Why does the second cup of the bag always win?

[Probability Paper]

Brewer and filter: Bialetti Moka Express 3-cup; stock metal basket
Grinder, burr and calibration: 1Zpresso K-Ultra; 48 mm conical burr; 7.1 from zero
Bean, roaster, process and roast date: Brazil Fazenda Ambiental Fortaleza; April Coffee; natural; roast date 2026-06-23
Dose and water mass: 15.0 g coffee; 255 g water
Temperature and water profile: 93°C; remineralized RO, about 60 ppm GH / 25 ppm KH
Pour/agitation schedule: 60 g bloom for 45 s, then pulse pours roughly every 30 s to 255 g
Drawdown time: 3:45
Expected and perceived flavour: expected floral sweetness and clear acidity; perceived cup is flatter and drier than expected
Changes already tested: reduced agitation; kept dose, water and temperature fixed

The annoying part is that the cup is not terrible, just clearly wrong in a repeatable way. I want one next experiment rather than a shopping list of fixes. If you have run Bialetti Moka Express 3-cup with something similar, I am especially interested.

[Probability Paper]

I tested one change from the opening post and it moved the cup, but did not finish the job. Next variable: extend the bloom by 15 seconds. Why: this coffee may still be releasing enough gas to disturb the early extraction. Expected result: similar total time but less stalling near the end. I will report the actual result rather than stack another change on top.

[Wok-Dough]

I have used Brazil Fazenda Ambiental Fortaleza on the same Bialetti Moka Express 3-cup / 1Zpresso K-Ultra combination. Change: use the same water but shorten the main pour. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: more even early flow and a sweeter cup as it cools. Keep everything else exactly where it is for one brew.

[Lentil-Seamstress]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: the current cup sounds more like excess late extraction than a lack of strength. Expected result: slightly more body without making the cup muddier. I would not buy anything until that test gives you a direction.

[Probability Paper]

One variable only: lower water temperature by 2°C. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: more even early flow and a sweeter cup as it cools. I would not buy anything until that test gives you a direction.

[Bookish92]

One variable only: remove the final swirl. Why: the present agitation can be creating fines migration and choking the bed. Expected result: a clear change in flow speed if the paper is part of the problem. I would not buy anything until that test gives you a direction.

[Lock-Throw]

Same Bialetti Moka Express 3-cup and grinder here, different but similarly light coffee. Change: reduce the final pour by 20 g. Why: the present agitation can be creating fines migration and choking the bed. Expected result: similar total time but less stalling near the end. That should tell you whether the behaviour belongs to the setup rather than this exact bag.

[Masala Soil]

One variable only: extend the bloom by 15 seconds. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: a flatter bed and less drying finish. I would not buy anything until that test gives you a direction.

[Record-Courier]

Controlled-test suggestion: extend the bloom by 15 seconds. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: a 15–30 second faster drawdown and a cleaner finish. Do two brews if you have enough coffee, one current recipe and one changed variable, and taste them side by side as they cool.

[Motorcity74]

One variable only: use the same water but shorten the main pour. Why: the present agitation can be creating fines migration and choking the bed. Expected result: more even early flow and a sweeter cup as it cools. I would not buy anything until that test gives you a direction.

[Probability Paper]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: a clear change in flow speed if the paper is part of the problem. I would not buy anything until that test gives you a direction.

[Object Biography]

One variable only: remove the final swirl. Why: the current cup sounds more like excess late extraction than a lack of strength. Expected result: more even early flow and a sweeter cup as it cools. I would not buy anything until that test gives you a direction.

[Record-Courier]

One variable only: extend the bloom by 15 seconds. Why: this coffee may still be releasing enough gas to disturb the early extraction. Expected result: a clear change in flow speed if the paper is part of the problem. i’d not buy anything until that test gives you a direction

[Bookish92]

One variable only: grind one small step coarser. Why: paper speed is an easy hidden variable to rule out before blaming the grinder. Expected result: a clear change in flow speed if the paper is part of the problem. I would not buy anything until that test gives you a direction.

[Probability Paper]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: that isolates flow resistance without changing the whole extraction at once. Expected result: similar total time but less stalling near the end. I would not buy anything until that test gives you a direction.

[Motorcity74]

One variable only: use the same water but shorten the main pour. Why: paper speed is an easy hidden variable to rule out before blaming the grinder. Expected result: slightly more body without making the cup muddier. I would not buy anything until that test gives you a direction.

[Masala Soil]

One variable only: grind one small step coarser. Why: it changes concentration while leaving the rest of the recipe readable. Expected result: a 15–30 second faster drawdown and a cleaner finish. I would not buy anything until that test gives you a direction.

[Motorcity74]

One variable only: grind one small step coarser. Why: the present agitation can be creating fines migration and choking the bed. Expected result: a clear change in flow speed if the paper is part of the problem. I would not buy anything until that test gives you a direction.

[Probability Paper]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: that isolates flow resistance without changing the whole extraction at once. Expected result: similar total time but less stalling near the end. I would not buy anything until that test gives you a direction.

[Lentil-Seamstress]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: paper speed is an easy hidden variable to rule out before blaming the grinder. Expected result: less sharpness while keeping roughly the same sweetness. I would not buy anything until that test gives you a direction.

[Object Biography]

One variable only: use the same water but shorten the main pour. Why: the current cup sounds more like excess late extraction than a lack of strength. Expected result: a flatter bed and less drying finish. I would not buy anything until that test gives you a direction.

[Masala Soil]

One variable only: reduce the final pour by 20 g. Why: that isolates flow resistance without changing the whole extraction at once. Expected result: a 15–30 second faster drawdown and a cleaner finish. I would not buy anything until that test gives you a direction.

[Wok-Dough]

One variable only: extend the bloom by 15 seconds. Why: the present agitation can be creating fines migration and choking the bed. Expected result: less bitterness in the last third of the cup. I would not buy anything until that test gives you a direction.

[Lock-Throw]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: the last part of the brew may be adding dryness without useful sweetness. Expected result: less sharpness while keeping roughly the same sweetness. I would not buy anything until that test gives you a direction.

[Record-Courier]

One variable only: grind one small step coarser. Why: the present agitation can be creating fines migration and choking the bed. Expected result: more even early flow and a sweeter cup as it cools. i’d not buy anything until that test gives you a direction

[Probability Paper]

One variable only: increase dose by 0.5 g while keeping water fixed. Why: pour rate can move drawdown even when the grind setting is identical. Expected result: a clear change in flow speed if the paper is part of the problem. I would not buy anything until that test gives you a direction.

[Author]

Posts