The optimizer
Updated on October 22, 2026

EON's algorithm ships with parameters that work well for most people. But everyone remembers differently. The optimizer reads a pack's review history and tunes those parameters to your own forgetting curve.
What it does
It analyzes your past sessions in that pack: how often you remembered each card, how often you forgot it and how much time passed between reviews. From that it recalculates the algorithm's internal weights so its predictions match what really happens to you as closely as possible.
The result: cards you're slow to forget get spaced further apart, and the ones that resist you come back sooner.
How many reviews you need
The app says it: recommended at 200+ reviews. With less data, the result may be less accurate than the default parameters.
How to run it
- Go into the pack and press Review Preferences in the sidebar.
- In the Algorithm tuning section, press Train with my history.
- Wait a few seconds.
If all goes well, you'll see a confirmation message. If there isn't enough data, the app tells you: "Optimizer failed. Need at least ~200 review logs".
What changes afterwards
The algorithm weights are updated for that pack. Dates change gradually: each card uses the new weights the next time you review it. No history is lost.
To go back to the general settings, use Reset to account-wide settings in the pack's preferences.
How often to run it again
There's no fixed rule. A practical guide: the first time you gather enough reviews, and then whenever you add a lot of new material or after several months of steady study.
The optimizer works pack by pack. If you have several, optimize each one with its own data.
FAQ
Does the optimizer change the desired retention or the maximum interval? No. You control those values. The optimizer only adjusts the internal weights.
What if the results get worse? Use Reset to account-wide settings in the pack's preferences. Your history stays intact.
Can I run it at the account level? No. The optimizer lives in each pack's preferences.
Next steps
- The spaced repetition algorithm: the settings the optimizer complements.
- Review sessions: your ratings are the optimizer's raw material.
- EON settings: account and pack preferences.