The sample omits transaction signatures and instruction indices, so finer within-slot ordering cannot be independently resolved here.
  2. Find the sell row with the greatest sol value. Compare its pool_sol and price with the immediately preceding row. The highlighted pair belongs to consecutive slots.
  3. Use source s for the horizontal axis. These are recorder-relative seconds, not a promise of transaction execution latency. Lines connect supplied states; there is no extrapolation beyond the last row.
  4. For each minute, count buy and sell rows and report the last supplied pool state in that bucket. An empty bucket has zero recorded rows and blank state fields; it does not prove zero chain activity.
  5. Coverage metadata is from the companion graduates CSV. Field definitions are in the sample README.

Source JSONL SHA-256:
9513e2258b8f9b4f8891fbe6f02612ad9cec9a406e9e4268bf240d1d7feeec77

Mint:
Ff41mqYhF7CqcVBCkFifKUnzavpf9qxByHtrcgbFpump

Your question, next

A focused report for US$99.

One agreed question, clear charts, supporting data and a method you can check. Send the question first; we confirm coverage and scope before payment.

Ask about a report →
← How custom research works