ACCURACY AND PRECISION IN MATKA
Accuracy is the percentage of “Pass” games for every 100 Attempts of Play or Testrun of a strategy / method. If your game passes any 75 days in 100 days, your accuracy is 75%. All rules, methods, protocols, strategies revolve around one and only one concept : Accuracy. Accuracy defines and validates a method/protocol/strategy. Higher the accuracy, the better chances of making money.
The more accurate your game is, the more money you make. Simple as that.
Always target higher accuracy rules / methods or protocols even if they yield more picks and work downwards combining them and testing them to reduce more picks. This type of development proved to be successful many times for you.
The Accuracy of Online Guessers Explained
Almost all guessers you find on any other websites or Facebook do not care about Accuracy much. It is why most of the time, your experience with a guesser does not end up satisfactorily or in profit. You get into a lot of loss, just because most guessers don’t measure and care about accuracy. A more accurate method can make you a millionaire even if you play 6 singles open to close. That’s the power of Accuracy. When you contact / want to subscribe a Guesser, follow their game and calculate accuracy.
The Problem : The Anomaly of Accuracy based on Averages
During our earlier years of working on methods and strategies, we made a mistake, and it took atleast 2 years to figure it out. Everything looks normal during test runs and everything goes awry the instant we take it for practical use.
When we found out what was going wrong….we had to laugh at ourselves.
At first, we used to measure accuracy based on large sets of data. A rule when tested, if yielded 80% accuracy, it is considered. We developed hundreds of rules which are 80% accurate over large sets of data. Kalyan’s test set was over 11000 days of data (1974 – 2014)
For Example, if a method/rule assessed to 80% accuracy, it didnt mean it was passing almost all the time with short breaks. There we long runs of fails and long runs of passes. Meaning that if a 50% accurate method/rule is taken into account, it didn’t perform as one would expect : Almost every alternate day. Instead, it fails many days consecutively somewhere and passes many days consecutively adjusting accuracy average. If a rule failed 50 days consecutively in a 100 day sample and passes 50 days , it would still amount to 50% accurate. And the problem, is , The rule can fail alternate days and still is valuable but when it fails consecutively for many days, you’d go ruin. Most Continuously failing rules made it into our deployment rules and we got irrevocably fucked up. 😥 We found it out when we manually looked at the readings data week by week.
The Solution : Segmental Accuracy
To solve this problem that mislead us for over two years, we came up with a concept called Segmental Accuracy. Segmental accuracy is explained in detailed with practical examples, in the Professional section of Learn Matka article base !