I do recommend playing around with it, dragging handles and seeing if you notice the changes. There's a simple logic to it, they are laid out from the lowest frequencies to the highest.

If you know anything about music, if you ever played an instrument like piano or guitar, it might also be helpful to compare to the frequencies of notes. The lowest note on a standard 88-key piano is 27.5 Hz, and the highest ~ 4 kHz. Middle C is ~ 260 Hz. An octave above that, 520 Hz is in the range of a male tenor, and two octaves above middle C is ~ 1kHz, which might be important for female vocals. Over 2kHz and you start getting to frequencies that are more important for harmonics, and things like cymbals. You probably can't hear over ~ 15 kHz.

So it probably won't sound good to randomly swing one slider up and the next down. But if you think the bass end is a little weak, you might want to boost 30Hz-80Hz by a few decibels, and then taper down to 0 boost by ~ 120 Hz. Or if you think vocals sound a little recessed on your headphone, you might try boosting somewhere in the midrange (200Hz to 2kHz). It might be best not to adjust anything by more than around 5 decibels to start (well other than to play around). That should be enough to hear some changes.

In any case, once you understand how these curves work with EQ, you will then have a better understanding of headphone frequency response curves. They are measuring the same thing.

If there are presets available, they can be a good place to start. If someone has made a tuning intended to fit the Harman curve, that should sound good to nearly 2/3 of people. But to me, using someone else's EQ settings can be a bit like using someone else's eyeglasses. Your ears aren't the same size and shape as everyone else, so you may hear a different frequency response then someone else's measurement suggests.