I use a recommended formula I found almost 20 years ago and it is very consistent. It has been used in finance (by my research) by traders and options dealers.
It gives a 50% chance of probability but it seems to be a 99% chance of occurrence
LOG(2/n) / LOG(1-s)
where n is number of races/events/games
where s is the strike as a percentage over 100.
In your example 55 wagers is a very low sample or period of trade. I try to get what the strike would be over 1000. (when researching markets I have +5000 before i go in).
There are 2 issues with your example. You stated the strike rate is 25.45% then you say the average odds are 5.5 (18.18% sr)
So which one is it?
Efficient market theory says the strike rate of an expected outcome will equal its rate of return. So in perfect markets with no abnormal information (drugs, corruption, illegal play, dodgy referees) a $2.0 chance will win 50% of the time yet still experience a run of 9 losing events. This occurs in very efficient markets like EPL.
In your example I would be using 18.18% as my strike rate, as this is your rate of return. As chance of outcome diminish, the efficiency also diminishes where say a $50 chance only wins 1.8 times in every 100, not twice.
Therefore in 1000 races your run of outs will be approximately 31. In 55 the ROO is 17, in 200 it is 23. Favorite punters will incur a ROO of over 20 (in >3500 races), yet few plan for it.
The next question is money management, how much of your bank do your risk per selection/event without tanking?