Here's my post from back in 2011 on the MA crosses - note, I have not updated the statistics but it doesn't really matter, as you can see the MA cross system with the whipsaw confirmation method has far superior statistics - note I've attached two chart examples showing how the whipsaw confirmation method works, read the text and see the examples on the charts.

However even still, I would still not trade such a system, for long term stuff like that my SPX 401K system from 1961 blows any MA cross system away as does my BPT MA on a month SPX chart. 

Regarding 50/200 Day EMA Crosses:

Over 20 years here's the back testing results.  Backtesting shows that this moving average system works better with a confirmation candle!  what I mean by that is, you need a confirming candle to validate the cross, this filters out a lot of the whipsaws.

see the first example as to how to use a confirmation candle (shows a confirmation candle using 13/34)

here's two sets of statistics, 1 shows the 50/200 cross with and without the confirming candle

Confirming candle: Profit factor of 8 and 70% winning trades

Non comfirming candle: Profit factor of 5.5 and 43.7% winning trades

I've found that no matter what moving average crossover you use, the confirming candle always makes it better.