Last month I did the least pleasant thing I have done as a developer. I lined up thirteen of my own games and prototypes, built a scoring rubric with no loyalty to me, and graded every one of them like a critic who had never heard my name.
Five cleared the bar. Eight did not. The eight are shelved.
The rubric
The problem with judging your own work is not that you are too kind. It is that your kindness is selective. You forgive the flaw you understand (“the tutorial is rough, but I know what it’s trying to say”) and you overweight the part that was hard to build, because effort feels like quality from the inside. I needed something that would refuse to care how hard anything was.
The rubric scores twelve dimensions — systems depth, moment-to-moment feel, clarity, escalation, audiovisual identity, meaning, and so on — each weighted differently by genre, because a puzzle game and an action roguelite do not owe the player the same things. Every game gets an integrity check on top: no dark patterns, no engagement bait, nothing that makes money by disrespecting the player. That one is pass-fail.
Two rules make it work. The first is that the bar is external. Each game is scored against the best game in its genre, the one it will actually sit next to in a Steam library. My deckbuilder was measured against Balatro, which I put at 4.35 out of 5 on the same rubric. Not against “pretty good for a solo dev.” Nobody browsing Steam grades on effort.
The second rule is that scores get re-derived cold. Weeks later, no notes from the first pass, score it again from scratch. If the two passes disagree by much, the optimistic one is wrong. It is always the optimistic one. Early versions of my scores did not survive this, which is precisely the point. A number you cannot reproduce blind is a mood with a decimal point.
What the numbers did
Wild Crazy 8s went through the gauntlet like everything else. It scored 4.78 against Balatro’s 4.35 bar, and it earned that across multiple cold re-scores, which is the only reason I let the sentence “it stands next to Balatro” appear anywhere near my marketing. When the store page promises fun, that promise has a paper trail.
And there is a receipt for the other side too. The rubric has dimensions it refuses to score from the inside — whether a run feels tense, whether you reach for one more game at midnight. Those get flagged as unknowns until real players answer them, because a rubric that fills in feelings with hope is just flattery with columns.
The eight games that failed were not garbage. A couple of them were one strong system away from clearing. That was the seductive part, and it is exactly the trap the rubric exists to spring: “one system away” is a phrase that has consumed entire years of solo developers’ lives. The honest version of the sentence is “not good enough yet, and not the best use of the next six months.”
Why I am telling you this
Indie marketing is mostly adjectives. Every store page says addictive, polished, deep. I cannot prove my games are fun in a blog post — but I can show you the mechanism that stops me from shipping the ones that are not. The five that cleared, cleared blind, against named competitors, on a rubric that has already killed eight of my own projects. That is what the bar means at this studio.
The first survivor reaches Steam on November 10. It is a deckbuilder about sharing your deck with the dealer, and if this post did its job, you now know exactly how much scrutiny stands behind that wishlist page.