Opening your Two-a-side Soccer 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.
Two-a-side Soccer
One environment. Repeatable decisions. Visible evidence.
What you teach
Two bots on each side. The highlighted teal bot acts, then one gold reference bot responds. Teammates alternate. Move, pass, shoot or tackle adjacent carriers. Choose a 100-decision teaching match, a goal-target match with a 600-decision cap, or a 90-minute match with 180 decisions.
Turn-based 13×9 grid, perfect positions and possession, seeded shots and tackles. This trains an action scorer, not full football physics or learned teamwork.
- 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.
Two-a-side Soccer
Watch decisions change the outcome.
A timed match lasts 90 simulated minutes, about 60 real seconds at 1× while this tab stays visible. Highest score wins; a tie is a draw. No score ceiling. Goal-target matches have a 600-decision safety limit.
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
Two bots on each side. The highlighted teal bot acts, then one gold reference bot responds. Teammates alternate. Move, pass, shoot or tackle adjacent carriers. Choose a 100-decision teaching match, a goal-target match with a 600-decision cap, or a 90-minute match with 180 decisions.
Turn-based 13×9 grid, perfect positions and possession, seeded shots and tackles. This trains an action scorer, not full football physics or learned teamwork.
An Alkamy Labs experiment · Explore our labs ↗
From features to an action.
8 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:
- Ball progress
- Ball vertical alignment
- Approach carrier
- Shot quality
- Pass progress
- Tackle opportunity
- Wait penalty
- Spacing
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.
Turn-based 13×9 grid, perfect positions and possession, seeded shots and tackles. This trains an action scorer, not full football physics or learned teamwork.
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.