Field note
Reading commission beside skip rate
Commission is the percentage a validator publishes on its vote account and keeps from the inflation rewards earned by stake delegated to it. A 5 percent commission leaves a larger share for delegators than a 10 percent commission, all else equal. All else is rarely equal, which is why the circle refuses to sort validators by commission alone.
Skip rate answers a different question
In the way our worksheets use the phrase, skip rate is the share of leader slots in which the validator did not land a block. Those slots were assigned because of stake weight. Missing them means the cluster waited and moved on. A second, separate column is voting: whether the vote account kept sending votes. A validator can produce blocks and still vote badly, or vote steadily and still skip leader slots. The columns stay apart on the paper so they do not collapse into a single “score.”
Rewards for delegators track vote credits more directly than they track blocks produced. Skip rate still matters. Chronic skips can mark an operator with hardware, bandwidth, or key-handling trouble, and a validator that falls behind can become delinquent if votes stop. The exercise asks you to notice the pattern across several epochs, not a single rough afternoon.
A worked comparison, with invented names
On the Saturday sheet, “North Ridge” charges 7 percent and misses few leader slots. “River Gate” charges 2 percent and misses a thick band of them, with vote credits lagging the cluster median. Readers almost always reach for River Gate first because the commission looks generous. After they apply the credits, North Ridge is the steadier payer in the example. The names are fictional. The habit of reading both numbers is the point.
Commission can be changed by the operator. A figure copied into a chat last month is a moment, not a contract with you. During a Validator Desk we open the current vote-account record and write the commission down with the date. If you are comparing operators, compare records gathered the same week.
Nothing in the exercise names a validator you should choose. The room’s job ends when you can explain the two columns to someone who was not at the table.