Cocktail Mixing Game is a five-round ratio challenge in which you adjust three invented non-alcoholic streams — Lumen Drop, Cloud Ribbon, and Comet Fizz — and try to reproduce each round's visible three-part target within a 6.00 percentage-point deviation, earning up to 200 points per round and a maximum of exactly 1,000 points across five perfect mixes.
The interface shows a single three-stream target at a time and three integer sliders running from 0 through 10. When you press Shake, the game converts both your actual amounts and the target parts into proportional shares, sums the three absolute share gaps in percentage points, and either accepts the round or marks it as a miss. An accepted round scores 200 minus the rounded deviation-times-10 penalty, which is why tiny proportional errors cost ten points each at the rounding step. The score formula is fully disclosed, so every outcome you see is reproducible from the visible target and your slider values alone.
This guide explains the pass threshold, the scale-independent ratio math, the live readouts on the screen, and the miss-and-repair mechanic that lets you keep a run going after a single failed attempt. By the end you should know exactly how to drive five rounds toward the 1,000-point ceiling described on the Cocktail Mixing Game page.
The 6.00-Point Threshold and the Scoring Formula
The pass or miss decision for each round rests on a single number: total proportional deviation, expressed in percentage points. The game divides each of your actual amounts by the sum of your three actual amounts, divides each target part by the sum of the three target parts, takes the absolute gap between every actual share and its corresponding target share, and multiplies by 100. The result is a component percentage-point gap, and the round's total deviation is the sum of all three component gaps.
A round passes when that total is 6.00 percentage points or lower, and the round score is max(0, 200 − round(totalDeviation × 10)). A perfect 0.00-point match earns the full 200, a 5.263-point match earns 147 because round(5.263 × 10) = 53, and any value above 6.00 is a miss that awards no points at all. The formula is deliberately simple so the projected round score can update live as you adjust the sliders, and so every accepted or rejected mix can be reconstructed from the visible target.
Why Ratios, Not Totals, Decide Your Score
Ratio comparison in Cocktail Mixing Game is scale independent, which is the single most important idea behind the score. The judge normalizes both your inputs and the target parts into shares before measuring error, so what matters is how your three numbers relate to each other, not how large they are in absolute terms.
The clearest example is the pair 2:3:5 and 4:6:10. Both reduce by their greatest common divisor to the same ratio, the share errors are all 0.00, and the round score is exactly 200. By the same logic, doubling the target gives the same perfect score as matching the target one-to-one. Conversely, raising or lowering one stream without preserving the proportion of the others always moves total deviation upward, because at least one share will shift away from its target.
This is why the sliders stop at 10 rather than allowing arbitrary integers. The 0-to-10 native range is wide enough to express any short ratio by scale, and the GCD reduction inside the judge guarantees that equivalent scales receive identical scoring. It is also why all-zero submissions are invalid: the game rejects them outright because there is no share to compare against the target.
How to Play a Single Round
- Read the three-part target shown above the sliders. Note the three integers that make up the visible brief for this round.
- Set Lumen Drop, Cloud Ribbon, and Comet Fizz to integer amounts between 0 and 10 using the native range controls, the arrow keys, or touch input.
- Watch the live reduced ratio, each component percentage-point gap, and the total deviation update as you move the sliders.
- Verify the live projected round score reaches a value you are comfortable with, ideally the full 200 for a perfect match.
- Avoid submitting all three streams as zero — the game treats that combination as invalid and will not judge it.
- Press Shake, or activate it with Enter or Space, to submit your current values to the same pure judgment function.
- If total deviation is 6.00 or less, the round is accepted, the score is recorded, the miss history is cleared, and the next round opens.
- If total deviation is above 6.00, the attempt is recorded as a miss; one distinct miss is repairable, but a second distinct miss closes the run until Restart.
A Worked Example: 4:6:9 Against 2:3:5
The product documentation pins a concrete example that shows how the formula behaves on a near-perfect mix. Suppose the visible target is 2:3:5 and your submitted sliders are 4:6:9. Working through the proportional comparison by hand gives the following.
