Harbour Line, one year on
What we got right, what we got wrong, and the update we pulled two days before release.

Harbour Line has been out for a year. It has been played by about nine hundred thousand people, which is more than we planned for and less than the projection someone made in a spreadsheet, and both of those facts are fine.
What we got right: no timers. We were told repeatedly that a management game without timers would not retain, and it retained better than Tessera did in its first year. People come back to Harbour Line because they left something unfinished, not because a bar filled up.
What we got wrong: the tutorial. It was eleven minutes long and it taught the tide system third. The tide is the whole game. We rewrote it twice and the current version teaches the tide in the first ninety seconds, which we should have done at the start.
The update we pulled: two days before the 1.4 release we cut a cargo-priority feature that worked correctly and made the game worse. It removed the moment where you realise you have berthed the ships in the wrong order. That moment is the game. It is still in the branch and it is still not shipping.