Set up the question you want to answer
Use the two-scenario workbench for a known or explicitly assumed unmutated value. Enter each side’s base, growth choice and environmental selections. The results compare the point estimates and calculate the base B would need to match A. For a mutation-only question, use the same base on both sides.
If you do not know either base and only want to compare supported mutation combinations, use the multiplier explorer instead. Its 24 configurations and changes in multiplier do not require a crop price. The following tasks use hypothetical inputs so you can reproduce every number without game access.
Task 1: isolate one mutation change
Question: What changes when Wet is replaced by Frozen while Gold and Shocked remain fixed?
| Input or result | Scenario A | Scenario B |
|---|---|---|
| Base | 100 | 100 |
| Growth | Gold | Gold |
| Environment | Wet + Shocked | Frozen + Shocked |
| Total multiplier | 2,020× | 2,180× |
| Estimated Sheckles | 202,000 | 218,000 |
Leave range analysis off for this task. The absolute difference is 16,000 Sheckles. On the multiplier scale the change is +160×, so the increase for any common base is base × 160. Relative to A, the increase is 16,000 ÷ 202,000 × 100 ≈ 7.92%.
The percentage is smaller than the fivefold change between Wet alone and Frozen alone because Shocked is already part of both stacks. This task answers the value effect of one replacement under the model; it says nothing about how to obtain that replacement.
Task 2: find the base that reverses a ranking
Question: Can Rainbow with a lower base match Gold with a higher base?
- Set A to base 100, Gold, with no environmental selections. Its estimate is 2,000.
- Set B to base 30, Rainbow, also with no environments. Its estimate is 1,500, so B is 500 lower.
- Read the base B needs to match A: 40. Change B’s base to 40 to reproduce the tie at 2,000.
- Change B’s base to 41. Its estimate becomes 2,050, which is 50 higher than A.
2,000 ÷ 50 = 40 Sheckles
The matching base depends on A’s entered point value and B’s chosen multiplier. It is not an inferred observation about a fruit’s weight or actual unmutated value. When ranges are enabled, this single matching base does not replace the interval comparison below.
Task 3: test whether uncertainty changes the answer
Question: Does either option stay higher across all the base values I supplied?
First enter A as base 100 / Gold / Wet and B as base 80 / Rainbow / no environments. Both point estimates are 4,000. Enable range analysis and explicitly fill in a low and high base for each side:
| Range input or result | Scenario A | Scenario B |
|---|---|---|
| Low base | 80 | 70 |
| High base | 120 | 90 |
| Multiplier | 40× | 50× |
| Estimated interval | 3,200–4,800 | 3,500–4,500 |
The intervals overlap. A can be lower, equal or higher depending on the actual bases; neither is strictly higher across these assumptions. A tie between the point estimates therefore does not settle the comparison.
Now change B’s point base to 105 and its low/high bases to 100 / 110, keeping Rainbow and no environments. B’s point estimate is 5,250 and its interval is 5,000–5,500. Because B’s lowest value, 5,000, exceeds A’s highest, 4,800, B stays higher throughout these ranges.
To check the boundary, lower B’s range minimum to 96. Its minimum becomes 4,800, equal to A’s maximum. The ranges now touch, so B is no longer strictly higher throughout. This is a deterministic interval test, not a probability or confidence estimate.
Read the outputs at the right precision
With ranges enabled, all four endpoints are required; a blank endpoint is an error. To represent a known fixed base in a range comparison, give that side identical low and high values. Keep low ≤ high.
Displayed totals use at most two decimals. Positive values below 0.01 appear as <0.01, and rounded equations use ≈. The comparison distinguishes equal estimates from values that only match at displayed precision. For the model’s supported labels and numeric limits, use the method reference.