Kategoria:
About us
ESBC PowerAI
ESBC PowerAI is the result of countless research hours and was implemented by our research department in 2016. We have integrated all factors and dependencies that, according to our research, affect the probability distributions of matches. We have optimized the significance of all statistical data available from matches concerning the teams' future performance. Our system compares actual statistics to expected ones and objectively adjusts the team's performance indicators based on the direction suggested by the facts. We have also optimized the rates of change for different factors to maximize the accuracy of predicting a team's future performance. After its launch in 2016, we have of course continued to work on its development, updating the program continuosly with our latest research results to make it even more precise.
We have been delighted to see the rise of expected goals thinking, which has been utilized by ESBC and pioneered by Jorma Vuoksenmaa. He was an early adopter of the method, using it decades ago with the traditional pen and paper method while watching TV or being in the stands. Nowadays, we don't need as much pen and paper, but ESBC PowerAI also incorporates expected goals (ESBC PowerAI xG) in our modeling, albeit with appropriate weight and without overreacting to them.
The foundation of ESBC PowerAI lies in the ESBC PowerAI Ratings, which indicate a team's current baseline performance level. Our program fine-tunes these ratings meticulously for each match, taking into account all special factors, such as player absences, new reinforcements, team fatigue, mental state, starting goalkeeper, motivation factors, and more.
For our experienced odds compilers, ESBC PowerAI provides a detailed, scientific, and objective tool to arrive at highly accurate probability estimates. These estimates serve as the basis for our betting recommendations and our own betting activities.
We continuously study the accuracy of our model and estimates using statistical methods, comparing them to odds from various bookmakers. Simultaneously, we determine optimal weightings to refine ESBC PowerAI's probability estimates to align even better with reality. Sports betting can never be reduced to a simple probability calculation like roulette, and a bit of "smoothing" to rational divergent views almost always improves results slightly.
While we do not reveal our precise research findings on a broader scale, here's a concrete examples from "under the hood":
- ESBC PowerAI crushed the market on its trial by fire in EURO 2016 - earning us and our customers massive 127,3 % Return of investment (ROI)
- In Finnish Liiga, ESBC PowerAI made quickly history. Our pure predictions offered the best predictive power already on the 2nd year of ESBC PowerAI's operation (2017-2018) and that accuracy also remained the next season (2018-2019):
ESBC PowerAI 100% - Bookmakers 0%.
Of course the most important thing is long-term results and statistically very meaningful sample of the first 10 000 ESBC PowerAI recommendations (2016-2021) yielded very nice 103,6 % ROI per bet. In betting compound interest plays key role in long-term net profit, but in this graph stakes are calculated just from static original bankroll i.e. compound interest doesn't even show in this impressive graph.
ESBC - Together More from the Games