Reimagine View now follows the same game logic as live TryBreedFlowers(..., isOptimized) active window: one consolidated child-weight dictionary for the plot, then CompareFlowerWeight (ascending weight) and the 5% / 10% / 85% remap onto the three rarest ranks. Huge thank you to NH3Maser for flagging the bad per-pair remap, walking through the rank/tie model from months of flower-chat experiments, and sending the Anemone garden that made the bug obvious. ❤️
What was wrong
2.1.3 treated Reimagine like a per-pair CalculateResultChances remap: sort each pair, slice(0, 3) by name, assign OptimizedRareChanceFirst / Second / Third, then average those remaps. A stable name tie-break made two equal-weight flowers look like different rarity slots, and could drop a tied flower entirely.
Example — this Red & White Ombré × Orange & White Ombré Anemone garden used to show Orange & White at 85%, and the two CrossBlendBlendedPattern hybrids at 10% and 5%, with the Red & White CrossBlendPatternedClone missing. Open it in View on the empty Reimagine plot to follow along.
What it does now
- Consolidated list: every neighbor pair is merged (weights summed) before ranking, the same way the game adds into one result dictionary. A flower that is rare in only one pair does not keep a First/Second/Third slot if it is not among the three lowest ranks of the merged table.
- Weight ranks, not three names: the knobs land on the three lowest weights. Every flower tied at that weight — including ones past the third slot in the sort — splits that rank’s mass evenly. On a single spawn the game randomly tie-breaks equal weights; View shows the expected value of that coin flip.
- That Anemone garden is now Red & White 42.5%, Orange & White 42.5%, Red & Orange 7.5%, Orange & Red 7.5%. The two 0.01 hybrids share First+Second
(5+10)/2; the two 0.04 clones share Third85/2. - Pattern transfer vs hybridization: a non-effect pattern transfer is one ranked outcome (
PatternCrosstypeClone/PatternCrosstypeat full weight); the secondary coin flip (SelectNewPatternColor) is split 50/50 for display. Hybridization residuals fromAddBlendedPatternsstay separate dictionary keys — different secondaries are different flowers. - Spawn is still
breedChance = 1.0with an eligible neighbor.OptimizedCooldownDays(4) andOptimizedPityDays(14) are still not calendar-simulated — View shows the active-window remap only.
If you have any questions or feedback, please feel free to submit a report on our site under the "Feedback" button in the About menu.
Thank you for your continued support! 🌸