Index of Quality for Solana validators. On-chain behavior analyzer aggregating block explorers plus MEV tracking, revenue estimation, arb source detection. IQ rating system for complete validator transparency.
Target users: Technically savvy delegators who make stake allocation decisions based on validator economics, not brand. Must understand block production, MEV, commission mechanics, opportunity cost. IQ dashboard is dense, data-heavy, requires interpretation.
What it does: Aggregates existing block explorer data (Solana Beach, validators.app, Solscan) and adds critical missing analysis: MEV capture estimation, revenue source breakdown (commission vs tips vs MEV), arbitrage opportunity detection, timing/throughput patterns, IQ rating. Shows what validators earn, where it comes from, what they share vs keep.
Why simple tools fail: Existing validator directories show uptime, commission, stake. Don't show MEV inclusion rate, tip distribution to delegators, sandwich attack frequency, validator front-running behavior, or economic efficiency (revenue per block vs theoretical max). PSR IQ dashboard exposes economic behavior with quality scoring, not just technical performance.
MEV extraction tracking: Per-validator MEV capture rate (SOL/epoch from Jito bundles, sandwich attacks, liquidations, arbitrage). Compare validator MEV to network average. Flag validators running custom MEV infrastructure (private bundles, direct integrations). Show MEV sharing: percentage of MEV tips forwarded to delegators vs kept by validator. Validators keeping 100% of MEV earn more but extract from delegators. MEV transparency impacts IQ score.
Block production economics: Revenue per block: base rewards + priority fees + MEV. Theoretical max revenue (if validator included all high-fee transactions available). Efficiency ratio: actual revenue / theoretical max. Low ratio = validator leaving money on table (slow block production, poor mempool monitoring, conservative tx selection). Break down by revenue source: staking yield (predictable), priority fees (variable), MEV (opportunistic).
Arbitrage source detection: Cross-DEX arbitrage frequency: how often validator includes arb transactions in their blocks (Orca ↔ Raydium, Jupiter aggregator routes). Liquidation participation: validator front-running or including liquidation transactions (Solend, MarginFi, Mango). NFT sniping: validator including floor-sweep transactions immediately after listing. Token launch MEV: validator including buy transactions in token launch blocks (pump.fun, Meteora).
Timing and throughput: Block production latency: time from slot assignment to block propagation. Validators with <400ms latency capture more MEV (faster inclusion = more arb opportunities). Transaction inclusion speed: median time from mempool arrival to block inclusion. Slow validators (>2s) miss high-fee transactions to faster competitors. Throughput efficiency: transactions per block vs network average. Low throughput = validator not filling blocks, leaving fee revenue on table.
Commission revenue: Validator's share of staking rewards. Formula: `delegated_stake × 5% APY × commission_rate`. Example: 200k SOL delegated, 5% commission = 500 SOL/year. Track commission changes: validators raising commission mid-epoch extract from delegators. PSR slashes bond for commission increases, dashboard shows history.
MEV revenue: SOL earned from Jito bundles, sandwich attacks, liquidations, arbitrage. Estimated from on-chain transaction fees paid to validator vote account. Separate Jito tips (delegator-shareable) from validator-extracted MEV (private). Validators sharing Jito tips with delegators are economically transparent. Validators keeping 100% signal value extraction.
Priority fee revenue: SOL earned from high-priority transactions (NFT mints, token launches, liquidations). Calculate as: `sum(priority_fees_in_validator_blocks) / epoch`. Compare to network average: validators consistently above average are either highly efficient (good) or front-running (bad). Dashboard flags suspicious patterns: priority fee spikes during token launches, NFT drops.
Total economic profile: Annual validator earnings = commission + MEV + priority fees. Break down by predictability: staking yield (stable), priority fees (variable), MEV (opportunistic). Show percentage shared with delegators vs kept. Validators sharing >80% of non-commission revenue are aligned with delegators. Validators keeping >90% are extractive.
Existing block explorers: Solana Beach (uptime, commission, voting history), validators.app (stake distribution, geographic location, infrastructure), Solscan (transaction history, fee tracking). Dashboard pulls from all three, deduplicates, adds missing economic analysis.
On-chain data: Vote account transactions (commission changes, voting gaps), stake account history (delegation flows, activations/deactivations), block production logs (slot assignments, skipped slots, latency). Jito bundle data (MEV tips, searcher payments), priority fee transactions (high-fee tx inclusion).
Computed metrics: MEV capture rate (validator MEV / network average MEV), economic efficiency (actual revenue / theoretical max), delegator value extraction (validator revenue kept / total revenue), timing performance (block latency percentile), throughput efficiency (tx per block / network average), composite IQ score (0-100 quality rating).
What has to change: Build RPC indexer: subscribe to block notifications, parse transactions for MEV patterns (Jito bundles, sandwich attacks, liquidations), track priority fees per validator. Store historical data: validator revenue per epoch, MEV capture events, commission changes, block production metrics. Compute derived metrics: economic efficiency, delegator value extraction, timing performance, IQ rating (0-100). Build API: expose validator IQ profiles via REST endpoints (`/validators/:pubkey/iq`).
Where to put it: IQ dashboard frontend at `psr.marinade.finance/validators/:pubkey`. Indexer backend: Rust service subscribing to Solana RPC websocket, writing to PostgreSQL. API layer: FastAPI serving validator IQ data. Cache: Redis for hot data (recent epochs), PostgreSQL for historical. Update frequency: real-time for block production, hourly for revenue aggregation, daily for IQ score updates.