You might not know this, but Jennie and I don't just build Game Oracle. We have our own game development ambitions and we're currently working on our first full commercial project Pixel Pyros. We think this is important because, fundamentally, we're building Game Oracle to help with the very problems we face with our own game development: how do we find a good idea? How do we know what players want? How do we know if this feature is a good addition and not just feature creep?
All these questions we plan to answer with our Game Oracle platform and our philosophy is, if we can answer these questions in a useful way for our own projects then they will help other game developers too.
So today I'm back again with another round of exciting updates for Game Oracle, all driven by our own experience using the platform on a real project. This new version ships with:
Player Insights: a new tool in your workspace that allows you to "talk" with player reviews, ask design questions, and have all your project context gathered to inform your design and help plan features
Clone Radar: I know there is a lot of anxiety online right now around AI clones so I've added a tool to the Steam Map to help see where clones could be occurring and how prevalent they really are
Workboard improvements: there was some very annoying gaps in the workboard we've closed
Expanded MCP Tools: you can now read and write to your workspace directly from your agent using the MCP server, meaning all your market research, designs, and project management can sit alongside your local development stack
Bug fixes...of course
Player Insights: ask design questions and get answered informed by real player review data
Reading reviews of comparable games is one of the most useful things you can do when working out what to build. Players will tell you what they love, what frustrates them, and what they wish the game did differently. The difficult part is turning all that feedback into something useful for your own project.
Player Insights gives you a place to do that inside your workspace. Start with a saved Data Table or Competitor Radar, choose the games you want to learn from, and open a conversation about their player reviews. You get an initial overview of recurring positives and negatives, then you can ask your own design questions.
For example:
- What makes gathering and crafting feel satisfying, and when does it become a chore?
- What do players find frustrating about the controls or interface?
- Which complaints would matter most for the kind of game we're making?
The answers include findings with supporting review excerpts, so you can inspect what players actually said. You can also see the strength of the evidence behind the answer.
A live review-backed answer in a demonstration project, framed around a cosy co-op game vision
This is where your project context becomes useful. Player Insights can use your project's Game Vision and Draft GDD, along with its description or linked concept, to frame its suggestions. If you're making a relaxed co-op game, advice about punishing survival mechanics needs to be considered in that context. Something can be popular in a comparable game and still be wrong for yours.
And when a finding does suggest something worth building, you can turn it into a workboard ticket. The draft carries the finding and supporting quotes with it, plus a link back to the conversation. You review it, choose the board, and add it to your backlog.
Keep the reason for a feature attached to the work, so you can come back to it when priorities change.
That last part matters to me. It's easy to fill a backlog with ideas and then forget why any of them seemed important. I want us to be able to ask, "What problem were we trying to solve here?" and still have the answer.
Player Insights is included with Individual and Business subscriptions, with a monthly question allowance for each person.
Clone Radar: a closer look at crowded parts of Steam
There is plenty of discussion about AI clones and waves of very similar games appearing on Steam. As a developer, it's understandable to wonder whether the space you're considering is suddenly becoming crowded.
Clone Radar adds a **Clones** tab to the Steam Map. It highlights detected bursts of near-identical new Steam pages, lets you narrow the results to an area of the map, and opens up the listings behind each hotspot. You can inspect the games, how many developers are represented, and the period in which those pages appeared.
Related waves are grouped into families, so you can explore a broader pattern without having to piece it together from separate cards. The tool also distinguishes clone waves from page factories, making those different patterns easier to browse.
Explore the detected patterns, then open the underlying listings to judge the similarities yourself.
There is an important distinction here: similar store pages do not prove that one developer copied another, or that anyone used AI. Clone Radar shows patterns in listings. It gives you something concrete to investigate, but it cannot tell you how a game was made.
Those signals also appear where you're making decisions. Game Gap can warn you about segments affected by a clone wave and lets you filter those segments out. Concept Compass can flag when your concept or its competitors overlap with a detected wave, with a link to explore it.
For me, the useful question is whether there is something happening in a particular part of the market that we should understand before committing months of development to it.
Workboards: we've tried to make it easy to use
Using the workboard for a real project has been a very effective way of finding all the little things that get annoying after the tenth time you do them.
In v7.2, you can select several tickets and move them together. Drag a selection around cards, use the selection shortcuts, or select from a card's menu. Moving a group of related tasks into the next stage should now be much less fiddly.
You can also change a ticket's colour directly from its card menu, or colour a whole selection at once. It makes grouping related work much quicker when you're looking over the board.
Select related tickets and apply a colour together, without opening each ticket's editor.
There's now an archive for tickets you want out of the way but may need again. Archive from a card, a selection, or by dragging into the archive lane, then search and restore tickets later. Long columns and backlogs also show a shorter preview, with the option to expand them when you need the rest.
These are small improvements individually, but they make a difference when the board is somewhere you spend time every day.
Your workspace, available to your development agent
The MCP server already lets compatible assistants and agents query Game Oracle's market research. This update brings much more of your workspace into that workflow.
With the appropriate access, an agent can browse your projects, read design documents and saved research, inspect workboard tickets, and create or update supported workspace content. It can also read stored project files and upload supported images and whiteboard content.
A particularly useful addition is document editing by section. You can ask your agent to update a specific part of your design document, append a decision, or replace a piece of text, while keeping the rest of the document intact.
Imagine starting a development session with a request like this:
"Read the game vision and the ticket I'm working on. Help me plan the implementation within that scope. Once we've agreed the approach, add the decision to the relevant section of the design document and update the ticket."
That's the kind of connection I want between Game Oracle and our work on Pixel Pyros. The research and the plan should be available while we're building, and decisions made during development should be easy to carry back into the workspace.
Workspace access through MCP is part of the Business tier. Your key needs the relevant read and write permissions, and it can only access projects you own or that are shared with you. If you're using an older key, you will need to issue a new key with the workspace permissions before trying the new tools.
The [MCP setup guide](https://www.game-oracle.com/help/mcp-server) explains how to connect a compatible client.
And the usual collection of bug fixes
Alongside the bigger additions, we've tidied up a few things that were getting in the way:
- Game Gap browsing: better loading and prefetching as you scroll, so the list doesn't look finished when more results are still on their way. - Concept Compass feature search: search the full feature catalogue directly from the feature deck builder. - Project setup messages: progress messages stay tied to the relevant concept, and completed notices clear themselves away. - Workspace stability: fixes for backend database issues and an editor error when leaving a document. - Blog pages: empty video embeds no longer leave a dead player behind, and the introductory drop cap stays out of list items.
Business teams can also access each other's Concept Compass concepts, making it easier to work from the same research.
Back to building
Working on Pixel Pyros keeps giving us reasons to improve Game Oracle. We need to understand our audience, make choices about scope, and turn those choices into work we can actually finish. I suspect those problems sound familiar to quite a few of you.
This release brings those steps closer together: investigate a question, check the evidence, decide what fits your game, and carry it into development with the context still attached.
I'd love to hear what you try with Player Insights, and which parts of your own research and planning workflow still feel more difficult than they should. We'll be putting these tools to work on Pixel Pyros as we keep building both projects.
Our Products
Dive into the data yourself with our research tools for unrestricted market research.
-Our Steam Map allows you to find and analyze games with
similar art style and mechanics
-Analyze uniqueness and performance in one glance with
our bespoke statistical scores
-Validate your game ideas and develop realistic targets
with our forecasting tools