In another thread, Whit Haydn made a comment that gives me enough "Food for thought" that I'd like to discuss it separately, rather than hijacking that thread.

The quote from Whit:


So, top-down or bottom-up. Which do you prefer? Why? What do you see as the strenghs and weaknesses of each approach?

Geoff

I do both, and I'd assumed that so do most other creative magicians.

Often I have a vision as to what I'd like to happen, the story I'd like to tell and try to figure out how to make it happen, but this can be as much of an intellectual exercise as much as anything else, as it neccesarily involves a long period of working out, testing and rebuilding before it yields anything resembling a finished product and may not even be something I'm practically able to realise yet anyway. As such I've only a few performable routines that fit that category and those are still being developed, but then I haven't been in magic nearly long enough to expect to do everything in my notebook.
So I will admit that my best completed routines so far have come from dressing up existing tricks in new presentations to suit my character, in which I can concentrate entirely on the theatrical aspect. But ask me again when I've been in magic long enough to finish some of the long term projects I have in my notebook and "experimental junk" cupboard and I may give a different answer.

I like to think of an effect that makes me go WOW!, then work out why I went WOW! and try to create this for other people.

Saxmangeoff,

I had posted the following in that topic, to which Whit responded with the quotation you posted.

Perhaps that's not the greatest impediment to the advancement of the art of magic, but in my opinion it is much more of an impediment than what was proposed in that topic, which was that card magic was holding other magic back!

I do both type of design too, but I start with bottom-up design much more often.

Pure bottom up design is typically over-constrained by effects. Imagine writing down a totally random list of tricks and then trying to use each one in a routine. It would likely not flow very well. First a card changes color, then I produce a coin in my hand, then I produce a dove, then I change the card to a different card, then a dog is produces from a production box. I’m sure all that could be forced into a routine, but it would be pretty contrived (Now that I’ve written that, someone will come up with a brilliant presentation with all that.)

Pure top-down design is typically over-constrained by the presentation. A presentation is created without any concern for whether the effects can be achieved. Thus it’s likely they won't be achievable.

A magicians who starts with a bottom-up design will eventually revert to top-down design so that they create a coherent routine. There is usually a bottom-up and top-down iteration many times. I believe this is the chiseling process to which Whit refers. (And of course, over time as new ideas are realized and feedback is obtained, a routine undergoes additional modifications).

Pure top-down design offers exciting challenges, but successes will be rare.

Surely there must be an objective that needs to worked towards?

Partizan, you're right. The purpose of a "design" is to achieve a specific objective. In this case, the objective is to create something that entertains an audience using one or more seemingly magical effects.