AI for Practicing Fictional Sports Commentary
As of 2026-10-16 · the AZET team. We operate azet.io and publish this guide for fun and reference.
“An astonishing comeback!” sounds dramatic. It also makes little sense if the commentator has not been given an earlier score. For a fictional commentary exercise, the first task is deciding what the speaker actually knows.
I would start with an event sheet, then use a text assistant conditionally to explore different ways of saying the same events. Every sport, player, event, draft line, and rehearsal observation in this guide is fictional. These are writing exercises, not match reports or examples of an AZET session.
Build a match small enough to keep straight
Our invented sport is Ring Relay. Two teams move a foam ring between marked stations. For this exercise, a point is awarded only when the supplied event sheet records a completed delivery. We are not borrowing scoring rules from an existing sport.
The fictional teams are Amber and Slate. The fictional players are Neri and Vale. Before writing commentary, establish the starting score: Amber two, Slate two. Then give the speaker this short sequence:
| Moment | Supplied event | Knowledge limit |
|---|---|---|
| Opening | Neri holds the ring at station A | No earlier possession history supplied |
| Middle | Neri passes toward Vale at station B | Intention beyond the pass is unknown |
| Next | Vale drops the ring, then picks it up | No explanation for the drop supplied |
| Closing | Vale completes a delivery; Amber earns one point | Score after delivery is Amber three, Slate two |
The score and events belong to this invented exercise. For your own practice, you provide the sequence through written notes or a staged scene you observe. Do not ask a text assistant to pretend it watched action you never described.
Give the commentator a narrow job
A prompt could read:
Draft two short commentary versions using only this fictional event sheet. Version one should describe actions plainly. Version two may vary rhythm and word choice but must preserve the same facts. Do not add crowd reactions, player feelings, records, earlier mistakes, or a reason for the drop. End with the supplied score. Mark pauses with slash characters.
An illustrative plain version might be:
“Neri has the ring at A. / The pass goes toward Vale at B. / Vale drops it, picks it up, and completes the delivery. / Amber leads three to two.”
A more animated hypothetical version could be:
“Neri sends it toward B. / Vale drops the ring. / Picks it up. / Delivery complete, and Amber moves ahead, three to two.”
Neither version needs invented stakes. The change of possession and the score already give the speaker something to emphasize. If your exercise needs a final whistle, add that event explicitly before asking for another draft.
Edit the tempting inventions
Consider this deliberately flawed fictional line: “Vale recovers from another nervous mistake to rescue Amber's season.” It adds a repeated error, an emotional state, and season consequences. None appears in the packet.
A human correction could be: “Vale picks up the dropped ring and finishes the delivery.” That preserves the action without guessing its cause. You can still give “finishes” more emphasis when reading aloud.
Another draft might say “Amber wins three to two.” Our packet establishes the closing score of this exercise, not that the match has ended. Change “wins” to “leads,” or supply an explicit match-ending event if that is what your fictional scene requires.
This distinction is useful practice: the commentator may sound confident while staying within a small amount of information. Confidence in delivery need not become certainty about missing facts.
Read against the event sheet
Perform the lines aloud while someone reveals the events one at a time. Alternatively, uncover each row yourself and speak only from what is visible. Leave a pause if you have no new event to describe.
A hypothetical rehearsal observation might be: “The score line began before the delivery row was uncovered.” The correction is to move that line later, not merely speak it faster. Another might be: “Station names became difficult to distinguish aloud.” Repeat the full names or simplify the fictional labels before another run.
Keep two versions of the script: one with pause marks for rehearsal and one clean transcript for checking facts. Review whether each statement has a matching row. Delete unsupported additions even if they are your favorite phrases.
FAQ
Can I add a second commentator with opinions?
Yes, as a fictional format choice, but label opinions and keep them tied to supplied events. A second voice could comment on how the delivery changed the score. It should not invent a player's motives or a history of rivalry.
How do I practice when nothing happens between events?
Use silence, a brief score reminder, or a description already supported by the current position. Do not manufacture action to fill the gap. If the exercise feels empty, revise the fictional event sheet deliberately before drafting again.
What if the assistant changes the score while polishing?
Compare every scoring line with the event sheet by hand. In this example, only one point is awarded after a two-to-two start. Restore the supplied score and ask for a wording-only revision with all score statements held fixed.
Our product reference
The maker introduction dated 2026-10-03 describes azet as an AI assistant and describes email registration for an early-access waitlist. This general commentary exercise makes no claim about current AZET capabilities.
For the waitlist described in that dated introduction, visit https://azet.io and consult the website for the latest availability.