For studios making co-op and party games

You will spend a year of dev time
on one of your ideas.

The wrong one costs you the year. Finding out which is right costs $2.99, and the first two runs are free.

Show it a future hit and a future flop and it puts the hit first 82% of the time.

↓ weakest in their genre, by us strongest, by us ↓

266 games launched after we froze the model, in the order we ranked them. A bar is seven games, its height their reviews. Highlighted: one of the seven took off.

Does it work?

We froze the prediction model in June.
Then 44 co-op job sims launched.

Ranked before a single review existed.

5of our top ten cleared the bar
against
0of our bottom ten
And it was not luck

All 266 games of that test, lined up by the place we gave them

A bar is one game, its height its reviews. The highlighted ones cleared the bar.

← we ranked it weakest in its genre we ranked it strongest →
What that is worth to you

Regular pick vs our pick

You have ten ideas and time to build one. How often is the one you build a hit?

You pick one of the ten yourself
8out of 100 became hits
You build the one we rank first
26out of 100 became hits

And if you can afford more than one shot

a random one of ten
8
the one we rank first
26
either of our top two
34
any of our top three
39

Hits per 100 choices.
Measured on 6,000 batches of ten released co-op games; every game scored by a model trained without it.

What lands on your screen

The ranking, and why

1How a run works
1–10ideas you paste in
4images for each
1,900scored games we compare yours against
1ranked list, in 10 minutes
2A real run, start to finish
Eight co-op job sims launched this summer. We ran all eight before any had a review.
Open a row for the axes and the released games on the same score.
Click a name to open it on Steam.
We rate an idea 0–100 inside its genre.
100 – we put it above every released game there.
0 – below all of them.
Idea map our score, 0–100 · what actually happened
Price

Pay in tokens. Start with two free runs.

A run of up to ten ideas is 80. A deep run of one to three ideas is 500. Bought tokens never expire.

Free
2 runs

One idea each, against nine released games of its genre. Shows how high the idea can get.

Start free
Tokens
100
$2.99

One run of up to ten ideas.

Buy
Tokens
800
$6.99

A deep run and three runs: where each idea stands, how high it can get and what to change first.

Buy
Best value
Subscription
8,000
$29 a month

16 deep runs or 100 runs a month, for the weeks you keep reworking ideas. Cancel any time.

Subscribe
Publishers and funds. Volume, team access and a shared portfolio of the pitches you review.
Request access
Tokens back if we cannot separate your ideas. If your ten land inside the band where we are a coin toss, the tokens come back.
Before you pay

Four things we cannot do

Measured, not guessed. If one of them is what you came for, keep your money.

Rate borrowed art

Pick a look from our gallery and you are scored on that studio's execution, not yours: accuracy drops by 11 points and the score flatters you. It will move once you put your own images in.

Predict exact sales or revenue

We predict attention in the first 90 days, counted in reviews. Not copies, not dollars. Free games collect far more reviews per dollar earned, so compare your ideas at the same price.

Split two decent ideas

Between games of 100 to 500 reviews we are right 47% of the time, a coin toss. We separate a likely hit from a likely flop, not a good idea from a slightly better one.

Triple your odds exactly

The 8-against-26 figure compares us to picking blindly among released games. Your own shortlist is already filtered by your taste, so your real gain is smaller.

Score one idea free

Two runs free. Run a game you already know succeeded or failed, or one of your ideas, and see how high it can get.

.foot .foot-links{display:flex;gap:18px;margin-top:18px;font-size:13px} .foot .foot-links a{color:#7e7e88;text-decoration:none} .foot .foot-links a:hover{color:#f2f2f3}