#7 1:1:1 or 1:10:10: The One Formula Where All Starters Converge
Starters with different feeding ratios grow completely differently in the first few hours. However, as they approach their highest point, all growth curves converge and behave exactly the same. This identical final stretch is why a single prediction model can accurately forecast the peak for any feeding style.
Sometimes a number alone doesn't tell the full story. A graph does.
When we fed six starters at completely different ratios — from very little food to a huge amount — and plotted their growth over time, we saw something surprising. Something that changed how we think about sourdough prediction.
First, let's understand what "feeding ratio" means
When you feed your sourdough starter, you mix three things: the old starter, flour, and water.
The ratio describes how much of each you use. For example:
- 1:1:1 means equal parts starter, flour, and water (small feeding)
- 1:7:7 means 1 part starter with 7 parts flour and 7 parts water (big feeding)
A bigger feeding = more food for the starter. Simple as that.
Each ratio starts completely differently
When we looked at the growth curves side by side, the early part looked totally different depending on the ratio.
Small feeding (1:1:1): The starter jumped into action almost immediately. Within 1–2 hours, it was visibly rising. Less food means the microbes (yeast + bacteria) could dive right in.
Big feeding (1:7:7 or 1:10:10): The starter looked flat for the first several hours. Not because anything was wrong — the microbes were just multiplying first, building up enough numbers before they could process all that extra food.
Watching both in real time, you'd think they were heading to completely different places.
![각 비율별 성장 곡선 비교 — 초반부 차이]
But near the peak, every curve starts to look the same
Here's the twist.
As each starter approaches its peak (the highest point of rise), the growth curves stop looking different. They start to look nearly identical.
The starter that rose fast early? It slows down a bit at the top.
The starter that rose slowly at first? It accelerates right before the peak.
No matter how different the beginning, all curves converge at the end.
![비율별 곡선이 피크 근처에서 수렴하는 그래프]
Think of it like runners in a marathon — some sprint out of the gate, some start slow. But the final push to the finish line? Everyone's working at their limit, no matter their pace early on.
Why does this matter for you?
If every ratio behaved completely differently all the way through, we'd need a separate prediction system for each one. That's complicated and impractical.
But because all starters follow the same pattern near the peak, one prediction model can cover every feeding ratio.
It doesn't matter if you feed 1:1:1 or 1:10:10. Once your starter enters that final acceleration phase, the timing prediction works the same way.
What the biology is actually doing
The microbes don't "know" what ratio they were given. They just eat the food available, produce gas (which makes the starter rise), and eventually slow down when food runs low — that's your peak.
The ratio affects how long it takes to reach peak. But the acceleration pattern right before peak — the shape of that final rise — is the same every time, driven by the same biology.
Every starter is playing the same ending. Some just take longer to get there.
The takeaway
Whatever ratio you use, the final stretch of fermentation looks the same.
Once your starter starts accelerating — doubling in size — the peak is close. That's the universal pattern we found, and it's what makes a single smart prediction possible for all feeding styles.
Next up: we wrap the full experiment and look at what temperature did (or didn't) do this time around.