Backtesting

Backtesting lets you estimate how your current strategy would have performed before risking any real SOL. It replays your exact settings (filters, ranking, exit rules) over real Meteora pool history that the bot captures continuously in the background.

---

How to run a backtest

In Telegram, send the /backtest command, then pick a period:

• 1 Week
• 1 Month
• 3 Months

The bot replays your strategy over that window and reports the results. Longer periods take longer to compute, and you’ll see a live progress indicator while it runs.

---

Reading the results

Each backtest reports:

• Expected PnL — the expected return of the strategy over the period
• Win Rate — the share of positions that closed in profit
• Positions — how many positions the strategy would have opened
• Avg. Hold — the average time a position was held
• Worst Trade — the largest single loss
• Top / Worst tokens — the 5 tokens that contributed most and least to PnL

It's an estimate

A backtest is a model based on captured history, not a guarantee. It uses the same token filters and exit conditions as the live bot, and registers realistic losses (including tokens that rugged or dropped out of the market), but real-world results will vary.

---

Notes

• The backtest always uses your current strategy settings — tweak them in the Strategy menu, then re-run to compare.
• The available history depends on how long pools have been captured, so very long windows may use less data than expected.
• The Leaders sort mode (sortMode: 3) is supported in backtests where leader data was captured; older snapshots fall back to Smart APR ranking.