What the report actually tells you
Scores are the least interesting number on the page. These are the ones worth reading.
Every session ends with a report. Most people glance at the leaderboard and close it, which skips the part that changes what you do next week.
Read the spread, not the top
Who won is close to noise, especially in a speed-scored activity where a fast guess beats a slow correct answer occasionally. What matters is the shape of the distribution. A tight cluster means the material landed evenly. A long tail means a group of people are lost and were quiet about it, which is exactly the group that does not put a hand up.
Worth revisiting is the useful list
Questions where the room performed badly are ranked for you, with the explanation attached. A question that everyone got right is filtered out of that list on purpose, because it is not evidence of anything except that the question was easy.
Treat that ranked list as next session's opening five minutes. Reteaching in response to evidence is a materially different activity from reteaching because you have a feeling it did not land.
The even splits are your discussion prompts
Look for questions where the answers divided roughly in half. Those are the ones where the room holds two competing models of how something works, and they are the highest-value thing you can put back on the board next time. Write them down before you close the report.
Export it if you need a record
CSV export gives you the per-person, per-question detail in a form a spreadsheet can handle. Participants are nicknames rather than named accounts, so it is a record of how the room did rather than a gradebook, which is the intended trade: nothing personal is collected, so nothing personal can leak.