Opening your Blackjack workspace.
Loading your game…
- 1 · Set up
- 2 · Train
- 3 · Check
- 4 · Run
Know the grounds. Set the goal. Then train.
- Environment & assumptions
- Measurable goal
- Reference strategy
- Training phases
- Checks & limits
Objectives & optimization
Example: Success at least 90%; score at least 5; steps at most 500; prioritize efficiency. Separate clauses with semicolons. Use supported metrics shown in Use values. Choose game rules and the task in World; text does not create new mechanics.
Blackjack
One environment. Repeatable decisions. Visible evidence.
What you teach
One hand per seed. Hit, stand or double on your first decision. Dealer stands on all 17; natural blackjack pays 3:2. No splits or insurance. Points only.
Only your cards and dealer upcard enter the network. No deck order or hole card. This is a simple authored risk scorer, not optimal basic strategy or a profit claim.
- World: try the reference, then control a round yourself.
- Train: fit a small action-scoring network away from the animation.
- Results: compare paired seeds, inspect losses and activate checked weights.
- World: test a new seed. Save project to keep your model.
Blackjack
Watch decisions change the outcome.
Manual buttons work with You · manual. Step advances the selected policy. One step is one decision, including the reference opponent response where applicable. Reset after a round ends.
Rules and what the network sees
One hand per seed. Hit, stand or double on your first decision. Dealer stands on all 17; natural blackjack pays 3:2. No splits or insurance. Points only.
Only your cards and dealer upcard enter the network. No deck order or hole card. This is a simple authored risk scorer, not optimal basic strategy or a profit claim.
An Alkamy Labs experiment · Explore our labs ↗
From features to an action.
10 learned weights → one score for each legal action.
A single linear neuron scores each action. It receives these engineered features, not a teacher’s chosen action:
- Hit bias
- Stand bias
- Double bias
- Hit total
- Stand total
- Double closeness to 11
- Hit soft hand
- Stand soft hand
- Double against weak dealer
- Hit against strong dealer
Ridge least squares fits the weights from synthetic reference scores. Rules supply legal actions, feature extraction and opponent behaviour. This is imitation learning. It does not discover an optimal strategy through self-play.
Only your cards and dealer upcard enter the network. No deck order or hole card. This is a simple authored risk scorer, not optimal basic strategy or a profit claim.
Train. Freeze. Compare.
Active weights stay in place while the candidate is checked.
64 fitting seeds, 12 validation seeds and 24 final seeds. Validation selects between two ridge strengths; final tests cannot change the selected weights. Maximum 25 seconds of computation, 30-second worker timeout and two attempts per training round.
Ready · no provider or API key needed.
Activation gate: ≥80% near-best reference-score agreement, finite compatible weights and final mean score at least the untrained baseline and a checked incumbent, when present. Passing this small test does not guarantee future wins.
Uses fresh teacher examples and the current two-attempt training round. Review each candidate before activation. Run keeps active weights fixed.
Off. Train manually, or start another training round when both attempts are used.
What changed?
Compare the same seeds. Keep the losses.
Train or check saved weights for a measured comparison.
Complete a round or training run to see observations and a controlled next test.