Simple, curious, yes/no question: Can you put a resistor in backwards?
While the original poor choice of wording "Does it matter if..." lead to the question being closed, there's now at least a window of opportunity to add an answer.
The answer by @Passerby is clear and definitive - currently it says 'Resistors lack polarity. There is no backwards. I think this applies perfectly to the definition of the resistor as a circuit element and should therefore be the accepted answer.
However, the comment by @RussellMcMahan brings up the point that it may "matter" during an optical inspection process (visual or machine-based) as part of circuit board assembly.
I just want to add that the term "resistor" can also apply to the actual components themselves, and while we expect discrete devices to be absolutely non-polar, this might not always be 100% true with absolute certainty. For example, "resistors" fabricated within semiconductor substrates may not actually be adequately described without their full equivalent circuit diagram, which may have non-linear and polar behavior as well as asymmetric RF parasitics. Some may even require a DC bias to operate properly as resistors in AC circuits.
While most semiconductor resistors are embedded well within integrated circuits and an end user does not "put them in" backwards or any other way, a designer could certainly 'drop them in' incorrectly during layout (if they are opening up circuit libraries or possibly turning off design rule enforcement), and if bulk semiconductor resistors (as opposed to thin film silicon resistors) ever did appear as discrete devices it may indeed matter electrically in some way how they are inserted.
So while the question gives the impression we are inserting a carbon resistor into a circuit with the gold band on the left vs on the right, the actual thing which we call a resistor may be quite different in different situations.