by Warning: This post is an extremely mathematical one.
I often think of technique as an optimization problem. There are hundreds of different parameters varying from body position to breathing pattern and the goal is to find a combination of these parameters that maximizes our efficiency. When solving an optimization problem it is important to try your best to avoid a local maximum and instead try to find the global maximum. Simulated Annealing is one approach to this.
This is the best analogy I have for Simulated Annealing.
Imagine being on a large mountain range (with many hills), and you are trying to find the highest peak. However it is very foggy and you can only see one step ahead in every direction. The simple strategy to finding the peak is to just keep taking the neighboring step that takes you higher until there is none left. However the problem is that you may get stuck on what is called a local maximum. It is a peak but not the highest peak. Simulated Annealing deals with this by occasionally taking steps that takes you to a lower spot (in hopes that you will reach an even higher peak later). How often does it take a lower step? Simulated Annealing uses a measure called temperature to determine this. The higher the temperature the more likely it is to take a step that goes lower.
The temperature changed as the optimization goes on. In most Simulated Annealing algorithms, the temperature starts very high and slowly decreases as time goes on. This is called a Cooling Schedule.
Optimizing technique in swimming or any discipline involving repetitive movement patterns should be viewed the same way. Start the season with a higher temperature and slowly decrease it as you get closer to championship meet.
So I encourage everyone to increase the temperature and make a jump down once in a while. The worst that can happen is you find out the new “hill of technique” you are on is not as optimal as the previous one. In that case you can always revert back to the old technique or try another hill.