Arbusers

Hi fellas, I don't know why but I don't feel like playing Japan B league is safe. What are your thoughts and do You place bet on such a league?

Mystic wrote: ↑

Tue Oct 24, 2023 3:34 pm

Hi fellas, I don't know why but I don't feel like playing Japan B league is safe. What are your thoughts and do You place bet on such a league?

Japan's Soccer competitions (J1, J2, Emperor's Cup, etc.), their Basketball B1, B2, and Nippon Pro Baseball league are widely covered by various legendary agencies, including valuable offers from time to time.

I suspect you've been caught up by some weird Asian Handicap results over the past few months, like Blaublitz or Shimizu S-Pulse's outcomes or similar.
In the past, I spent many hours specifically observing Blaublitz's Statistical flow until a minimally reliable Model was created. But it is unstable, therefore the time to introduce new fine-tuning in the current Statistics Model for someone's team selected by you in that J2 league is quite different from the average team's Model in J1. I hope that helps you generate your own ideas.

By the way, did you watch the shocking World Cup 2022 match Germany vs Japan? The reason for the outcome was the goalkeeper of the Shimizu S-Pulse team, Gonda Shūichi.

In short, if these criteria are covered, you may consider a dedication here:
- You've got a suitable Model that works for you, tuned up regularly by you;
- Fully accept the fact that the Time/Reward Ratio is average compared to other leagues around the Globe;
- You know some specifics in the lower leagues, for example, skilled players in the starting lineups;
- You have at least minimal idea when minotaurs like SBO, ISN & IBC are wrong in their estimations;
- And so forth.

arb12 wrote: ↑

Tue Oct 24, 2023 6:55 pm

Mystic wrote: ↑

Tue Oct 24, 2023 3:34 pm

Hi fellas, I don't know why but I don't feel like playing Japan B league is safe. What are your thoughts and do You place bet on such a league?

Japan's Soccer competitions (J1, J2, Emperor's Cup, etc.), their Basketball B1, B2, and Nippon Pro Baseball league are widely covered by various legendary agencies, including valuable offers from time to time.

I suspect you've been caught up by some weird Asian Handicap results over the past few months, like Blaublitz or Shimizu S-Pulse's outcomes or similar.
In the past, I spent many hours specifically observing Blaublitz's Statistical flow until a minimally reliable Model was created. But it is unstable, therefore the time to introduce new fine-tuning in the current Statistics Model for someone's team selected by you in that J2 league is quite different from the average team's Model in J1. I hope that helps you generate your own ideas.

By the way, did you watch the shocking World Cup 2022 match Germany vs Japan? The reason for the outcome was the goalkeeper of the Shimizu S-Pulse team, Gonda Shūichi.

In short, if these criteria are covered, you may consider a dedication here:
- You've got a suitable Model that works for you, tuned up regularly by you;
- Fully accept the fact that the Time/Reward Ratio is average compared to other leagues around the Globe;
- You know some specifics in the lower leagues, for example, skilled players in the starting lineups;
- You have at least minimal idea when minotaurs like SBO, ISN & IBC are wrong in their estimations;
- And so forth.

This is very good explained and interesting to read, I really appreciate your time you take to answer me and yes, I was talking about basketball I didn't even mentioned it.

Can you briefly explain how and based on what are you creating your own models? Do You maybe use markov chain monte carlo method or you simply searching for average just by using stats of "X" previous matches of specific team / league. I have some programming skills and always wanted to create model for local football leagues / cups (mostly HNL) to check few interesting things like average goals, how first half ends (1/X/2 / goals / handicaps / etc.), what's most played HT/FT / etc. but I never knew from where to start. Maybe previously mentioned things are not even thing for you so I'm really interested on how you create your ones.

Mystic wrote: ↑

Tue Oct 24, 2023 11:10 pm

arb12 wrote: ↑

Tue Oct 24, 2023 6:55 pm

Mystic wrote: ↑

Tue Oct 24, 2023 3:34 pm

Hi fellas, I don't know why but I don't feel like playing Japan B league is safe. What are your thoughts and do You place bet on such a league?

Japan's Soccer competitions (J1, J2, Emperor's Cup, etc.), their Basketball B1, B2, and Nippon Pro Baseball league are widely covered by various legendary agencies, including valuable offers from time to time.

I suspect you've been caught up by some weird Asian Handicap results over the past few months, like Blaublitz or Shimizu S-Pulse's outcomes or similar.
In the past, I spent many hours specifically observing Blaublitz's Statistical flow until a minimally reliable Model was created. But it is unstable, therefore the time to introduce new fine-tuning in the current Statistics Model for someone's team selected by you in that J2 league is quite different from the average team's Model in J1. I hope that helps you generate your own ideas.

By the way, did you watch the shocking World Cup 2022 match Germany vs Japan? The reason for the outcome was the goalkeeper of the Shimizu S-Pulse team, Gonda Shūichi.

