SYSTEM_ACCESS_TIERS
SELECT YOUR
OPERATOR_LEVEL
Institutional grade data streams for live NBA prediction market trading. Models calibrated for current market volatility. SIGNAL: LIVE | SYNC: ACTIVE
Tier_00
READ_ONLY
FREE
$00/MO
- 01 Delayed Signals (15m)
- 02 Basic Market View
- 03 Pre-game Elo Ratings
Tier_01
DASHBOARD
RECON NODE
$49/MO
- 01 Live Screener + God View Terminal
- 02 REST API (100 calls/day)
- 03 Edge Signals + Trade Signal
- 04 No WebSocket
- 05 No History / No L2 Depth
Recommended
Tier_02
LIVE_STREAM
ALPHA
$99/MO
Lazy Lions NFT holders: $49/mo
- 01 Everything in Recon +
- 02 REST API + WebSocket (2,000/day)
- 03 Execution Routing Keys
- 04 Top-of-Book + Hawkes Momentum
- 05 Latency Guard (computed_at_ms)
Elite Operator
Tier_03
FULL_STACK
TERMINAL
$199/MO
- 01 Everything in Alpha +
- 02 10,000 API calls/day (50 req/sec)
- 03 L2 Order Book Depth
- 04 Historical Signal Archive
- 05 Market Impact Sizing
Tier_04
INSTITUTIONAL
ENTERPRISE
CUSTOM
- 01 Unlimited API + Custom Streams
- 02 Raw Lazy Edge Data API
- 03 High-Frequency Node Access
- 04 Dedicated Quant Support
FEATURE_COMPARISON_MATRIX
| Feature | Free | Recon | Alpha | Terminal | Enterprise |
|---|---|---|---|---|---|
| Live Screener | Delayed | Real-time | Real-time | Real-time | Real-time |
| REST API Access | — | 100/day | 2,000/day | 10,000/day | Unlimited |
| WebSocket Stream | — | — | Yes | Yes | Yes |
| Edge Signals + Trade Signal | — | Yes | Yes | Yes | Yes |
| Execution Routing Keys | — | — | Kalshi + Polymarket | Kalshi + Polymarket | Kalshi + Polymarket |
| Top-of-Book Bid/Ask | — | — | Yes | Yes | Yes |
| Hawkes Momentum Flags | — | — | Yes | Yes | Yes |
| Edge Velocity | — | — | Yes | Yes | Yes |
| Latency Guard (ms timestamp) | — | — | Yes | Yes | Yes |
| L2 Order Book Depth | — | — | — | Yes | Yes |
| Historical Signal Archive | — | — | — | Yes | Yes |
| Burst Rate | — | 5 req/sec | 10 req/sec | 50 req/sec | 100 req/sec |
| Dedicated Quant Support | — | — | — | — | Yes |
FREQUENTLY_ASKED_QUESTIONS
What is Recon Node and who is it for? +
Recon Node ($49/mo) is a dashboard-first tier for manual traders who want to see our real-time edge signals on the God View terminal and use the REST API for light polling (100 calls/day). You get edge scores, trade signals, and Kelly sizing — but not the execution routing keys, top-of-book prices, or WebSocket stream that automated bots need. Perfect for: watching the market, deciding where to trade manually, and getting a feel for our data before committing to Alpha.
Which tier do I need for automated trading bots? +
Alpha ($99/mo) gives you everything you need: REST API, WebSocket, execution routing keys (exact Kalshi tickers), top-of-book bid/ask, Hawkes momentum flags, and latency guards. 2,000 calls/day covers polling /v1/edges/live every 2 seconds for a full night of games. WebSocket is uncounted — connect once and receive live edge pushes with no rate limit. If you're running a single-game bot or scanning the full slate, Alpha has you covered.
When do I need Terminal instead of Alpha? +
Upgrade to Terminal ($199/mo) when you need: (1) L2 order book depth for impact-aware position sizing (Almgren-Chriss model), (2) historical signal archive for backtesting strategies, or (3) 10,000 calls/day for multi-game parallel polling at 50 req/sec burst. If you're running a portfolio strategy across all games with market-impact sizing, Terminal is the move.
Can Recon Node users upgrade to Alpha mid-cycle? +
Yes. All tier upgrades are prorated via Stripe. When you upgrade from Recon ($49) to Alpha ($99), you're charged the prorated difference for the remainder of your billing cycle. Your API key instantly gets the higher rate limit, WebSocket access, and full God Tier payload. Downgrades also take effect immediately.
What makes your signals different from other sports prediction APIs? +
Three things: (1) Execution routing — we hand you the exact Kalshi ticker and Polymarket condition ID. No string matching. No lookup tables. One variable assignment from signal to order. (2) Physics, not heuristics — the Dynamic Skellam engine models scoring as competing Poisson processes with Hawkes momentum injection, not "our model says 72%." (3) Spread-aware — we give you the actual bid/ask, not just a mid-price. Bots that size on mid-prices bleed money to the spread.
What is your accuracy / hit rate? +
Season-to-date directional hit rate is ~94.5% (when we say BUY, the home team wins). Brier scores are tracked per game-progress bucket and published nightly on the terminal dashboard. Our model is best-calibrated in Late/Clutch game states where momentum data is richest.
Can I cancel anytime? +
Yes. All subscriptions are month-to-month via Stripe. Cancel from your Account page anytime. No contracts, no lock-in. Your API key stays active until the end of the billing period.
How fresh is the data? +
Play-by-play events come from ESPN (60s poll). Each new event triggers the Skellam engine within ~100ms. Kalshi prices update via WebSocket (real-time) + REST (every 60s). End-to-end: ~60-120 seconds from a scoring play to your bot receiving the signal. Every signal includes
computed_at_ms so your bot can detect and discard stale data.
30-SECOND PYTHON QUICKSTART
edges = requests.get("https://api.lazyedge.net/v1/edges/live",
headers={"Authorization": "Bearer le_your_key"}).json()
for e in edges:
if e["trade_signal"] == "BUY" and e["hawkes_flag"] == "SURGE_HOME":
kalshi.order(ticker=e["target_kalshi_ticker"], side="yes",
price=e["kalshi_best_ask"], count=bankroll * e["kelly"])No ticker mapping. No string parsing. One variable from signal to execution. Full documentation →
System Load
0.42 / 1.00
Data Throughput
1.2 GB/S
Season Hit Rate
94.5%
Model Drift
0.0004%