I’ve been slow on email responses for the past few weeks because I’ve been working on and testing a new algorithm for the ball deceleration. The idea was to develop an algorithm that suits all conditions and wheels, without being difficult to use. I’ve achieved it now and completed testing last night.
For testing I set up my wheel to be very difficult to beat. With some algorithms the edge was reasonable (+15%), and with other algorithms the edge was only around 7%. With the new algorithm, the results are below:
No Targeting (15-20% edge), Betting 2 Numbers
With Targeting (70-80% edge), Betting 1 Number
Key Points of Algorithm
- Compatible with targeting.
- More accurate learning of complex ball behavior.
- Very quick to learn changes and adapt.
- Compatible with both Uber manual clicks and Hybrid auto timings.
- Compatible with algorithm to determine how early predictions can be, while maintaining accuracy.
There’s more and this algorithm will be the default one used by all players.
Next I’ll prepare the coding specifications for the programmer. This will take about 3 days between other work. I don’t know when he’ll be able to code it. I’m guessing the updated and tested version will be available in around 2-3 weeks. It mostly depends on the programmer’s availability.