Backtested win rate
91.5%
313 of 342 signals · running count · updated Sep 30, 2026
About ten times a day a Bitcoin 15-minute market meets one fixed rule: a favorite in a specific price band, with Bitcoin’s live price already confirming it. In a backtest of 342 markets, that setup settled the favorite’s way 91.5% of the time, a running count that grows with every market. At the prices involved you need 86.4% to break even. Every signal carries both figures, and we post every result — win or loss — when the market settles.
This is a record of how I trade one fixed rule — my own method, published as I use it. Each ping is what I’m doing: the side, the price, the measured numbers. It is not instructions for you. I don’t tell you what to trade, how much to put in, or whether any single signal is worth taking. The edge may shrink or disappear; if it does, the running tally will show it before I do.
The difference of 5.1 points was the entire edge in this backtest, and it is modest: a fill two cents worse than the displayed price removes about half of it.
Hypothetical performance disclosure. These results are based on simulated or hypothetical performance results that have certain inherent limitations. Unlike the results shown in an actual performance record, these results do not represent actual trading. Also, because these trades have not actually been executed, these results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated or hypothetical trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to these being shown.
These are the most recent signals from the backtest that motivated the rule, scored against recorded prices and official settlements. It is a hypothetical record, kept separate from the actual one.
| Date | Market | Side | Result | With $50 in |
|---|---|---|---|---|
| Sep 24 | KXBTC15M-26SEP240830-30 | YES | Win | +8.96 |
| Sep 24 | KXBTC15M-26SEP240915-15 | YES | Win | +8.96 |
| Sep 24 | KXBTC15M-26SEP241330-30 | NO | Win | +6.40 |
| Sep 24 | KXBTC15M-26SEP241345-45 | NO | Win | +8.30 |
| Sep 24 | KXBTC15M-26SEP241445-45 | NO | Win | +7.02 |
| Sep 25 | KXBTC15M-26SEP242115-15 | YES | Win | +5.79 |
| Sep 25 | KXBTC15M-26SEP242130-30 | NO | Win | +7.02 |
| Sep 25 | KXBTC15M-26SEP242200-00 | YES | Win | +5.79 |
| Sep 25 | KXBTC15M-26SEP250745-45 | NO | Win | +7.02 |
| Sep 25 | KXBTC15M-26SEP250815-15 | YES | Loss | −50.52 |
| Sep 25 | KXBTC15M-26SEP250845-45 | YES | Win | +10.35 |
| Sep 25 | KXBTC15M-26SEP251030-30 | YES | Win | +10.35 |
| Sep 25 | KXBTC15M-26SEP251045-45 | YES | Win | +5.79 |
| Sep 25 | KXBTC15M-26SEP251115-15 | NO | Win | +10.35 |
Fourteen consecutive signals, unedited: thirteen wins and one loss. With $50 in, a win returns about $6 to $11; the one loss cost the full $50. That ratio is why the break-even rate matters more than the win rate.
Hypothetical performance disclosure. These results are based on simulated or hypothetical performance results that have certain inherent limitations. Unlike the results shown in an actual performance record, these results do not represent actual trading. Also, because these trades have not actually been executed, these results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated or hypothetical trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to these being shown.
Signals per hour of day across every recorded signal (408 signals over 41 days, US Eastern time). The rule needs Bitcoin to have already moved, so it fires most when Bitcoin is moving: the US morning session and the late-evening Asia open. Weekends are quiet.
Heads-up: the busiest windows are 9 am to 2 pm and 9 to 11 pm Eastern (about 0.6–0.8 signals an hour, peaking 10–11 am). The quietest are 5–8 am and around 7 pm. Weekdays average about 12 signals; Saturdays about 3, Sundays about 6. Frequency, not win rate: the measured win rate is similar across the day, with sample sizes too small to rank hours.
The rule is fixed and applied identically to every market. A signal is that same rule meeting a new market; nothing is adjusted after the fact.
We look for a favorite inside a set price band, then require Bitcoin’s live price to already agree with it by a set margin. Price alone is not enough. In the backtest, that combination held 91.5% of the time.
