A specific look at 52 Hawaiian eruptions since January 1832 shows the same sort of pattern.

"Nearly twice as many eruptions have occurred nearer fortnightly tidal maximum than tidal minimum." Hawaiin Volcano Observatory scientists have noted that the Pu'u 'O'o fountaining episodes each occurred remarkably close to fortnightly tidal maximums and that the first set of eruption pauses in 1990 (periods during which the eruption turned off for up to a few days) occurred remarkably close to fortnightly tidal minimums.

Great! Now let's start predicting eruptions based only on this information. The fortnightly tidal maximum occurs at full and new moons, every 14 or so days. The next tidal maximum will be the new moon on November 15 -- will Mauna Loa erupt then? Almost certainly not.

Although this is a fascinating correlation, there are just too many tidal maximums and too many volcanoes to base predictions on tidal cycle alone. In the Hawai'i example of 52 eruptions since January 1832, there have been nearly 3,900 tidal maximums, of which roughly 3,850 went by without causing an eruption. Statistically, this is about a one percent chance that any tidal maximum will affect the start of an eruption.

The correlation is more important as a clue to how volcanoes work. The effect of the tides suggests that a volcano can remain in a state of near eruption for a period of time before some threshold is exceeded and an eruption starts. There are probably many possible mechanisms for exceeding that threshold -- the lunar tides are but one. (source - Hawaii Volcano Observatory).