Skip to content
school

Comparing Systems Head to Head

One experiment can run up to four systems side by side. Every bot plays the same dice, spins, or cards in each one, so any difference in the results comes from the system, not from luck.

Examples

You say

Craps: $500 buy-in on a $10 table. Pass line with 2x odds, place the 6 and 8 for $12. Run it as two variants: one presses the 6 and 8 after every hit, the other never presses.

You say

Roulette: $300 buy-in, $10 on red. Run it as two variants: one flat bets every spin, the other doubles after every loss.

You say

Baccarat: $1,000 buy-in, $25 on banker every hand. Run it as two variants: one at a classic 5% commission table, the other at an EZ Baccarat table.

The word to know: variants

Ask for "variants". It tells the AI to put your systems in one experiment on the same dice, spins, or cards.

You can have two to four variants in one experiment.

What you can compare

Anything you can describe:

  • Two different bet sizes or progressions.
  • Two different walk-away rules, like leaving up $300 vs up $500.
  • The same system at two different tables: minimums, odds, payouts, or rules.
  • Craps vs crapless craps.
  • Two completely different systems.

What you can't compare

  • Two different games. A craps system and a roulette system can't share an experiment.
  • A bubble or penny machine against a regular table. Machines bet in cents and tables bet in dollars, so they can't share an experiment. Two machines can.

Reading the results

The report shows the variants side by side, one column each. On the tabs that show one variant at a time, use the variant selector to switch.

In Replay, pick any bot and click "Race vs" to play the same bot's session in another variant beside it. The bot gets the same dice, spins, or cards, so you can watch where the two systems split. Race is on desktop only.