Thursday, November 15, 2012

Gaming Models and Presidential Elections

Copyright 2012 J. Lee Lehman, PhD

For many years, I have been interested in mundane methods for predicting the outcome of sporting events. Originally, I worked primarily with William Ramsey's rules. Later, I also worked with Bonatti's rules of warfare. In the latter case, I developed methods for working with presidential elections based on quantitative models: assigning points to each of the criteria used. I believe that this work is vital, because only quantitative models avoid the judgment calls that make historical work truly useful, rather than speculative.

The pitfalls of non-quantitative models was illustrated all too well in Sylvia De Long's 1981 work, Charting Presidential Elections. In this book, she presented a model using Saturn to rule the Democrats, and Jupiter to rule the Republicans. She used elevation and disposition as two of the factors to judge which planet was stronger in the chart for sunrise on Election Day.

The problem? The one advance chart she gave was 1984, in which she predicted that the Democrats would win, instead of 1984 becoming the reelection of Ronald Reagan. Oops. Now - we all will be wrong sooner or later in our predictions. But the point, I think, is that, when you know what the outcome was, it is difficult to be truly objective in rating the charts. Unless you have a friend run the data, then chop the chart information off the print-outs, it's hard to actually retrospectively predict rather than explain.

I made my own predictions for 2012 in August at a talk I delivered to the Northern Illinois Chapter of NCGR. I wanted to share a couple of my slides, to illustrate some ideas about mundane prediction, especially using gaming models.

I would also like to note that the media has given little coverage to the fact that not only were there several of the media statisticians like Nate Silver exceptionally accurate - but so were the betting models. Jamie Partridge has also mentioned this. During the run-up to the election, sites like and were showing Obama winning consistently. Predictwise was likewise very accurate on the results for the US Senate.

So: about the astrology. I presented two different discussions: a comparison of natal charts based on chart comparisons from prior elections where we had good birthdata for each candidate.

For example, comparing the last three outer planet transits before the election gave these results:

This wasn't a stunning call for Obama, but given that Saturn aspects seemed more likely in winners, the fact that his last three aspects were all Saturn did argue for Obama. When I counted up which transiting planet aspected the natal chart, I did not break the data out by transited planet: off-hand, Saturn square Saturn seems harder than most.

Another example was the lunation cycle. I saw a number of predictions based on Romney going into the 8th phase as a negative.

This was a technique that was not predictive in this race, although I saw it used by several astrologers as an argument for Obama. Here;s the issue. This graph shows the actual historical count. Romney was born in the 6th phase, which, according to the graph, has more winners than losers. He had just gone into the 8th phase in September 2012. Given the description of the 8th phase in Rudhyar's work, this makes absolute sense as a call for Romney to lose. In the actual data going into this election, there was only a small elevation of losing associated with this phase.

Contrast this with Obama. Obama's birth phase, the 7th, is not historically good for presidential winners, whereas Romney's birth phase was. Obama's phase at the 2012 election had heretofore only been occupied by losers.

We don't have a lot of data points to work with here, so these results so far are just a start.

We have more data for mundane models, and this is what I will review now. My model is based on the following:

The major question is: what charts work? This is my summary slide of results leading up through 2008.

It was typical in the classical period to examine all outer planet conjunctions. Since the Outer Planets hadn't been discovered yet, that meant the Mars-Saturn and Mars-Jupiter conjunctions, as well as the more famous and more powerful Jupiter-Saturn. Here, the Mars-Jupiter and Mars-Saturn would potentially be interesting, because there is always at least one such conjunction between each election. Thus, one cycle of Mars and Jupiter or Saturn can only apply to one presidential cycle.

This being said, that doesn't guarantee that either of these cycles is predictive. In fact, the Mars-Jupiter does not appear to be. Similarly, the prior solar eclipse and the Inauguration Day chart are not predictive, at least using this particular rules of war model. This does not mean that some other model could not use these charts.

We may further note that the Libra Ingress turns out to be one of the more predictive charts: but in reverse. This is what is called a contrarian model: you take the winner as whatever the model deems the loser.

Examining all these charts together gives the following prediction, based on the historical model:

Going back to 1912, this combined model has been wrong only three times: in 1932, 1952, and 2008. It is interesting that 1932 was the first election after the onset of the Great Depression, and 2008 the beginning of the Great Recession: major economic downturns which clearly were outside the bounds of the model.

My future concern - that Election Day charts may become obsolete as more of the country moves toward early voting - may in fact be borne out by this year's data. The last time both Election Day charts were wrong was 1972: and these two charts have only both been wrong twice in the last century. We shall have to watch this trend.


Bonatti, Guido. The Book of Astronomy by Guido Bonatti. Trans. Benjamin N. Dykes. Golden Valley, MN: Cazimi Press, 2007.

Brady, Bernadette and J. Lee Lehman. Twelfth century castle besiegement in sport. The Astrological Journal 39(3): 27-44, 1998.

DeLong, Sylvia. Charting Presidential Elections. Tempe, AZ: American Federation of Astrologers, 1982.

Ramesey, William. Astrologia Restaurata, or, Astrologie Restored Being an Introduction to the General and Chief Part of the Language of the Stars : In 4 Books ... : With a Table of the Most Material Things Therein Contained. London: Printed for Robert, 1654.

No comments: