Game designer & developer @Toonminator . Producer @RoshkaStudios . Hole and Fill 24M, Junkyard 3.9M, Push Titans. Sharing game dev notes & resources.

Joined June 2026
5 Photos and videos
Victor retweeted

14
8
151
10,940
I spent the whole weekend trying to add a third comparison to my second test, this time using a local model, Qwen 3.6… but it has been a pain in the ass...
First test with Bezi, Codex and Claude (these 2 with Unity MCP server by AnkleBreaker Studio) for Unity prototyping medium.com/@vcorva/testing-a…
1
4
I'm running this on my machine (mbpro m4 pro 48gb ram, 2tb, 14c)... I tried with Claude Code but at some point it just stops responding...
1
2
Closing my Sunday with VSCode/Cline. So far, it’s making some progress, but I’ve had to use a lot of “continue” prompts and ran into a few endless loops like the one in the video. Let’s see what it delivers at the end… (if there is one 🥵)
1
Just finished a second test of Unity Prototyping with Claude, Codex and BEZI. Similar results as the first one ( medium.com/@vcorva/testing-a… ), Claude straightforward but expensive, Codex cheaper but needed extra prompts, BEZI worked! similar experience to Codex
Victor retweeted
Recently had the opportunity to build 3 demos to demonstrate the power of game feel! #gamedevelopment
318
3,182
36,804
5,102,174
Victor retweeted
Never forget this airplane picture. During World War II, the military studied planes that came back from combat. They saw bullet holes all over the wings and tail, so the obvious answer was: “Add more armor where the bullet holes are.” But one statistician noticed something most people missed. These were only the planes that survived. The planes that got hit in the engine or cockpit probably never came back. So the real lesson was not where the visible damage was. The real lesson was where the missing damage was. This is survivorship bias. And developers need to understand this deeply. We only see the indie devs who made money. The SaaS founders who got users. The creators who blew up. The apps that reached the top charts. So we copy what survived. Their tech stack. Their content style. Their launch strategy. Their “simple idea.” But we rarely study the hundreds of similar people who did the same thing and failed silently. That’s why market research matters. That’s why distribution matters. That’s why testing demand matters. Don’t just ask: “What worked for winners?” Also ask: “What killed everyone who didn’t make it back?”
3
1
14
1,042