In short, if these criteria are covered, you may consider a dedication here:
- You've got a suitable Model that works for you, tuned up regularly by you;
- Fully accept the fact that the Time/Reward Ratio is average compared to other leagues around the Globe;
- You know some specifics in the lower leagues, for example, skilled players in the starting lineups;
- You have at least minimal idea when minotaurs like SBO, ISN & IBC are wrong in their estimations;
- And so forth.

This is very good explained and interesting to read, I really appreciate your time you take to answer me and yes, I was talking about basketball I didn't even mentioned it.

Can you briefly explain how and based on what are you creating your own models? Do You maybe use markov chain monte carlo method or you simply searching for average just by using stats of "X" previous matches of specific team / league. I have some programming skills and always wanted to create model for local football leagues / cups (mostly HNL) to check few interesting things like average goals, how first half ends (1/X/2 / goals / handicaps / etc.), what's most played HT/FT / etc. but I never knew from where to start. Maybe previously mentioned things are not even thing for you so I'm really interested on how you create your ones.

Hiya again,

Very good question from you.
You understand very well why exactly I cannot afford to explain publicly the algorithms I've created. I suspect the agencies also read that Forum.
My individual Model is the result of many years of observations, experimentation, stubborn overcoming of endless errors and continuous evolution. Therefore, that Model is in a constant beta stage, almost every new modeled event requires specific fine-tuning, and is de facto a creation in a stage of infinite evolution, and not a final finished intellectual product, but in a good "sporting form".


I spotted your enthusiasm here in the Forum and your desire for self-development, congratulations! That's why I'll write some short ideas to you, and some milestones in the development of my creation, in such a way, that can't harm the existence of my intellectual work so far.

- Read Jim Makos' trading experience.
- Read literature by Prof. Spyros Makridakis. Especially search for Forecasting and Time-Series Analysis.
- Learn in detail the Probability Theory & Stats in general.
But initially, statisticshowto.com is the elementary entrance here.
- Now you can generate some ideas.




- Deep fundamental analysis outside of Stats, as deep as it's possible. You like Soccer, don't you?

Croatia vs Türkiye, held on 12th October. The majority of agencies offered Asian Handicap (-0.75) pre-match. Really adequate "prognosis"? The reading of the starting line-ups? First tragic inertial minutes and tactical approaches by Vincenzo Montella and late reaction by Zlatko Dalić and the inertia and indifference of some players despite his later tactical attention to them? Modrić's accumulated fatigue and nervousness far before that match (due to Real Madrid) and Montella's tactical approach to him, knowing that fact? The personal estimations of every Croatian player during the initial minutes before the goal? Why? Therefore, the probability of failure is ...? The probability of turnaround is ...? And so on, and so on. Thus, the cumulative probability for (-0.75) AH, calculated by your side is ...? Why? The agencies offered close to Evens, why? The hypothetical expected value was positive, why? If you react later in this match, (-0.75) transforms into (-0.5) and new iterations are required, etc.
But why the Croatian attacks picked up more power (estimate it XXX points) when Dion Beljo appeared on the pitch? The probability of scoring after the 80th minute due to Beljo? The coach Montella and the Turkish defensive dept didn't react well here, so the lack of scoring after the 80th minute was due to ....? Why?


Another example: Uruguay. Marcelo Bielsa and his tactics. Asian Handicaps love Marcelo Bielsa. He and Jorge Sampaoli are the last minotaurs from an old South American Soccer school, very good indeed, but very predictable. I've been following them for years. Pep Guardiola worships two very different Master Tacticians - the defence-oriented Carlo Mazzone and the offensive-oriented Marcelo Bielsa.
Bielsa's open mind does wonders in every team when several conditions appear. One of those conditions is that romantic attacking old-school soccer fails in some matches (due to much more modern defensive-complex tactics by the opponents) and Bielsa rethinks.
He prepares specific pseudo-defensive tactical pitfalls. When it is combined with other things, spotted by me, the probability that the Asian Handicap bomb will explode is very high... Spotted that in Leeds United under Bielsa's management many times. Uruguay was blocked vs Ecuador, but all the conditions I was looking for didn't appear against Colombia, so the triggers came noisy days later. - Bielsa's Uruguay vs Brasil, 2:0 in the end. AH markets went mad.

"Tata" Martino? A very different case, but a very predictive story, he fails the next time, and the next... That lasts until he changes some specific tactical things under pressure... But here the market is completely different.



- Use ELO ratings in a very dynamic way. Pre-match & in-play. Every player, team etc;
- MCMC (Markov Chain Monte-Carlo) very good algorithms indeed, but I advise you to look at methods, called ABC (Approximate Bayesian Computation) for simplification purposes.

- Despite the abovementioned, it's much better as a data-rookie to start with Poisson & Weibull and to develop them further in your way, via experiments, before computing the real odds. For example, the "average" parameter to be searched, mentioned by you in your question could be simply λ parameter in most elementary cases.

