27 ms·
Start with a small one, predict how data will turn out and see how close you get. Then go for a bigger one. Disclaimer: This opinion based on (almost) no real e
by ektimo 17y ago
Start with a small one, predict how data will turn out and see how close you get. Then go for a bigger one. Disclaimer: This opinion based on (almost) no real experience.
- ektimo 17y agoAlso: Ask people you know, especially people who will be qualified/candid enough to tell you what is wrong. Asking HN was a clever idea! Asking people in person is also important. I think this is an important question. Anyone else have suggestions?