Continuing on through 2014...
So I discovered Graham and Greenblatt (about 13 years too late, but I'm a slow learner). Graham advocated buying cheap, cheap cheap. In his time, opportunities to buy net-net cheap seemed more prevalent relative to today. This may be a function of more participants today looking for net-nets compared to during Graham's time. The act of more participants constantly looking has probably arbitraged long standing net-net opportunities away. Huge multi-national institutions employ scores upon scores of teams of analysts, who's only job is to scour the earth looking for net-nets (which aren't cigar butts). <On the flipside, this could actually be an opportunity for nimble individual investors, who, by not being limited by the same investment and capital allocation constraints institutions are subject to, can potentially venture into areas of the market that large institutions cannot due to these same constraints.>
Graham proposed that buying a basket of net-nets would a) provide a margin of safety, and b) insulate an investor from being wrong on any individual net-net.
While Graham fully expected that a handful of net-nets would be cigar butts, he also fully expected that the remaining net-nets that weren't cigar butts would more than compensate the investor for the risk of holding the cigar butts. And his system worked! Why? Because he operated with a huge margin of safety. He was operating under the premise of buying $1's for $.50's, and all he needed to do was sell his accumulated $.50's for $1 once the market properly valued the non cigar butt net-nets. And over time, this is exactly what ended up happening.
Greenblatt's methodology, in my opinion, is a modernized extension of Graham's methodology. Recognizing a dearth in modern-day net-nets, by buying a basket of the highest ranked good and cheap stocks relative to all other stocks, Greenblatt eliminated the need to scour the rubble for net-nets. Relatively good combined with relatively cheap, spread across 30 stocks would theoretically share many of the same characteristics as a basket of 30 net-nets, except that a high return on capital would likely add an additional layer of margin of safety not found in a basket of net-nets who's only redeeming factor was their relative cheapness.
I believe that Buffett (and Munger) 's approach parallels these two approaches in certain ways, but the distinguishing factor with Buffett (and Munger) is that they seem to relax the cheapness constraint in order to amass significant positions in companies that are good now, and have the potential for remaining good for a long time, hence the often reproduced (and misused) quote that it is, "better to buy a wonderful business at a fair price than to buy a fair business at a wonderful price". What Buffett (and Munger) may end up sacrificing in terms of excess return in the short run by not just buying cheap, they more than make up in terms of compounding over multi-year periods.
This has allowed Buffett (and Munger) to buy and hold a concentrated portfolio of companies with indisputable franchises/moats for decades on end, and as a result, their core holdings have become the gold standard by which buy and hold investors benchmark themselves.
If an investor truly wanted to have a dividend growth portfolio run on autopilot, he/she could do a lot worse than simply buying and holding Buffett (and Munger) 's current holdings in whatever proportions he/she was comfortable with.
You can find Berkshire Hathaway's current holdings here: