There's another position that's similar but less gross, which is to have some humility about what you think is important vs. what work your manager/manager's manager think is important. Try to work on what they think is important not (just) because of a cynical desire to game the system and get ahead, but because they have probably more experience and (often a lot more) context than you do.

Of course, this only works if you have managers you feel like you can trust and respect in the first place.

That's definitely true too. There's a lot of developers who spend way more time beautifying code than they actually do driving the business. While sometimes great code leads to great business results, that's not a law of nature. The inverse and converse can certainly be true too.

Spot on. Most organizations spend time on how to do thinks than actually doing it. What i found out was to get some data as soon as possible, even if it has large caveats. That is more important than getting it right, becaude you wont.

But the author's manager gave him very high ratings.

Great point. Developers are usually very smart folks, and we like to believe that we know everything. There's no substitute for context.

I mean, there's a balance you can definitely reach. In alot of organizations, depending on the process, playing some Machiavellian game in order to try to get promoted the quickest is going to backfire.

But that doesn't mean you should be consistently selfless all the time. If you think the projects you are working on aren't going to get you promoted, discuss it with your manager. Tell them some of your thoughts, and if they aren't amenable to giving you other work, consider moving to a different part of the organization.

To add on to this, many jobs won't have good metrics in place, so promotion will depend on emotion and cognitive biases. Learning how to work these two factors is crucial.

If you have any insight into this, would you mind sharing?

Is this bad? If you are working on something without any metrics you should be skeptical, regardless of promotion systems.

You can do half and half. You don't have to game everything to make sure it boosts your metrics, but you also have to make sure you're not doing nothing that boosts your metrics.