Here you are:
Dynamic Bivariate Poisson:
papers.tinbergen.nl/12099.pdf

Discrete Weibull:
air.unimi.it/bitstream/2434/640408/2/authorFinalVersion.pdf


Final words: all that above has to be modified your way one day, via evolution and experiments. Wish you success.

arb12 wrote: ↑

Tue Oct 24, 2023 6:55 pm

Hiya again,

Very good question from you.
You understand very well why exactly I cannot afford to explain publicly the algorithms I've created. I suspect the agencies also read that Forum.
My individual Model is the result of many years of observations, experimentation, stubborn overcoming of endless errors and continuous evolution. Therefore, that Model is in a constant beta stage, almost every new modeled event requires specific fine-tuning, and is de facto a creation in a stage of infinite evolution, and not a final finished intellectual product, but in a good "sporting form".


I spotted your enthusiasm here in the Forum and your desire for self-development, congratulations! That's why I'll write some short ideas to you, and some milestones in the development of my creation, in such a way, that can't harm the existence of my intellectual work so far.

- Read Jim Makos' trading experience.
- Read literature by Prof. Spyros Makridakis. Especially search for Forecasting and Time-Series Analysis.
- Learn in detail the Probability Theory & Stats in general.
But initially, statisticshowto.com is the elementary entrance here.
- Now you can generate some ideas.




- Deep fundamental analysis outside of Stats, as deep as it's possible. You like Soccer, don't you?

Croatia vs Türkiye, held on 12th October. The majority of agencies offered Asian Handicap (-0.75) pre-match. Really adequate "prognosis"? The reading of the starting line-ups? First tragic inertial minutes and tactical approaches by Vincenzo Montella and late reaction by Zlatko Dalić and the inertia and indifference of some players despite his later tactical attention to them? Modrić's accumulated fatigue and nervousness far before that match (due to Real Madrid) and Montella's tactical approach to him, knowing that fact? The personal estimations of every Croatian player during the initial minutes before the goal? Why? Therefore, the probability of failure is ...? The probability of turnaround is ...? And so on, and so on. Thus, the cumulative probability for (-0.75) AH, calculated by your side is ...? Why? The agencies offered close to Evens, why? The hypothetical expected value was positive, why? If you react later in this match, (-0.75) transforms into (-0.5) and new iterations are required, etc.
But why the Croatian attacks picked up more power (estimate it XXX points) when Dion Beljo appeared on the pitch? The probability of scoring after the 80th minute due to Beljo? The coach Montella and the Turkish defensive dept didn't react well here, so the lack of scoring after the 80th minute was due to ....? Why?


Another example: Uruguay. Marcelo Bielsa and his tactics. Asian Handicaps love Marcelo Bielsa. He and Jorge Sampaoli are the last minotaurs from an old South American Soccer school, very good indeed, but very predictable. I've been following them for years. Pep Guardiola worships two very different Master Tacticians - the defence-oriented Carlo Mazzone and the offensive-oriented Marcelo Bielsa.
Bielsa's open mind does wonders in every team when several conditions appear. One of those conditions is that romantic attacking old-school soccer fails in some matches (due to much more modern defensive-complex tactics by the opponents) and Bielsa rethinks.
He prepares specific pseudo-defensive tactical pitfalls. When it is combined with other things, spotted by me, the probability that the Asian Handicap bomb will explode is very high... Spotted that in Leeds United under Bielsa's management many times. Uruguay was blocked vs Ecuador, but all the conditions I was looking for didn't appear against Colombia, so the triggers came noisy days later. - Bielsa's Uruguay vs Brasil, 2:0 in the end. AH markets went mad.

"Tata" Martino? A very different case, but a very predictive story, he fails the next time, and the next... That lasts until he changes some specific tactical things under pressure... But here the market is completely different.



- Use ELO ratings in a very dynamic way. Pre-match & in-play. Every player, team etc;
- MCMC (Markov Chain Monte-Carlo) very good algorithms indeed, but I advise you to look at methods, called ABC (Approximate Bayesian Computation) for simplification purposes.

- Despite the abovementioned, it's much better as a data-rookie to start with Poisson & Weibull and to develop them further in your way, via experiments, before computing the real odds. For example, the "average" parameter to be searched, mentioned by you in your question could be simply λ parameter in most elementary cases.

Here you are:
Dynamic Bivariate Poisson:
papers.tinbergen.nl/12099.pdf

Discrete Weibull:
air.unimi.it/bitstream/2434/640408/2/authorFinalVersion.pdf


Final words: all that above has to be modified your way one day, via evolution and experiments. Wish you success.

You don't realize how much I appreciate your post, thank you so much for the details!

I will definitely read everything, but do you have any video/videos to suggest because I have a problem with reading, but considering that this is actually something I am really interested in, as I already said, I will definitely read it.

And considering why everyone doesn't write about some things on the forum, he would like to see some (of course "regulated") group where there will be people who obviously have experience and share with each other.

Glad to help you, here's random video:
youtube.com/watch?v=sbbYntt5CJk
But you'd much better buy books/PDFs and read former Olympian Prof. Makridakis, Time-Series Analysis, Forecasting.