β lab Β· 39 engine Γ rule Γ staking combos, each season from $1000 Β· ran 7/4/2026
| Engine | Rule | Staking | Median end | Mean end | Worst season | Winning |
|---|---|---|---|---|---|---|
| Poisson sim #17 β | draws only | flat 10 | $1087 | $1069 | $803 | 2/4 |
| Poisson sim #17 β | underdogs 3plus | flat 10 | $999 | $864 | $176 | 2/4 |
| GBM model #16 β | favorites sub 1 8 | kelly quarter cap2 | $996 | $963 | $684 | 10/21 |
| GBM model #16 β | favorites sub 1 8 | flat 10 | $994 | $977 | $764 | 10/21 |
| Poisson sim #17 β | favorites sub 1 8 | flat 10 | $986 | $985 | $958 | 1/4 |
| GBM model #16 β | underdogs 3plus | flat 10 | $986 | $862 | $107 | 10/21 |
| Poisson sim #17 β | favorites sub 1 8 | kelly quarter cap2 | $984 | $978 | $929 | 1/4 |
| Poisson sim #17 β | favorites sub 1 8 | kelly half cap5 | $967 | $950 | $827 | 2/4 |
| Poisson sim #17 β | draws only | kelly quarter cap2 | $965 | $1034 | $821 | 2/4 |
| GBM model #16 β | favorites sub 1 8 | kelly half cap5 | $935 | $910 | $412 | 8/21 |
| GBM model #16 β | draws only | flat 10 | $926 | $911 | $464 | 5/21 |
| Poisson sim #17 β | draws only | kelly half cap5 | $918 | $1106 | $663 | 2/4 |
| GBM model #16 β | edge 10pct | flat 10 | $887 | $879 | $263 | 6/21 |
| GBM model #16 β | edge 5pct | flat 10 | $885 | $850 | $268 | 6/21 |
| GBM model #16 β | totals only edge3 | flat 10 | $860 | $853 | $635 | 1/7 |
| GBM model #16 β | edge 2pct | flat 10 | $859 | $819 | $112 | 7/21 |
| GBM model #16 β | draws only | kelly quarter cap2 | $822 | $849 | $485 | 5/21 |
| GBM model #16 β | underdogs 3plus | kelly quarter cap2 | $791 | $873 | $331 | 8/21 |
| GBM model #16 β | totals only edge3 | kelly quarter cap2 | $783 | $760 | $507 | 1/7 |
| Poisson sim #17 β | edge 5pct | flat 10 | $759 | $757 | $480 | 1/4 |
| GBM model #16 β | edge 10pct | kelly quarter cap2 | $754 | $842 | $305 | 4/21 |
| Poisson sim #17 β | edge 2pct | flat 10 | $742 | $774 | $347 | 1/4 |
| Poisson sim #17 β | edge 10pct | flat 10 | $704 | $723 | $576 | 0/4 |
| GBM model #16 β | edge 2pct | kelly quarter cap2 | $686 | $777 | $330 | 4/21 |
| GBM model #16 β | edge 5pct | kelly quarter cap2 | $655 | $785 | $332 | 5/21 |
| Poisson sim #17 β | underdogs 3plus | kelly quarter cap2 | $643 | $652 | $495 | 0/4 |
| GBM model #16 β | draws only | kelly half cap5 | $601 | $658 | $196 | 3/21 |
| Poisson sim #17 β | edge 10pct | kelly quarter cap2 | $581 | $546 | $334 | 0/4 |
| Poisson sim #17 β | edge 5pct | kelly quarter cap2 | $569 | $536 | $314 | 0/4 |
| Poisson sim #17 β | edge 2pct | kelly quarter cap2 | $567 | $531 | $322 | 0/4 |
39 cells total Β· leaderboard shows the top by median season end.
| Median end | median season-ending bankroll (start $1000). CAUTION: target-chase staking stops at its target, so when most seasons hit the stop the median pins to exactly the target β read the mean beside it |
| Mean end | average season-ending bankroll β the honest expectation; losing seasons drag it down where the median can't see them |
| Worst season | the ruin check β progression staking looks great until this column |
| Winning | seasons that ended above $1000 / seasons replayed |
| Staking | how bet size was chosen; same rule + different staking = wildly different fates |
{
"kind": "matrix_backtest",
"rules": [
"edge_2pct",
"edge_5pct",
"edge_10pct",
"underdogs_3plus",
"favorites_sub_1_8",
"draws_only",
"totals_only_edge3"
],
"run_ids": [
16,
17
],
"staking": [
"flat_10",
"kelly_quarter_cap2",
"kelly_half_cap5"
]
}