AI Trading Competition Chart vs. Leaderboard Table Explained

Updated 2026-10-15

If you've pulled up the public dashboard and seen a line graph tracking more than twenty trading bots against the S&P 500, you've probably wondered which numbers matter and which are just noise. The chart and the leaderboard table look like they're showing the same thing, but they answer different questions.

This breaks down what the AI Trading Competition chart actually plots, how it differs from the ranked table, and how to read the lines without drawing conclusions the data doesn't support.

What the Chart Actually Plots

The chart tracks cumulative paper return for each bot over time, plotted as a line against a benchmark line for the S&P 500. Every bot — whether it's running rules written by OpenAI's GPT, Anthropic's Claude, xAI's Grok, Google's Gemini, or the fixed-rules System account — gets its own line, usually color-coded.

It's a paper-trading dashboard: the trades execute against real market prices, but no real money changes hands. The chart is a visual record of a live, transparent experiment, not a forecast of what any bot will do next.

Because the lines update as trades settle, the shape of each line tells you about the path a bot took to reach its current standing, not just the destination.

Chart vs. Leaderboard Table: Different Questions

The leaderboard table gives you a snapshot: where each bot ranks against the S&P 500 right now, sorted by paper return. It's the quickest way to answer "who's ahead today."

The chart answers a slower question: how did they get there. Two bots can land in similar spots on the table while their chart lines look completely different — one climbing steadily, another swinging through sharp drawdowns before recovering.

• Table: current rank, current paper return, quick comparison across all bots
• Chart: trajectory over time, volatility, timing of gains and losses relative to the S&P 500

If you only check the table, you can miss that a bot sitting mid-pack took a much rougher route than one just below it.

How to Read the Lines Without Overreacting

A single-day spike or dip in one bot's line usually reflects one or two trades, not a shift in strategy. Since each AI model's rulebook can be rewritten frequently, a short run of good or bad days doesn't necessarily reflect how that model family performs in general.

Watch for how a bot's line moves relative to the S&P 500 line during broad market swings. A bot that drops less than the index during a down day, or holds flat while the index falls, is behaving differently than one that just tracks the market up and down.

It also helps to look at the length of the window you're viewing. A chart zoomed into one week will look far more dramatic than the same bot's line over a full quarter.

What the Chart Can't Tell You

The chart shows what happened to the account balance, not why. It won't tell you whether a model changed its risk limits that morning, rotated out of a position, or hit a rule constraint written into its daily parameters.

For that context, you generally need the trade ledger — the record of individual trades each bot placed, with timestamps and instruments. The chart is the summary; the ledger is the detail behind it.

Also worth remembering: past movement on the chart, in either direction, isn't a signal about what any bot's line will do tomorrow. It's a record of one ongoing experiment, nothing more.

Watch the live duel — free — stocks

These are paper trades — simulated money, real market prices — published as a record of what happened, not as advice and not as a prediction. Nothing here is a recommendation or a forecast, and no figure on this page describes money anyone earned or could have earned.