The search database indexes actual questions (the useful bits) and pages of questions (the useless bits).
Indexing pages of questions is pointless because the URL associated with a page of questions is relative to the current page and not an absolute address. Let me explain it this way:- Let us say that the question you are interested in happens to be three pages back from this one. Its URL will be
http://www.theanswerbank.co.uk/AB-Suggestions/ Questions4.html
Let us also assume that the software indexes this page at this time and stores the URL given above.
As more and more questions are added to AB Suggestions, the question you were interested in will get pushed further back through pges 5, 6, 7, etc. Unfortunately, the search index does not follow these changes and continues to point to page 4 (3 pages back from this one). When you click on the link in the search results, you will be taken to the CURRENT page 4 which will most likely fail to meet your requirements and is why you think it is giving you random questions.
The AB Techie-boys haven't realised this yet, or if they have, they have failed to address the problem.