I am a firm believer in the "Cluster Theory of Life". Occurences, including winners in your latest betting system come in clusters.
There have been three shark attacks in as many days, and all of a sudden there is mass hysteria. A check of the stats shows that overall there are no more or less shark attacks than previous, just a cluster of them.
Which for system bettors is the reason you should have confidence in your system. If you have tested it for a long enough period of time, you should have confidence that you know what its strike rate is.
If it is 30%, you know that it may go as high as 40% or as low as 20% on occasions but it will always return to the 30% and you can adjust accordingly.

Darky I have also read plenty of Malcolm Knowles stuff but the I think the "Power of Ten" fails because it expects average dividends to perform consistently as strike rates for winners do.
Strike rates will always be consistent over time (ie: win % of favs) but I don't think the same can be said for average dividends, they can go right off.

Jeez I better get off before I bore people.