4 ms·
Fun little game, and I actually quite like the look of it. For something so minimal, it has a nice feel to it. A couple of thoughts/ideas: 1. It seems like th
by hyperpape 7y ago
Fun little game, and I actually quite like the look of it. For something so minimal, it has a nice feel to it.
A couple of thoughts/ideas:
1. It seems like there's not much of an incentive to have more than 3-4 employees. You can just keep stacking upgrades & self-starter or observation training on one of each, and it seems like it works just as well as spreading them out (hiring another employee gives you a small benefit without upgrades, but not much of one, and has a big downside in needing to click more).
2. The distinction between self-starting and observation is confusing (I know others have said it, but I want to repeat that).
3. Workers completing tasks overrides the description for items in the store, which makes it a pain to read.
4. You can't see how many items are in a given swim lane.
5. Not being able to apply upgrades to an employee while they're working on something is a little inconvenient, especially with the little gap that self-starter has.
6. The emoji for the workers made me think I had unhappy workers and I needed to fix them. That was a little frustrating.
Lastly, a big idea that might or might not be fun: there's no penalty for accumulating a big backlog of work (it's not perfectly optimal, because there's opportunity cost, but you still get the payoff). If projects got stale when they piled up in one column, that could be an interesting dynamic.
- LeonB 7y agoThis is a really good list of feedback hyperpape.
- hyperpape 7y agoThanks. I actually walked into my office Monday morning a few weeks ago, and told everyone "I've got it! I'm going to make an idle game about running a development team!" so I've spent a little while thinking about various mechanics.
- workingpatrick 7y agoI agree about having fewer workers stacked with upgrades seems to be the "meta" but I found that the upgrades don't seem to stack linearly. For example, one Dev with 51x mechanical keyboards is not significantly faster than 2x Devs with 1x mech. keyboards each, and costs orders of magnitude more $.