Example of Pruning:

As an example, we are assume that the four choices for player one, there we are looking only two moves ahead 1 for player one and 1 for player two such as player are known as moves M1, M2, M3 and M4. In fact if we do a depth first search for player one's move, we can work out the score they are guaranteed just for M1 before even bearing in mind that move M2. Think that it turns out that player one is guaranteed to score 10 with move M1. We can just need this information to discard move M2 without checking all the potential for player two's move. Just for this instance, visualize that the first choice much possible for player two after M2 from player one technique that player one will score only 5 overall. In such for this case, we just know that the maximum player one can score with M2 is 5 or less. Naturally, player one won't choose this, is means that M1 will score 10 for them. We see that there's no point of to checking all the other potential of possibilities for M2. Whenever this can be seen in the following diagram (ignore the X's and N's for the time being):

We that we could reject M2 straight away, because saving ourselves 3 nodes in the search space. We could reject M3 after we came across the 9, and in the end M4 turns out to be better than M1 for player one. In total, by proper using of alpha-beta pruning, we could say that we avoided looking at 5 end nodes out of 16 - around 30%. If there is the calculation to assess the scores at end-game states (or estimate them with an evaluation function) is computationally expensive, so this type of saving could enable a much more larger search. However, this kind of pruning can occur anywhere on the tree.