The most important thing about creating a backup strategy is to consider how you will be using your backup. If you are using it simply for disaster recovery purposes, then you're OK with most of the strategies listed here. If you're going to need the ability to recover a file (or database, or whatever) from the state that it was in on any given date, you're going to need a rotation that doesn't involve overwriting data. This is not that uncommon in certain highly regulated fields (finance, healthcare, law, etc).
My system is a 28 day rotation. We do full backups on Monday night, and incremental backups Tuesday-Sunday. Each full backup is done on a fresh tape that is sent to secure storage, so at the end of the year we will have 52 full tapes. Our incremental tapes are divided into Week 1 thru Week 4. After week 4 we start again with Week 1. The incremental backups are small enough that we can get roughly 6 of them on a tape. Once we have filled a set of incrementals (which are actually 4 weeks worth of tapes) they are sent off to secure storage with the full backups and we start a fresh group of incrementals.
It is fairly convoluted (and expensive), but if someone tells me "I need to see file x from June 12th of last year" I can have the tapes pulled and produce the file. If you're overwriting incremental backups as suggested above, you'll only be able to get to the week or month, not the day. That's fine for most applications (like disaster recovery), but depending on your business needs it may be a problem. So make sure that you and management are both clear on exacly what your backup system needs to accomplish, and then design a system that fulfills those needs.