How can I get better at estimating the time needed for coding tasks?

Honestly, estimating coding tasks can be incredibly difficult, even after years of experience. There are various methods out there, like story point poker or critical-path scheduling, but the truth is, estimating well is part of the challenge of programming. One solid piece of advice is to underpromise and overdeliver—this way, your estimates don’t put pressure on others relying on your timelines. Also, don't make your estimate the shortest time that could barely cover the work; try to be realistic. If you haven't yet, check out "The Mythical Man Month" for some great insights.

You can only really estimate tasks that you’ve done before, and even then it's still tricky. It feels a bit like a guessing game, but you'll find that with more projects under your belt, your estimating skills will naturally improve the more experience you gain.

Estimating isn’t a skill that can be easily taught; it really depends on your personal work style and skill level. It's something that you get better at over time. Early in my career, someone advised me to avoid overpromising. It’s much better to give yourself extra time and finish early than to promise something and fall short. It sounds like you might be leaning too much toward giving the most optimistic time frames. Building in some buffer can help, and remember, overestimating doesn’t mean you're being lazy—it's about ensuring you meet deadlines!