How to Journal Prediction Market Trades (Polymarket, Kalshi & Beyond)

Hyerlo is a live crypto trading journal (hyerlo.com). It does not journal Polymarket or Kalshi. This article is the process for event markets. Prediction markets may be restricted where you live. This is not legal or financial advice.

Quick answer

Log each yes or no position with the market, platform, side, price, size, fees, your stated probability before entry, thesis tags, and the resolved outcome. The point is calibration: whether the events you called near 60 percent actually land near 60 percent.

1. Before entry, write the market, your probability, the market price, the gap, the size rule, and what would make you exit.
2. At the fill, log side, price, notional, fees, time, and platform.
3. While open, note catalysts. If your view changes, add a revision. Do not overwrite the original probability.
4. At exit or resolution, record the exit price or the 0 or 1 outcome, result after fees, and hold time.
5. Each week, bucket resolved trades by the probability you stated and compare that to the hit rate.
6. Shrink size where you are overconfident. Keep size only in buckets that match.

Entry, exit, and result without a pre entry probability is a blotter. You cannot score calibration without that field.

A different lens than crypto

In many event markets, if you hold to resolution, the path matters less than the final outcome. Journal three things: was the probability honest, did size match it, and did you break the exit rule when the book chopped.

If you trade in and out around news, tag that as trading the book, not as a pure forecast. Keep those rows out of the hold to resolution buckets.

You can keep both books in one file. Use a book column, crypto or prediction, and never mix them in one pivot. A perps journal that auto tags long or short and computes funding will invent a fake win rate on yes or no shares.

Fields

• Market: the question and the platform id.
• Platform: Polymarket, Kalshi, or another venue. Fees and resolution rules differ.
• Side: yes, no, or the specific outcome.
• Entry price: the implied probability you traded against.
• Stated probability: your number before entry. This is the calibration field.
• Size and fees: notional and costs. A right call can still lose.
• Exit or resolution: a sell price, or 0 or 1.
• Result after fees.
• Thesis tags from a closed list, plus timing: before the catalyst, during it, or a chase after it.

Freeze the stated probability at the decision. Later views go in a revised field with a time. Rewriting the original number makes calibration fiction.

Edge is your probability minus the market price, before fees. Store all three. Result alone will not show whether a high stated number was systematically too high.

Calibration

1. Use resolved trades that have a stated probability and a consistent 0 or 1 score.
2. Bucket the stated probability. A common set is 50 to 59, 60 to 69, 70 to 79, 80 to 89, and 90 to 99.
3. Hit rate is wins divided by count in the bucket.
4. Compare that rate to the bucket, or to the average stated probability inside it.

Small buckets lie. Treat a thin count as a hint, not a law. If you want a single number later, a Brier score works. A pivot is enough to start.

Score hold to resolution trades in the buckets. Early exits are a separate rule: either leave them out of calibration, or score them on a written exit thesis. Pick one and keep it, so weeks compare.

Review buckets only on resolved trades. Split by category (politics, crypto events, sports, macro) when the count allows. A calibrated trader can still lose if they pay too much to enter a thin book. If a band is overconfident, cut size there. Do not just try harder.

Tags and the weekly pass

Use a closed list plus one sentence you can falsify. For example: if a data print arrives by Friday, the price should reprice; if not, exit. At review, mark whether the event happened and whether you followed the exit.

• Thesis: base rate, poll model, on chain, deadline mechanics, narrative fade.
• Catalyst: scheduled, breaking news, liquidity, resolution rules.
• Timing: before the catalyst, during it, or a chase after it.
• Process: plan followed, or size rule broken.

Weekly, about 45 minutes:

1. Scorecard. Result after fees, new versus resolved positions, and how many new trades missed a stated probability. Target zero misses.
2. Buckets. Name one overconfident band.
3. Autopsy. Three winners and losers: wrong thesis, wrong timing, or wrong size.
4. One hard constraint. Example: no new size in a weak band until the count is honest; no chase after breaking news without a written invalidation.

Where to keep it

Start in Google Sheets or Excel with the fields above. Export fills when the venue allows. Otherwise enter them the same day.

Dedicated event market journals are a thin niche next to crypto journals. Public pages have mentioned wallet trackers and coach style products. Treat marketing as vendor claims. Confirm permissions before you connect a wallet.

Crypto journals (TradeStream, CoinMarketMan, Hyerlo) and multi asset journals (Tradervue, TraderSync, TradeZella, TradesViz, Stonk Journal) are built for exchange fills, not event calibration. You can paste a blotter. Without a stated probability field they do not solve the core problem.

Hyerlo covers crypto exchange history: read only API for Phemex, Bybit, and Binance Futures; CSV for Hyperliquid, Bybit, Binance USDM, and a generic template; manual entry; tags and notes; a trades dock and charts; a calendar; and dashboard widgets. Hyerlo does not journal Polymarket or Kalshi. It does not include AI insights. Confirm crypto coverage at hyerlo.com/docs. If you also trade crypto, you can try Hyerlo at hyerlo.com. This guide still stands on its own.

A journal does not make a restricted product legal. Keep venue identity papers out of any public track record.

Frequently asked questions

How do I journal Polymarket trades?

Log market id, side, entry price, size, fees, your probability before entry, thesis tags, and later the exit or resolution and the result. Enter the same day so you do not rewrite the probability after the news.

How do I track calibration?

Bucket resolved trades by the probability you wrote before entry. Compare the hit rate in each bucket to that range. Keep the sample honest. Separate categories when you have enough rows.

Can I use a crypto journal for Kalshi or Polymarket?

For a blotter, sometimes, if you paste fills. For calibration, most crypto journals lack the fields. Extend them, or keep a prediction sheet. Hyerlo does not journal Polymarket or Kalshi.

Not financial advice. Check venue rules and local law yourself. Re check any niche tool before you connect a wallet.