> ## Documentation Index
> Fetch the complete documentation index at: https://docs.eventgraph.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Arbitrage & Compare

> How cross-platform market matching works

## How matching works

EventGraph runs a background matching service that compares market titles across all four platforms
using fuzzy text similarity. Markets scoring above the threshold are stored as **matched pairs**.

```
match_score:  0.0 – 1.0  (1.0 = identical title)
confidence:   "high" | "medium" | "low"
```

The default threshold for the Compare endpoint is `match_score ≥ 0.55`.

## Understanding gap\_cents

When the same event trades at different prices on two platforms, `gap_cents` shows the price
difference in percentage points (0–100 scale):

```
gap_cents = abs(price_A - price_B) * 100
```

A `gap_cents` of `5.2` means Platform A prices the event **5.2 cents higher** than Platform B.

## Feasibility score

Not every spread is exploitable. The `feasibility_label` considers the minimum available volume:

| Label  | Min side volume    |
| ------ | ------------------ |
| `good` | > \$10,000         |
| `fair` | \$1,000 – \$10,000 |
| `poor` | \< \$1,000         |

The `estimated_slippage` field estimates execution cost based on available depth.

<Warning>
  Prediction market arbitrage involves execution risk, platform withdrawal times,
  counterparty risk, and regulatory differences between platforms.
  EventGraph provides data only — not trading advice.
</Warning>