That would still be a mess, in mostly the same way. All it does is emphasize that the values have a coarse 1-year sampling rate. The real mistake is that the m

That would still be a mess, in mostly the same way. All it does is emphasize that the values have a coarse 1-year sampling rate. The real mistake is that the most rapidly-changing values should have been in the middle, with the slowly-changing values at both top and bottom. This would help align the values from one sample to the next, making them easier to follow across the chart.