Topic: Developing AI-software for betting on football (Read 113 times)
You've written an excellent example of idea. There is a lot of ideas and methods to increase a value. Only technics that have an effect to bets will be added to work.
It is an engineer solution. By general, there are two ways to predict. In the first case we build a blackbox with given data and one of NN and do experiments with it.
In the second case we build a machine with a set of algorithms of data processing from the roots. We have each of numbers under control, can see and find any dependencies, can build a model as lego.
In this way, these metrics more reliable to predict than results of games that can be randomically from one side and calculcated by bookmakers from other.
A statistic has a straight relation to a style of games and teams that we can see at charts. With these numbers it's a good fight for a data-ninja.
The target isn't just to predict whatever. Basically, the system is a workstation for bettors with some tools and features.
I've started to make a demo and materials about it. New content is coming soon.
In the second case we build a machine with a set of algorithms of data processing from the roots. We have each of numbers under control, can see and find any dependencies, can build a model as lego.
In this way, these metrics more reliable to predict than results of games that can be randomically from one side and calculcated by bookmakers from other.
A statistic has a straight relation to a style of games and teams that we can see at charts. With these numbers it's a good fight for a data-ninja.
The target isn't just to predict whatever. Basically, the system is a workstation for bettors with some tools and features.
I've started to make a demo and materials about it. New content is coming soon.
Wonder why you choose those metrics only, or are there more?
I think you could also predict a match fairly well by predicting how the coaches involved will respond to their opponents tactics.
By the way, I have wondered sometimes if lots of the players next moves can be accurately guessed during a match...so you get people to predict their every moves... and the ones who are consistently correct get rewarded with rank upgrade or something. I think a well trained predictors could do that.
Your metrics look basics and I wonder how reliable they will be.
I think you could also predict a match fairly well by predicting how the coaches involved will respond to their opponents tactics.
By the way, I have wondered sometimes if lots of the players next moves can be accurately guessed during a match...so you get people to predict their every moves... and the ones who are consistently correct get rewarded with rank upgrade or something. I think a well trained predictors could do that.
Your metrics look basics and I wonder how reliable they will be.
There are some reasons.
- That super machine is not fully ready yet.
- Developing, analysis, design, support take all my time. I'm coder, not a businessman, not a capper or everything else.
- Guys, who are doing an analysis of statistics, have around 100 matches at a weekend. It also takes all their time.
Early we hadn't coefs data in the system, someone of newbie alanysts had many troubles with gadgets and it was so hard. Guys had known what is red eyes are mean.
So, step by step, we are growing up models, skills and technical parts.
- Bookmakers control mostly clever users, there is own specificity.
- I haven't enough power for all of it. It is like a pet project for me.
Of course, I'm going by a plan such as you described.
May be, I said incorrectly about investing. I mean that a little crowdfunding company is a good idea in that case.
By example, if you'll go to cappers channels, for 100-200$ you can buy bloger's fantasies, random predictions and other good stuff
Probably, I just came to speech with a community, to start a dialogue. I'm not experienced in social media, marketing and selling.
By definition, it is project from guys to guys
- That super machine is not fully ready yet.
- Developing, analysis, design, support take all my time. I'm coder, not a businessman, not a capper or everything else.
- Guys, who are doing an analysis of statistics, have around 100 matches at a weekend. It also takes all their time.
Early we hadn't coefs data in the system, someone of newbie alanysts had many troubles with gadgets and it was so hard. Guys had known what is red eyes are mean.
So, step by step, we are growing up models, skills and technical parts.
- Bookmakers control mostly clever users, there is own specificity.
- I haven't enough power for all of it. It is like a pet project for me.
Of course, I'm going by a plan such as you described.
May be, I said incorrectly about investing. I mean that a little crowdfunding company is a good idea in that case.
By example, if you'll go to cappers channels, for 100-200$ you can buy bloger's fantasies, random predictions and other good stuff
Probably, I just came to speech with a community, to start a dialogue. I'm not experienced in social media, marketing and selling.
By definition, it is project from guys to guys
Now I need your support, because I have spent all of my resources for development gg
If you have already developed a system, why not take advantage of it yourself and earn the necessary funds at bookmakers?
With a money-making machine, investing support seems pointless.
Don't have an interest in betting? I'm too, it is not the best thing in the world ) But it is so funny to hack the system
Hi everyone!
I was banned on all betting forums when I tried to post this, but it is not spam, is not scam, this is my project that I've been developing for the last year.
Well, I've been developing a software system for betting on football.
It is a software for analysts based on such technologies as big data, stats and combinatorics algos, ML and neural networks (a lot of loud words, sorry), some tactics from ABC of poker and other tricks.
These technics work perfectly in many commerce and industrial companies and it may works great for bets also, I guarantee it xD
Notice, we work only with metrics like corners, shots, card and offsides, without goals and results of games.
Sounds hard, but really it is so easy. The main is know how to cook it (but I think it is a preschool level for people from crypto's party).
Now MVP is ready for manually work with statistics (dudes from our team work with it without problems), coefs integrated (just from one bookmaker), prediction models are under development.
We support all Europe top leagues, some other, CL and EL; championships of national teams are coming soon.
Collected data includes full info of ~100 000 matches.
For a year of hard work we have prepared base of knowledge with great info from zero to a way to auto predictions.
Now I need your support, because I have spent all of my resources for development gg
Guys, it is a great deal. It is not same great as cryptocurrencies of course, but it is so cool also.
Current target is join a small community, get some money, deploy the system for comrades, and back into the underground.
If you've figured out what is mean it and have a few hundred dollars to a small invest, please, you're welcome
You can write me to PM or to email that you can find in my profile.
I was banned on all betting forums when I tried to post this, but it is not spam, is not scam, this is my project that I've been developing for the last year.
Well, I've been developing a software system for betting on football.
It is a software for analysts based on such technologies as big data, stats and combinatorics algos, ML and neural networks (a lot of loud words, sorry), some tactics from ABC of poker and other tricks.
These technics work perfectly in many commerce and industrial companies and it may works great for bets also, I guarantee it xD
Notice, we work only with metrics like corners, shots, card and offsides, without goals and results of games.
Sounds hard, but really it is so easy. The main is know how to cook it (but I think it is a preschool level for people from crypto's party).
Now MVP is ready for manually work with statistics (dudes from our team work with it without problems), coefs integrated (just from one bookmaker), prediction models are under development.
We support all Europe top leagues, some other, CL and EL; championships of national teams are coming soon.
Collected data includes full info of ~100 000 matches.
For a year of hard work we have prepared base of knowledge with great info from zero to a way to auto predictions.
Now I need your support, because I have spent all of my resources for development gg
Guys, it is a great deal. It is not same great as cryptocurrencies of course, but it is so cool also.
Current target is join a small community, get some money, deploy the system for comrades, and back into the underground.
If you've figured out what is mean it and have a few hundred dollars to a small invest, please, you're welcome
You can write me to PM or to email that you can find in my profile.
Jump to: