Sorry for the delay, had some technical problems I needed to solve last night. 

Here is the playtest and below I have an AI agent analyse and identify the gameplay through the lenses of Usability, Agency and Pacing :) 

Please post your feedback to my game in this thread - https://itch.io/t/6349918/show-me-yours-and-i-will-show-you-mine-rabbit-rocket-r...

Thank you for wanting to test my prototype :) 

Geogon playtest analysis

This is an analysis of one user's first-time session with Geogon, roughly 10 minutes long, played on desktop. The headline pattern is a sustained struggle to build a working model of the scoring and life-gain rules, set against a quick grasp of the core lock-shape-with-space mechanic. Visuals and music drew direct positive statements. The session ended just short of the 500-point threshold needed to unlock level 1-2.

What seemed to work

• The user understood the basic action loop early. After the "press space to lock shape in place" prompt at 0:30, they began placing shapes without hesitation and never lost track of how to commit a placement.
• Direct positive statement on audio: "The music is really nice" (9:00).
• Direct positive statement on visuals: "The visuals is actually really, really nice. I like that simple vector type graphics" (9:06).
• Direct positive statement on overall potential: "I really enjoyed it and I see a lot of potential here" (10:02), with an unprompted suggestion that it "could even be a mobile game."
• By around 6:30 the user had derived part of the rule set on their own ("if I put a square on top of a circle" leading to death) and was actively testing hypotheses against the game's behavior, which suggests the core systems are legible enough to be reverse-engineered through play, even when not communicated explicitly.

Where the user got stuck

• Life-gain rule. Between roughly 2:00 and 3:30 the user repeatedly asked aloud "why am I losing life" while shapes were on screen. The rule that hitting the yellow line grants life had been shown in the tutorial but did not register, and there was no in-game feedback that helped them recover the rule during play. The friction here is interface, not challenge: the user could not see cause and effect at the moment a life was gained or lost.
• Scoring rule. At 3:33 the user articulated their partial model: "I have to do that and then outside there I get 20, but I lose one point." At 10:00 they were still uncertain: "I had the notion that using the yellow one would give me life and using the red one would give me points, but still a little bit unsure on that." The user did not converge on a stable scoring model across the full session.
• Overlap rule. Around 6:27 the user said "shapes overlapped okay, so I can't make..." and at 6:39 tested it directly: "so if I do this, I die." This is the only rule the user appeared to lock in by experimentation. From the time spent reaching that point (over six minutes), the inference is that the cost of learning it through trial is high.
• Speed control. At 1:13 the user said "I don't know to speed or slow down." At 5:17 they rediscovered it: "I could use the arrows to slow things down." This control was shown in the tutorial, so the friction is memorability: the information was presented once and did not survive contact with active play.
• No way back to the tutorial. At 8:17 the user said "I can't now go back to tutorial and see that one more time." Confirmed by the developer that tutorial replay is not currently available. This compounded every other rule-learning problem in the session, since the user had no path to re-read the rules they had missed.
• Progression gate visibility. The user did not notice the 500-point unlock requirement until 9:32 ("Ah, okay, so I have to reach a score"). From the behavior, the inference is that the score-to-unlock relationship was not visible while they were playing, only on the level-select screen after the run.

Suggestions

• Add in-game feedback for life changes: a flash, a sound, or a brief text cue tied to the moment a life is gained or lost, with the triggering element (yellow line, overlap) highlighted. The user's repeated "why am I losing life" suggests cause-and-effect is the gap, not the rule itself.
• Add a persistent tutorial replay option from the main menu or pause menu. The user explicitly asked for one mid-session.
• Consider a brief, always-visible legend or HUD hint for the first few runs, summarizing the three core rules (lock with space, hit yellow for life, no overlaps). The user retained the space-bar rule but not the others, which suggests the tutorial alone is doing too much work.
• Surface the unlock threshold during play, not only on the level-select screen. A "next unlock: 500" line near the score would let the user calibrate their effort. They came close to 500 without knowing the target existed.
• Consider staging the rules across the first few placements rather than front-loading them all in one tutorial pass. The overlap rule was the only one the user discovered through play; the life rule and scoring rule were not.

Overall

The session shows a strong core mechanic and a strong audiovisual identity sitting underneath a learnability problem. The user reached a workable understanding of the game by minute six or seven, but most of that time was spent guessing rather than playing. The highest-impact change is in-game feedback at the moment a life is gained or lost, since that single gap drove most of the confusion in the transcript. A tutorial replay option is the cheapest second win.