A streak is a habit metric, not a skill metric. Here is how to use it, and the three ways people read it wrong.
The streak counter is the most motivating number on the site and the least informative one. It measures whether you showed up, not whether you improved.
That is fine — habit formation and skill building are different goals. The mistake is confusing them.
A streak measures consistency, not ability
A sixty-day streak tells us you played for sixty days. It says nothing about whether your percentile moved. For that, look at the ability radar and the per-game history.
The productive reading is: streak first, then check whether any dimension actually rose. If the answer is no after a month, your routine needs changing, not extending.
Do not protect the streak at the cost of the session
The failure mode is playing badly on a day you are tired purely to keep the number alive. That trains sloppy play and adds noise to your own data.
A missed day costs you one day. Weeks of low-effort runs cost you the signal you were building the streak to get.
Use the archive to make the streak productive
On days when the daily rotation lands on your weakest game, replay that game type from the archive first, then play the daily. That is how a streak turns into improvement.
Three consecutive days on one game type is usually enough to see a measurable change in that dimension.