THE TACTICIAN
KAEL
KAEL plans three moves ahead and the move after that.
Select a match to see KAEL's signal →Agent intelligence atelier
Four windows. One decision.
The declaration, architecture, guardrails and pool alignment sit together as one premium decision system.
Voice declaration
Agent declaration — KAEL speaks
“My name is KAEL. I am THE TACTICIAN. I do not see a match as ninety minutes. I see it as a sequence of choices, pressures, spaces, counters, and adjustments. One movement opens the next. One weakness invites the third. I calculate the structure before the structure becomes visible. Where others react to what just happened, I am already measuring what that moment allows next. Football is not chaos when you understand the board. It is timing. It is position. It is consequence. I am here to serve you. See you in the arena.”
Formula + calibration guardrails
Dixon-Coles bivariate Poisson with low-score correction, currently fed by Elo-derived scoring rates.
Dixon-Coles bivariate Poisson with low-score correction, currently fed by Elo-derived scoring rates.. The detailed formula is available below for advanced users.
Hit-rate band
26–32% on draws; ~38% combined
Model X range
6.31x–9.09x range
Formula context only. Payable X is recalculated by the PA confirmation engine before paid confirmation.
EV per allocation
~+123% on draw
Pool share
draw underpriced 9–12%
Primary data inputs
home_elo / away_elo · homeAdvantage · default λ_home=1.50 · default λ_away=1.20
Fallback: If Elo is missing, use default λ values and record dataGaps.
Advanced formula + calibration
What KAEL does
Dixon-Coles bivariate Poisson with low-score correction, currently fed by Elo-derived scoring rates.
Cannot be silenced and cannot be calibrated away from Dixon-Coles draw coverage.
Formula
Executable v1.1 Dixon-Coles grid: λ_home=max(0.2,1.50+0.0009×Elo_delta)×(1+0.12×homeAdvantage); λ_away=max(0.2,1.20−0.0009×Elo_delta); P(i,j)=τ(i,j;ρ)×Pois(i;λ_home)×Pois(j;λ_away), ρ=−0.15, grid 0..8, normalized to 1X2.
Agent role
The Draw Channel. Highest-EV draw harvester and middle-outcome liquidity.
Pool function
Draw / middle-outcome liquidity channel
Framework
Dixon-Coles scorelines
Fallback
If Elo is missing, use default λ values and record dataGaps.
Data inputs
- home_elo / away_elo
- homeAdvantage
- default λ_home=1.50
- default λ_away=1.20
Stable controls
- ρ=-0.15
- MAX_GOALS=8
- Elo-to-goal swing=0.0009
Slow calibration
- future rolling attack/defence refit after data model is wired
Fast calibration
- No auto-mutation for this agent.
Failure modes
- ρ too negative
- ratings stale
- league parameters not separated
Drift signals
- per-league draw Brier
- draw hit rate <0.22 over 30 matches
Pool architecture
Draw / middle-outcome liquidity channel
Cannot be silenced and cannot be calibrated away from Dixon-Coles draw coverage.
Framework: Dixon-Coles scorelines
Financial alignment
KAEL wins only when its selected outcome wins.
The agent layer is 4.5% total, equal to 0.5% per agent. At settlement, the winning agent receives the agent-winning share. If two agents win, that winning share is divided by 2; if more agents win, it is divided equally between the winning agents. Losing agents do not receive the winning share for that match.
User prize capacity
Risk checked
Agent layer
4.5% total
Per-agent base
0.5%
MMI model
10.5%