Silver's model predicts that the four states (Maine, Maryland, Washington and Minnesota) which voted on the side of marriage equality in November 2012 would have done so, including 16 others. More interestingly, the model also has predictions for 2016. If one assumes a linear relationship between the 2012 and 2016 results one can interpolate a 2014 prediction as well.
Note that Silver's predictions for the results in 2012 are greater than the actual results that occurred at the ballot box for Maryland (55.0% predicted versus 52.4% actual; +2.6), Minnesota (52.3% predicted versus 51.2% actual; +1.1), Washington (55.7% predicted versus 53.7% actual; +2.0) but not Maine (51.1% predicted versus 51.5% actual; -0.4).
But his model also has good news for LGBT activists who may want to move forward with a pro-marriage equality measures in California and Oregon in 2014. A linear interpolation would indicate that California's measure would pass with approximately 57.5% of the vote while Oregon's would pass with 56.8%. Oregon's measure is much more likely to actually appear on the 2014 ballot than California's.
Another thing to notice about the list of states is that Rhode Island is at the top of the list, yet that state has not enacted marriage equality yet. Other states, like Illinois and Delaware, which are considering marriage equality legislation this year are also well into the blue territory of majority support for marriage equality.
The future of marriage equality is blue skies ahead!