I have calculated a running win/lose ratio. It is actually winners as a proportion of total trades. Unfortunately IG Index dates the trades at the time they were closed, not when they were opened; as winners tend to run for longer than losers, this means that there is a positive bias to later data points (this is why I haven’t calculated a 10-trade moving ratio, for example, because it wouldn’t give a fair picture of what was going on — the chart shows the ratio since inception at each point). The chart shows my win/lose ratio to be stuck around 20%. I need to work on this — it should be 50% or higher.

So why haven’t I lost a lot of money? Well, I have run my winners and cut my losers. The following chart shows a running ratio of the average gain on winners to the average loss on losers. This isn’t too bad.

If 20% of trades make money and the average winner is 3x the size of the average loser, then the expected return from trading is negative. 
Update: Of course there is something of a trade-off here. If one holds trades that seem to be winning for longer, there will be many trades that move one risk unit in one’s favour before reversing and hitting the stop. But I can’t see how it is possible to wait for big profits if one is inclined to take small ones. This means that what I am trying to do is to create a system where 50% of entry signals herald decent moves.