State of the Lab — Nov 15, 2026
State of the Lab — week ending Nov 15, 2026
This week the lab ran 20 seats across 3 desks and settled 35,879 trades, nearly all simulated, 44 of them real money. Best seat: Mention markets. Roughest seat: Early leaning favourite.
Real money this week. The last minute bot's second block of 60 real contracts finished on 14 November, and it failed. It won 41 of the 60 and lost 1.72 cents a contract after fees, minus $1.03, against a pass mark of plus 5 cents. The first half lost 9.76 cents a contract and the second half made 6.31, so most of the damage came from a few bad days early in the month. That is two blocks in a row under the bar: the first finished at plus 2.95 cents, this one below zero. The week on its own was its best in a while, 31 real trades and 24 won for plus $1.05, which leaves it at minus $0.60 on 123 contracts since it went live on 14 October, every fill from Kalshi's own records. What the bot does next is this week's decision, and the next issue will say what we chose and why. The cheap side lane, the same signal under 55 cents, has 13 contracts since 1 November and has won 2 of them, minus $2.59. It needs 60 before it is judged.
The Bitcoin hedge, a Kalshi short with the same amount of real Bitcoin so price moves cancel, collected $0.79 in funding this week and $4.96 since 18 October. The whole package stands at plus $0.69, including about $5 of one time setup fees.
Two verdicts came in, both failures, so both free. The last minute rule on hourly Bitcoin and Ether markets reached its verdict date on 11 November with 143 confirmed paper fills at plus 1.62 cents a contract. The bar was plus 5 cents with both halves positive. The first half lost 3.03 cents, the second made 6.19, and a single day made more than the whole profit. It has done better in the days since, which does not change a verdict written down before the first trade. And five of the eight market adapter rules, the switches that only let a trade through in conditions where it has worked before, have now reached their trade counts and failed. The closest was rule 8, last minute only on coins that had a losing week: its kept trades made 2.14 cents a contract against minus 4.05 for the ones it skipped, but the gap was not large enough to rule out luck. Three rules are still collecting.
Two new paper tests started on 12 November. One runs the last minute rule on the daily crypto closes at 10 am and 5 pm Eastern, the same settlement average with fewer, larger markets; it is judged on or after 10 December. The other collects funding on gold and silver: short Kalshi's gold and silver perpetuals and hold the matching ETFs so price moves cancel, judged on 22 January. Three days in, the price moves between the two sides are bigger than the funding, so there is nothing to read yet.
The crypto desk settled 34,844 fifteen minute trades for minus $786.79, after $534.80 of fees. It lost $251.99 before fees too, so the problem this week was the picks as well as the toll.
Published this week: Can AI beat Kalshi?, the full record of 274 trades picked by an AI with web search, and the written study to go with it. On Tuesday, the three checks we run before buying a fifteen minute Bitcoin contract, from our real fills. Next in line is a viewer's idea about order books just before the close; we will write the rule down before we look. We are still hunting, and every one of these results narrows the search.
Members watch all of this move in the control room, rebuilt every hour, and run the strategies themselves in Ping, live on paper. Your first month is $5.
| Desk | Settled | Net | ¢/trade | Fees |
|---|---|---|---|---|
| live money | 44 | −$0.32 | −0.72¢ | $0.44 |
| Paper strategies | 991 | −$22.24 | −2.24¢ | n/a |
| Crypto 15-min rules | 34,844 | −$786.79 | −2.26¢ | $534.80 |
| all desks | 35,879 | −$809.35 | −2.26¢ | $535.24 |
Each desk name links to The Seats, which gives every strategy's real trigger and threshold plus the Kalshi tickers it trades, enough to run any of them yourself. The one exception is market making: which markets it quotes is listed, how it quotes them is not, and the About page explains why.
All 188 seats behind these lines, with win rates, fee drag, all-time results and the per-trade CSV, are in the members edition. 13 seats retired to date; their records stay on the board.
The market-making seats are simulated fills, and a simulation cannot model adverse selection: the tendency of a resting order to be taken exactly when someone better informed wants the other side. Where this lab has run the same strategy with real money, the live result has come in materially below the paper one. Read the maker desk as an upper bound on what the idea could earn, not as what it would have earned. The live desk below it is that check running in public: same ideas, one contract at a time, settled by the exchange.
Fill verification: 35,918 of 35,918 trades (100%) were checked against real order-book depth.
The rest are entered at the ask the exchange reported, but the lab does not record the book for every market. Challenger tennis, MLB, soccer, basketball and football have no depth archive, so those fills cannot be re-checked after the fact. Trades whose price level had no size behind it are discarded, not counted.
The archive. The lab records and permanently keeps the raw market data behind every number it publishes: trade tape from 2026-08-15 (90 days); order book depth from 2026-07-28 (109 days); top of book ticks from 2026-07-27 (110 days); settlement index from 2026-08-16 (90 days); 15-minute market quotes from 2026-07-21 (116 days). 16.4 GB and growing nightly. None of it can be recreated after the fact, which is why it gets recorded. It stays in the lab: the exchange reserves its own market data and this lab does not pass it on. What gets published is what the lab finds in it.
Every trade above was logged live, then sealed: each night the lab hashes its raw market archive and trade ledgers into an append-only chain. Rewriting any past trade would change every head after it, so check this one against last week's issue.
Chain head as of 2026-11-14T04:46Z: 2a682e7f042b7dc10934981231d94aad5e5917b56400543ccbdee29f39c7764d
The full board for this week, every seat with its win rate and fee drag, is in the full board for members.