// how it was made
An IT support shift compressed into eight minutes. Tickets arrive, you set a priority, you get one diagnostic question, and then you have to commit to a fix. I work in IT. This is that job with the paperwork removed and the judgement left in.
Play it →Every ticket runs through three steps. Set the priority, from P1 down to P4. Ask one diagnostic question, and only one, because the other two lock the moment you choose. Then pick a resolution from three options and read what it cost you.
Scoring is blunt on purpose. The right priority is worth 30, the right question 25, the right fix 45, and a missed deadline takes 25 back. That number becomes the ticket's satisfaction score, and the shift average is your grade. The top grade needs 92 percent and zero breaches, so one panicked guess ends a clean run.
Most people triage by counting users. Forty people locked out of a shared drive feels critical, one slow laptop feels minor, and most of the time that instinct is right.
Two tickets exist purely to break that habit. A new hire who cannot reach email on her first morning is one person, but her whole day of training runs through it, so she sits above a normal single user issue. A payroll portal throwing errors affects a team of four, and it is the most critical thing in the building, because the submission cutoff is in two hours and after that nobody gets paid on Friday. Priority is impact multiplied by urgency, not headcount, and that is the one idea I wanted the game to hand somebody who has never worked a queue.
If the bad option is obviously stupid, this stops being a game and becomes a quiz with one plausible answer. So every wrong resolution is something a real service desk does on a bad day: reboot the file server before anyone has confirmed the scope, reinstall a client that works fine on a different network, reset a password that is provably correct, grant access to finance data because a manager said it would be useful.
The feedback never just says wrong. It names the cost. Granting that dashboard is not a mistake, it is an audit finding, because the manager is not the data owner. Replying to a phishing sender to ask them to stop confirms a live mailbox to an attacker.
A new ticket lands every nine to twenty seconds and the queue stops accepting them at seven, which is about where a real morning stops feeling manageable. Each carries its own countdown, drawn as a bar that drains while it sits there. A P1 gives you sixty seconds. A P4 gives you over five minutes.
That makes reading carefully expensive, which is the honest part. The pressure in support work is almost never the difficulty of any single problem. It is that the careful answer and the fast answer are different answers, and the queue keeps growing while you decide. A breach is permanent once it lands, so there is no rescuing the shift by hurrying afterwards.
There are eight tickets, drawn at random, so a single shift can hand you the same one twice and by your third shift you are recalling answers rather than working them out. The judgement game quietly turns into a memory game. The fix is more tickets, not more mechanics, and that is a writing job rather than a coding one.
Plain HTML, CSS and JavaScript on a single page. No build step, no framework, no 3D, no images, no sound. It shares the small profile the other games here use, so counts, grades and badges stay in your own browser.