This is a good question. I have been searching for a stable strategy for 4 years, and still haven't found one. Most strategies are divided into scalping (short-term frequent trading with small profit goals) and the rest (mid- to long-term less frequent trading). There are many scalpers on the market. Most of them use pairs like EURGBP, EURCHF or EURUSD and trade during low-activity hours (e.g. asian sessions). The second place in the 2008 mql4 championship was won by a scalper. MQL championships have rules against scalpers, but with a fixed spread of 2-3 pips and allowing to open 15 lots during low-liquidity hours, these rules don't work. Try in real-life to send an order to buy 15 lots of EURGBP during night hours. So, I don't know how organizers are going to deal with the low liquidity of the real market during scalping hours. They need to have some rules disallowing placing large orders during low liquidity.
The mid and long term strategies are based on "if-then" decision trees or intelligent market models such as neural networks. These strategies are hard to design and then train. You can search the code database for perceptron based EA's as an example. The 2008 winner used a decision tree, that was trained by a RIPPER algorithm.
I feel that this year the championship will be flooded by Martingale strategies. It seems to be a fad. Check Ilan in the codebase as an example of a strategy using Martingale. There are many followers of Ilan. Good luck!