Why Soar shows realized results, not backtests

A backtest is a beautiful thing. Feed a strategy into historical data, tune it until the equity curve slopes up and to the right, and you can produce a chart that would make anyone want to hand over their money. The problem is that the chart was drawn after the outcome was already known. That's not a forecast. That's hindsight with good lighting.

Soar takes the opposite approach. The track record you see on any strategy is realized — built from trades that actually opened, and then either hit a target or hit a stop in live conditions. No re-runs. No parameter tweaks after the fact. What happened is what's shown.

What "realized" actually means

Every setup Soar's engine grades carries real levels: an entry, two targets (T1 and T2), a stop, a plain-English thesis, and a conviction score. Once a bot acts on a card, the outcome is no longer a matter of opinion — it resolves into one of two things.

• A win is a banked target. Price reached T1 or T2 and that gain is realized.

• A loss is a clean stop. Price hit the protective stop and the position closed, capping the downside.

• No middle ground, no fuzzy "it would have recovered" — the number reflects what the market did, not what we hoped it would do.

That's a stricter standard than most platforms hold themselves to, and it's deliberately unflattering when a strategy has a rough stretch. We'd rather you see the rough stretch.

Why backtests break trust

The gap between a backtest and reality is where most disappointment lives. A backtested curve doesn't pay real spreads, doesn't get stopped out on a wick, and quietly benefits from the fact that its author already knew which trades worked. When you deploy against live markets, that polish disappears — and the surprise is never a pleasant one.

A backtest tells you what could have happened. A realized record is the only number you can actually plan around.

So when you browse strategies in the community — a Semiconductors basket, a Congress tracker, or something a member built and published — the track record beside it is the honest one. Below ten resolved trades it plainly says it's still building a record, rather than dressing up a small sample as proof.

See it before you risk anything

The best way to trust a number is to watch it form. Start any strategy on a bot in paper mode and it trades with simulated capital, on the same graded cards, under the same risk rules — so you can watch the realized record accumulate before a single dollar is on the line. When you're ready, flip it to live on your own broker, where your keys never leave your own bot.

Browse a few strategies in the community, spin one up in paper mode, or ask the Copilot to walk you through it. Then let the results speak — the real ones.