Some games have a settled answer and some do not. Where the boundary sits.
Tic-tac-toe, gomoku and connect four all show the whole board, involve no chance, and have finitely many moves. Games meeting those conditions have a theoretically determined outcome. With best play from both sides, either the first player wins, the second player wins, or it is a draw. Which one differs by game.
The board is small enough to examine every possibility, and the conclusion is that flawless play from both sides ends in a draw. Winning at tic-tac-toe therefore means the opponent played something less than best somewhere. It is also why you rarely beat a computer at it.
Connect four has also been fully analysed and is known to be a first-player win with best play. On larger boards such as gomoku, the first player's advantage is so strong that competitive rules restrict the opening placements to rebalance it. Rules that look complicated usually exist to correct exactly this kind of imbalance.
Chess and go satisfy the same conditions, so their results are determined in theory. The number of possibilities is simply unmanageable, so the answer cannot be obtained. A solution existing and a solution being computable are entirely different problems, which is why these games remain interesting.
There are principles that work without memorising a complete solution: block the square where the opponent wins next, and build a shape where you threaten to win in two directions at once. Faced with two simultaneous threats, the opponent can only block one. Most games turn on who builds that shape first.
Each button runs this keyword on that search engine