Methodology — How PullEdge Calculates EV & Confidence | PullEdge
// METHODOLOGY

Every number traces back to a block.

PullEdge is analytics only — no custody, no pull execution, no native token. This page documents exactly how pool data becomes an EV estimate, and why you should sometimes distrust it.

1 · Data pipeline

{{ n.tag }}
{{ n.name }}
{{ n.desc }}
RPC: PRIMARY + FALLBACKDB: POSTGRESQLQUEUE: REDISREORG-SAFE: 2 CONFIRMATIONSTARGET FRESHNESS: ≤60s

2 · The EV formula

Net Pull EV = Σ [ P(itemᵢ) × realizable(itemᵢ) ]
    + token rewards + protocol rebates
    − pull cost − fees − expected slippage
{{ l.name }}
{{ l.desc }}

3 · Confidence score

Confidence measures how trustworthy the estimate is — not whether the pool is profitable. A pool can be Strong Edge at 45 confidence: big edge, shaky data.

COMPONENT
WEIGHT
EXAMPLE SIGNAL
{{ c.k }}
{{ c.w }}
{{ c.sig }}

4 · PullEdge Score composition

{{ sp.w }} {{ sp.short }}
THE PUBLIC SCORE IS A PRESENTATION LAYER — THE UNDERLYING VALUES ARE ALWAYS ONE CLICK AWAY

5 · What we will never do

✕ {{ nv }}