To Buy Or Not To Buy ….
Take a look at these actual results from a systematic method of selection;
Time…1970-1989 inc
3621 Bets
1340 Winners
2281 Losers
37% Strike rate
Winning Months 69.2%
Profit on Turnover 61.63%
What are your initial thoughts? This is a system with pre-defined rules and are results in action with no back-fitting of rules. You have a large sample size, a solid strike-rate, exceptional profitability and consistency. Would you purchase this system, or the selections provided by this system if it were offered to you for the next few years?
Congratulations are in order, if you answered “Yes” to purchasing the selections….here are the results for the next 10 years…..
Time…1990-1999 inc
2977 Bets
995 Winners
1982 Losers
33.4% Strike Rate
Winning Months 60%
Profit on Turnover 30%
Now what do you think? Even these results for the past 10 years, although below those of the 70’s and 80’s, are excellent. A large sample size, good strike-rate and excellent profitability and consistency. And now you have 30 years of results covering almost 6600 bets! If you didn’t buy the system in 1989…would you buy it now in 1999? How much more evidence do you need?
Let’s have a look at the results from 2000 to today…..
2731 Bets
713 Winners
2018 Losers
26.1% Strike Rate
Winning Months 52.3%
Profit on Turnover 6.21%
So, how can this occur? Still profitable, but now at a marginal level, and it would take someone with a steely psychological make-up to continue with such a methodology, given the previous performance levels.
Firstly, a confession, although the results you have read are genuine, they are systematic bets on financial markets rather than racehorses, but the same principles apply.
Sample Size is NEVER large enough. Forget what the Mathematicians and Statisticians tell you. Because conditions are ever changing, in the financial world there are oil crisis, war, inflation, credit crunch, in the racing world there are training methods, course changes, betting market changes, more information etc….
So, toss aside the system? That depends if circumstances/conditions have changed to cause the disruption to results? Or, if we could look at 100 years of results would 2000-2012 appear as a statistical blip? Clearly we cannot know for sure, but these are things to bear in mind when you are researching methodologies to bet upon. Something that can look incredibly alluring can often be a mirage. More important than results are the rules behind the results.
And the most important statistical truism of all….”your biggest drawdown (losing run) is always in the future…not the past”