For newer systems/file system based systems you tend not to have simple memory pointers (
http://hitmen.c02.at/files/yagcd/yagcd/frames.html ) unless the text is part of a binary. This can make some of the guides for older systems appear not as applicable, and while they technically still are I would agree to a statement of it not being an ideal teaching case for newer systems.
To that end you get to figure out what the devs used for that game, though you might be lucky and find the same format used for the sequels/other ports.
I usually start by finding the ends of text (hopefully it is a nice simple to search for value) for right after text finishes a new one starts and noting down the location. A good search in a hex editor will do this. You can always manually scan through the text as well looking at locations.
Put this list of ending values (or ending + however many to get to the next segment) next to a column or two of numbers from the start (or maybe end) of the text and you get to figure out the pattern that unites them.
You tend to get either pointers that start counting at the start of the file (conventional pointers), at the start of the text section (offset pointers) or take the address of the pointer within the file and add the contents of the pointer to it (relative pointers)
Sometimes the stuff at the start (or maybe in a similarly named file or end of the file) will be a list of locations, sometimes it will be a list of locations and lengths, sometimes it will have formatting data, sometimes it will just be a list of lengths, sometimes the text sections themselves will have a value at the start of them saying how long they are and there are probably some others.
Personally I like to play with the numbers -- take one from the next, see how far out the numbers I have are from what I expect them to be... you rarely get some kind of super convoluted difference between pointers and the eventual values.