Coups are less frequent these days, but more likely to succeed

In other words, last night was an outlier.  Here is Jonathan M. Powell and Clayton L. Thyne in the Journal of Peace Research:

We also see some interesting trends in the frequency of coup attempts over time. As shown in Figure 2, there is a fairly clear decline in the total frequency of coup attempts over time. The high point for coup attempts came in the mid-1960s, followed by two more bubbles in the mid-1970s and the early 1990s. The number of successful coups has likewise decreased over time. We saw 12 successful coups in both 1963 and 1966. The mid- to late-1970s also saw a brief burst of successful coups (ranging from 3 to 9 for each year). An interesting trend emerges when we look at the percentage of coup attempts that resulted in successful regime changes, which we plot on the right side of the Y-axis. The mean success rate is 48% during the entire time span. This rate saw early peaks around 1970 and 1980, and then a decline until the turn of the century. However, we see another spike in the success rate starting in 2003. Twelve of the 18 (67%) coup attempts since then have been successful, and only one of the most recent four coup attempts has failed. While coups have certainly waned over time, the recent success of coup plotters suggests that coups remain a key element of governmental instability.

I cannot readily pull out Figure 2 from the pdf, but it is on p.7 of the document.  Note that their data run up through 2010, and thus do not cover the Arab Spring.