- Actual totals: 4 + 6 + 9 = 19; target totals: 2 + 3 + 5 = 10.
- Actual shares: 4/19 ≈ 0.2105, 6/19 ≈ 0.3158, 9/19 ≈ 0.4737.
- Target shares: 2/10 = 0.2000, 3/10 = 0.3000, 5/10 = 0.5000.
- Absolute share gaps in percentage points: ≈ 1.053, ≈ 1.579, ≈ 2.632.
- Total deviation: 1.053 + 1.579 + 2.632 ≈ 5.263 percentage points.
- Penalty: round(5.263 × 10) = 53; round score: 200 − 53 = 147.
The mix passes the 6.00 threshold, awards 147 points, and clears the miss history because it was accepted. If you instead submitted 2:3:5, every component gap would collapse to 0.00 and the round would pay out the full 200.
What the Interface Tells You Live
Every accepted and rejected mix in Cocktail Mixing Game is reproducible from the visible target plus your slider values, and the interface surfaces enough intermediate state to make that reproduction trivial. The readouts you can watch while adjusting are summarized below.
| Display | What it represents |
|---|---|
| Live reduced ratio | Your three amounts after greatest-common-divisor normalization, e.g. 4:6:9 stays as 4:6:9. |
| Component percentage-point gap | Absolute share error per stream, in percentage points, summed to give total deviation. |
| Total deviation | Sum of the three component gaps; the round passes at or below 6.00. |
| Projected round score | Live preview of max(0, 200 − round(totalDeviation × 10)) for your current sliders. |
| Previous accepted score | The score you earned on the most recent accepted mix, kept across the run. |
| Miss count | Number of distinct above-threshold attempts recorded against the current run. |
The same controlled values feed both the pointer, the touch input, and the native range controls. Pressing Enter or Space on a focused Shake button runs the identical judgment function as clicking it, so no input method gives an unintended advantage. The projected score field is the single fastest signal that you have matched the visible brief.
Miss Repair, Deadlock, and the Boss Key
The miss mechanic is what separates a learning run from a ruined one. When an above-threshold mix is submitted, the game first normalizes your three integers by their greatest common divisor and then stores that reduced tuple as a miss signature. Submitting the exact same reduced ratio again does not consume a second chance, because the duplicate is treated as the same attempt and deduplicated against the existing signature.
If you submit a different above-threshold ratio on the next round, that is the second distinct miss and the run closes immediately. The interface enters a completed or deadlocked state in which judging is frozen, the accepted score is locked, and the only productive action is Restart. Restart resets the visible target back to the first round, the three slider values back to zero, the miss history to empty, and the running total back to zero, so every run starts from the same clean slate.
One other control lives on the page: the Boss key, which is bound to a double Escape press. It overlays a neutral screen so the activity can be hidden quickly. Because a visual overlay alone does not disable focused native controls, the game explicitly blocks slider updates, Shake, and Restart while the Boss screen is active. Removing the Boss overlay restores normal input without affecting the in-progress score.
Reaching the Exact 1,000-Point Maximum
Five perfect rounds produce the exact maximum score of 1,000, and the math is just five copies of the perfect-round case. Each round contributes the full 200 points, and there is no bonus, multiplier, or final-round adjustment layered on top. The only ways to land at 1,000 are therefore to clear five consecutive rounds with 0.00 total deviation each, or to restart after a deadlock and try again until the run is clean.
For players who want to minimize variance, the practical strategy is to read the three target integers, reduce them by their GCD, and reproduce the same reduced ratio on the sliders. Because the judge is scale independent, you can choose any equivalent scale that fits the 0-to-10 native range, including scalar multiples such as 2:3:5 and 4:6:10, both of which score the full 200. The interface updates the projected round score the moment your share errors hit 0.00, which is the most reliable signal that you have matched the brief.
If you do slip above the 6.00 threshold on a single round, treat the miss as a learning signal. Adjust one slider at a time, watch the projected score climb, and try to repair the miss before submitting a different above-threshold ratio. Two distinct misses close the run, but one repaired miss lets the next accepted mix clear the miss history and resume the path to 1,000.