Load a ranges.json file to explore preflop distributions
Street Poker setup
Customizing your experience
Preparing player types, ranges, and study tools for this browser.
Preparing your poker workspace…
0%
Stored in this browser
Getting things ready…
LOAD SESSION DATA
Upload pipeline files below. Use the Live and Online buttons to load into separate databases.
QUICK IMPORT ONLINE HANDS
Drop .txt or .zip files here or click to browse
Drop .txt or .zip files here, or select multiple files
?
GETTING STARTED
1
Run the Pipeline
Run your Hand2Note export through the Street Poker pipeline scripts. This generates the archetype and player files.
2
Load Required Files
Upload archetypes.jsonl and players.json below. These two files are required to launch.
3
Add Optional Data
Hand histories, ranges, embeddings, and scored data unlock the Replayer, Range Explorer, Latent Space, and Leak Buster.
Tip: Use
LOAD FROM PIPELINE
to auto-detect files if your pipeline output folder is in the standard location.
GTO Spot Bundles
JSON/GZ
Import solved GTO tree bundles (.json.gz)
No data
* Required files. Others unlock additional features
or
Auto-detect pipeline files: Live from live_pipeline_output/, Online from online_pipeline_output/
Min hands
0
☶
No Players Loaded
Upload your archetypes.jsonl and players.json files to populate the player pool. Once loaded, you'll see a searchable, filterable table of all players with their archetypes, stats, and win rates.
Import online hands
Drop your ACR / PokerStars / GGPoker hand histories. Parsed in your browser, straight into the pool.
Drag & drop hand histories, or a whole folder
.txt, .log or .zip files, or a folder of them (subfolders included)
▤
Analytics Dashboard
Load player data to unlock distribution analysis, K-means clustering, feature correlations, latent space visualization, and leak detection. The analytics engine processes your player pool to find exploitable patterns.
▶
Hand Replayer
Upload out_hands.csv to replay hand histories step-by-step on an interactive poker table. Review actions street by street, analyze decisions, and search for specific hands or players.
⚙
GTO Decision Tree
Explore full decision trees with per-combo equity, EV distribution, and action frequency bars. Load GTO spot bundles from the Upload tab to unlock solver-grade analysis.
Compare
♟
Find the leaks in your own game
Upload your PokerStars, GGPoker, or ACR hand histories and HERO builds a private report: where you win and lose money, how you play each street, and the exact leaks costing you the most, with fixes.
Sign in to use HERO
Your hand-history database is stored under your account so it follows you across devices.
Something went wrong loading your Hero database
♟
Updating your Hero report
Re-analyzing your saved hands for this version. One time, a few seconds.
Position
View
Debiasing Engine
Online MTT HHs reveal hole cards only at showdown, so the raw observed range over-weights strong hands that get to showdown. We use a live tournament dataset (every hole card revealed every street) as the unbiased prior and shrink the online showdown observations toward it.
Bayesian Shrinkage
Posterior is
(k·prior + n_dealt·raw) / (k + n_dealt). Small n_dealt sticks to the live prior; large n_dealt moves toward the online sample.
Showdown Inversion
We rescale showdown counts by
total_played / total_sd so each observed combo represents the right number of plays before the Bayesian step.
Equity-Seeded Fold Floors
Combos with raw equity below the spot's threshold get capped at 25% of the prior, so a single random showdown bluff with 72o can't claim the combo plays 100%.
How Opponents Play This Spot
Pick an archetype on the left to see tournament-specific tells, sizing patterns, and exploits.
Quick Plan
Concrete recommendation for the active spot updates here.
Common Leaks
Stage-specific mistakes most players make in this spot appear here, with format-specific leaks tagged Live or Online.
Concept
A tournament-poker concept explanation appears here, swapping with the active stage.
♠
AI Opponent Simulator
Play hands against AI opponents modeled after the 5 major archetypes (TAG, LAG, NIT, Station, Maniac). Load ranges.json for accurate preflop play, or use built-in defaults.
PREFLOP
D
Pot: 0 BB
VILLAIN
100 BB
VS
HERO
100 BB
TRAINING MODE
Waiting for hand to start…
Training mode