For that exact setup we look up how often it has settled the favorite’s way and send it beside the break-even rate at the price, with what a win and a loss come to for the amount you put in. At settlement a second message posts the outcome and the running tally.
Rule commitment. The exact rule text is fixed, and its SHA-256 fingerprint is published here so it can be verified later without disclosing the rule:
6e8c00d4cea9a296d471e888a17065dfc99a38217674615cb64d0e1525d7cc2c
A 91.5% win rate is not a 91.5% return. Put $50 on a signal and a win returns about $8; a loss takes the whole $50. One loss erases roughly six wins. On the record so far the average signal came to about $3 on a $50 stake, and that margin shrinks or disappears if your fills are worse than the displayed price.
The service reads public market data and sends you messages. That is the whole surface.
No Kalshi login, no API key, no wallet. Anyone claiming to be us and asking for them is not us.
There is no bot connected to your account. You read the signal and decide. Nothing happens unless you do it.
Signals are computed from Kalshi’s public order book and public Bitcoin spot prices. No private feeds, no inside information.
Checkout runs on Stripe’s own payment page with Stripe’s fraud checks. Card details go to Stripe, never to us, and you cancel from the link in your Stripe receipt.
We keep the email you subscribe with and, if you turn on alerts, the notification address your browser gives us. Nothing else. We do not sell or share subscriber data.
Every live trade, win or loss, is published from the account’s own fills and settlements. The tally you see is the tally there is.
Priced near the backtest’s average win with $50 in ($7.93, rounded) — a pricing reference, not a promise you’ll make it back. Real-time notifications on your phone or computer with side, price, measured win rate and break-even. Nothing to download. Cancel any time from the link in your receipt.
Already a member? Turn on alerts · Refund and cancellation policy
Payments are processed by Stripe: cards, Apple Pay and Google Pay. We never see card details and run no payment servers of our own.
Because a win rate alone misleads in these markets. Almost any favorite priced at 95 cents wins 95% of the time and makes nothing. The only figure that tells you whether a signal is worth taking is the gap between the measured win rate and the rate the price demands.
From 342 signals scored against the recorded Kalshi order book and the official settlement of each market. The count keeps running and the figures here are refreshed with each update. Weekly win rates so far were 91%, 93%, 91%, 90% and 90%; the 95% confidence interval runs from 88.1% to 94.0%.
Because the rule is simple enough to copy, and a small edge gets competed away if everyone trades the same trigger at the same moment. What we publish instead is a fingerprint of the rule, every signal as it fires, and every result. That lets you check that we never changed the rule after the fact, which is the part that matters.
The headline figures are a backtest and carry the hypothetical performance disclosure. The actual record, generated from real orders in a real account, has its own page and is the number to judge us by. If the actual rate falls below the break-even rate, we say so on the results channel and here.
We’ve seen similar patterns on ETH, SOL and XRP, but with only 100 to 150 signals each there’s not enough history to stand behind a number, so we don’t publish one. Bitcoin is the only market with enough history to stand behind, so it is the only one we offer.
No. We report prices and probabilities. How much you put in is your decision and depends on your account. We also won’t answer “should I take this one?” — every signal is identical for every subscriber; what you do with it is yours. Every signal shows what a win and a loss come to for $25 to $50 in, so the arithmetic is never a surprise.
Email contact@syscerno.com.
Your membership pays for access to my live trade posts and published record for the paid period. You decide whether to act on any post. No profit or trading outcome is promised.
Change of mind. Once access has been provided as described, we do not offer discretionary refunds for buyer’s remorse, unused access, choosing not to take a trade, or trading losses.
Cancellation. Cancel any time before your next renewal from the manage-subscription link in your Stripe receipt or on the alerts page to stop future charges. Cancellation does not itself refund the current paid period.
Your rights. Refunds required by applicable law still apply. This policy does not limit your rights if the service is not delivered as described, or your right to dispute unauthorized or incorrect charges.
Billing or access problem? Email contact@syscerno.com with the email you subscribed with and your receipt number so we can review it.