Nice video! I have currently implemented this in Python and I am currently figuring out how to weigh the previous results with respect to the quality of the team played against. I was thinking about logistic regression, but I am struggling on how to implement this ratio in the Goal expectance/Poisson distribution. Does anybody have a clue?
legit bullshit i been waiting for new content from this channel for quite a while youtubes algorithm just never put it on my homescreen clicked the bell icon now
Poisson is flawed as its weighted to lower values, thats why there is something called zero inflated poisson. Also all these percentages are calculated by multiplying the home score percentage by the away score percentage which implies the events are correlated, they aren't.
I saw Poisson models here on youtube where you multipyling Home Team Attack Strength with Home Team Defensive Weakness(?) You are doing on different way?.. And seems that I am not right but your way is the correct one!
@@thefootballbank2924 Very important moment! Thanks! Now poisson data has some sense... I don't like xG... Have you ever gave some thoughts to Pithagorical analyzes?
Absolutely good point - i use the Poisson to get an indicator on BTTS or home/away win. Precisely for the reason you state I don't think it can be relied upon to call draws. So in short no adjustments to the distribution there is just implicit understanding that the % chance of home win and away win is likely smaller and should be reallocated to the draw outcome.
Nice Content....Just want to ask what if the game played are not evenly just the way we have 6 game played in the content...can this poisson distribution method still work Incase there is an outstanding make the game played unevenly.
Good question, Poisson will still work in this case just make sure to divide the total number of goals by the number of games that specific team has played. Usually there isn't much difference in the number of games played, but with the disruptions to the leagues from last two years some of the variance has been greater.
I've added an average goal per game from the last six. I averaged goals on the season and the goals from the last six. Currently, I have them weighted 60/40 respectively. What do you think about that percentage?
Hello, I have the data scrapped and loaded into an excel file, what is the color pattern formular, for the first sheet, containing home and away . Thank you.
Nice content! I'm interested in knowing how much time does it take you to keep the table updated week after week for the whole premier league? And wouldn't it be easier just to check the given odds for all possible results considering that the house does the math right and more efficiently?
Thank you! So it takes a little work but I've automated some of the data collection. The analysis still needs to be performed manually. Of course you can use the betting companies' odds as an indicator, but keep in mind that they do need to make sure there is equal money on both sides of a bet, so if a lot of people are hammering a favourite the odds offered by the bookmaker will be lower than the true odds - another reason to stay away from heavy favourites!
Great video! Just the info I've been looking for. Thank you for taking the time to put it together.
Thank you - glad it was helpful :)
Can we get the excel file for download bro...?
Nice video! I have currently implemented this in Python and I am currently figuring out how to weigh the previous results with respect to the quality of the team played against. I was thinking about logistic regression, but I am struggling on how to implement this ratio in the Goal expectance/Poisson distribution. Does anybody have a clue?
Can you share the code?
Excellent calculation... Good work... We don't have any scale to measure the over or under matches where we can reach more than 55% in 2.00
legit bullshit
i been waiting for new content from this channel for quite a while
youtubes algorithm just never put it on my homescreen
clicked the bell icon now
Thanks for the support :)
Poisson is flawed as its weighted to lower values, thats why there is something called zero inflated poisson. Also all these percentages are calculated by multiplying the home score percentage by the away score percentage which implies the events are correlated, they aren't.
Hi Thank you for your time to advise us really help me to analyzed. hope we can learn more tips and advise from you. Thank you :)
Thank you for the support :)
Best information ever just what I needed my problem is can you please give us the template for this excel document🙏🙏
I saw Poisson models here on youtube where you multipyling Home Team Attack Strength with Home Team Defensive Weakness(?) You are doing on different way?.. And seems that I am not right but your way is the correct one!
Yea for home xG it should be Home Attack strength multiplied by Away Defensive strength :)
@@thefootballbank2924 Very important moment! Thanks! Now poisson data has some sense... I don't like xG... Have you ever gave some thoughts to Pithagorical analyzes?
Poisson always understates the likelihood of a draw - do you make adjustments for that in your distribution?
Absolutely good point - i use the Poisson to get an indicator on BTTS or home/away win. Precisely for the reason you state I don't think it can be relied upon to call draws. So in short no adjustments to the distribution there is just implicit understanding that the % chance of home win and away win is likely smaller and should be reallocated to the draw outcome.
we've missed U
Do you have an app or website where you publish the predictions?
Nice Analysis.
Pls how can I get the Excel setup for the poison distribution
it's easy to set up the poisson. I can create one for you. It's feeding in and maintaining the underlying data which feeds into it which is hard :)
@@thefootballbank2924 KINDLY CREATR THE EXCEL FOR US ..
THANKS
I have the data scrapped and done setting up the excel, except the color grouping patterns. How was the color pattern grouped.
Nice Content....Just want to ask what if the game played are not evenly just the way we have 6 game played in the content...can this poisson distribution method still work Incase there is an outstanding make the game played unevenly.
Good question, Poisson will still work in this case just make sure to divide the total number of goals by the number of games that specific team has played. Usually there isn't much difference in the number of games played, but with the disruptions to the leagues from last two years some of the variance has been greater.
great sir
Bro, Good Morning
We are waiting patiently for the Excel Downloaded link .
Thanks
Nice job!! Can I get this excel file? Thanks!!
I love that
Please share the excel document
does the understat website provide the xG only after the match has played?
I've added an average goal per game from the last six. I averaged goals on the season and the goals from the last six. Currently, I have them weighted 60/40 respectively. What do you think about that percentage?
I might lean the weighting more to form than season. Try running 70/30 in parallel and see what gives better results
Hello, I have the data scrapped and loaded into an excel file, what is the color pattern formular, for the first sheet, containing home and away . Thank you.
I did the calculation but the home went on loosing even though the odds favoured home win
My away teams poisson is not adding up to 100% and its driving me mad
We real need this money 😄
You still haven't made the excel available for us
Nice content! I'm interested in knowing how much time does it take you to keep the table updated week after week for the whole premier league? And wouldn't it be easier just to check the given odds for all possible results considering that the house does the math right and more efficiently?
Thank you! So it takes a little work but I've automated some of the data collection. The analysis still needs to be performed manually. Of course you can use the betting companies' odds as an indicator, but keep in mind that they do need to make sure there is equal money on both sides of a bet, so if a lot of people are hammering a favourite the odds offered by the bookmaker will be lower than the true odds - another reason to stay away from heavy favourites!
whats the website you use when you watch the number of xg 22.00?
understat is pretty good for European leagues
kinda need another rollover games
when will U be back for good
Upcoming season there should be more regular content as well as live streams on Saturday or Sunday - be sure to drop in!
Please can we download this exel file
Can you send excel file? Thanks!
Can we have a downloadable link?
Let me see how i can do this and then i'll share a link.
@@thefootballbank2924is there a link now?