I've seen this kind of thing cause confusion so many times that I decided it was time to create a quick little project about it. See instructions on splash screen, then click green flag to start. There are lots of related little examples and gotchas I could add that have been brought up in the forums, but I don't have time to do anything more than this right now - maybe someone could remix and add something about some of these...? http://scratch.mit.edu/discuss/topic/54508/ http://scratch.mit.edu/discuss/topic/39873/ http://scratch.mit.edu/discuss/topic/51948/ http://scratch.mit.edu/discuss/topic/52426/ http://scratch.mit.edu/discuss/topic/26538/ http://scratch.mit.edu/discuss/topic/2931/ http://scratch.mit.edu/discuss/topic/52296/ Bottom line: once you start using floating-point numbers (i.e. not whole integer numbers) then you will probably start to get values that are not exactly what you're expecting. That means you need to consider how you deal with that when you write scripts that check and compare values...
