About

Pondletter is the public record of a machine that trades Kalshi markets around the clock and reports what happens. Almost all of it runs on paper. Since August 2026 one desk has also traded real money, in deliberately small size, and its complete daily record is sealed by machine every night on the proof page.

The lab runs dozens of strategies at once. Each one is a seat with its own clock and its own money: a rule, running on one kind of market, logged at the price the order book was actually showing and charged the fee Kalshi would actually charge. Every Sunday the board goes up, showing what each seat did this week, what it has done since it was seated, and how much of the result was eaten by fees.

Most of them lose. That's the interesting part.

Why paper, and where it stops

Almost nothing on the board moves real money, and that is a deliberate choice rather than a limitation. Paper trades can run a hundred strategies simultaneously for years, including the ones that turn out to be terrible, which is exactly what you need to find out whether an idea works. A real account can only afford to be wrong a few times.

The exception exists because there is one question paper cannot answer: whether a simulated fill would ever have been a real one. So since 4 August 2026 a live desk has quoted real money, one contract at a time. It runs in pre-registered blocks: a fixed number of orders, a pass bar written down before the first one, and a kill rule that retires the lane if it fails. Size is tested the same way, one block at a time, and a two lot block that lost is on the record next to the one lot blocks that did not. It is marked as live wherever it appears, its results are never blended into the paper figures, and every fill is the exchange's own ledger, not ours.

The tradeoff is honest: simulated fills are not real fills. Every entry price is the ask the market was showing at that moment, and every trade pays the fee Kalshi would actually charge. But a price being displayed is not the same as a fill being achievable, so here is exactly how far the verification goes.

Where the lab records the order book, meaning Bitcoin and esports markets, every entry is checked against the depth that was actually there. That check has already caught and discarded hundreds of trades whose price level had no size behind it. Where the lab does not record the book, which covers challenger tennis, MLB, soccer, basketball and football, the entry price is what the exchange reported at the time, but we cannot prove after the fact that the size was there. Each week's issue prints what share of that week's trades were fill-verified, so you never have to guess which kind you're reading.

Three things the simulation does not model, stated plainly: the delay between deciding to trade and an order reaching the exchange, the price slipping while it travels, and adverse selection, meaning the tendency for your order to get filled precisely when someone better informed wants the other side. That last one matters most for the market-making seats, whose simulated results should be read as an upper bound rather than an estimate.

Settlement is not our judgment. Every win and loss comes from Kalshi's own resolution of the market.

Why you can check us

Most trading publications ask you to believe a track record. This one is built to be verified instead.

Every night the lab hashes its raw market recordings and trade ledgers into an append-only chain, where each entry commits to the one before it. Rewriting any past trade would change every subsequent link. The current head of that chain is printed at the bottom of every issue, so once an issue reaches your inbox, that week's history is frozen somewhere we can't reach. Compare this week's head against last week's whenever you like.

Members get the full board. Desk members get the per-trade files behind it, back to the lab's first week, so the numbers can be recomputed from scratch rather than taken on faith.

What gets published

The weekly board, every Sunday. Every seat, honestly, including the ones having a terrible month.

The live desk's block verdicts, pass or fail, on the day they land.

Studies. One question, answered properly. Does following large trades work? Are weather markets beatable? How much of a strategy's loss is fees rather than bad picks? Free readers get the verdict; members get the evidence.

The field guide. A standing reference on Kalshi's markets: where the liquidity actually is, what spreads really cost, which market types are structurally traps.

The Seats. What every strategy on the board actually does, with its real triggers and thresholds.

What I am holding while I write this

I trade these markets with my own money. That is worth saying plainly, because a publication about a market, written by someone active in it, should tell you where its author stands.

Three things are kept apart, and it is worth being precise about which is which. The paper lab is published in full. The lab's live desk is published in full, as results and as a sealed daily ledger. My own trading is published nowhere: its positions are not disclosed, and no figure from it appears in any issue, on any board, or in any file. Where all of this overlaps is that the lab tests ideas I care about for obvious reasons, and occasionally a seat is trading the same contract I am.

You should read the board knowing that. It is also why the market-making seats are reported as results rather than as a recipe. Entry filters that can be explained from public mathematics do get published here, and have been. Quote placement, skew, inventory limits and pull triggers do not, because on markets this thin those are the parts that compete for the same spread, and publishing them would degrade the edge for everyone running one, including me.

What this is not

This is not advice, and nothing here is a recommendation to trade. No customer funds are managed, no trades are placed for anyone, and no personalized guidance is offered. It is a research publication about a market, written by someone running experiments on it.

Prediction markets can lose you money quickly. The lab loses money on paper most weeks, and says so in public, which is rather the point.

Who writes it

The Heron. A heron stands still at the edge of the water for a long time before it does anything.

Questions, corrections, and theories worth testing: reply to any issue.