It's beginning to sound a lot Christm.... Soccer Mystic.Derek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
Soccer Poisson database
-
- Posts: 4327
- Joined: Tue Oct 25, 2016 12:50 pm
That must have taken a fair bit or work, thanks for sharing mateDerek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
+1Trader Pat wrote: ↑Fri May 28, 2021 10:04 pmThat must have taken a fair bit or work, thanks for sharing mateDerek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
Apart from "mate". I don't think we have that status
- Silverthorn
- Posts: 26
- Joined: Sat Mar 13, 2021 2:47 pm
- Location: Australia
Awesome job Derek. thanks soo Much. What a Fantastic tool!! You just gave my Soccer Trading a shot in the arm. Took a LOT of setting up and just a tad of tweaking to suit my system but so much easier selecting suitable events now!!
Thanks. I really created it as a toy and VBA exercise, but I'm beginning to see potential uses for it for my own trading as well.Silverthorn wrote: ↑Wed Jun 02, 2021 3:02 amAwesome job Derek. thanks soo Much. What a Fantastic tool!! You just gave my Soccer Trading a shot in the arm. Took a LOT of setting up and just a tad of tweaking to suit my system but so much easier selecting suitable events now!!
- Silverthorn
- Posts: 26
- Joined: Sat Mar 13, 2021 2:47 pm
- Location: Australia
A wise old Engineer once told me something early in my Engineering career that has done me well in so many aspects of life.
A good tradesman doesn't blame his tools because he doesn't own crap tools. A great tradesman cares for and modifies his tools to make them a perfect fit for him. You can build a house with just a tape measure a hammer and a saw but, while it's possible it's not advisable. Get good tools. And for Gods sake stop driving home nails with the back of your saw. Every tool has a purpose.
I think this is a great tool but you can't drive nails with it. Thanks.
A good tradesman doesn't blame his tools because he doesn't own crap tools. A great tradesman cares for and modifies his tools to make them a perfect fit for him. You can build a house with just a tape measure a hammer and a saw but, while it's possible it's not advisable. Get good tools. And for Gods sake stop driving home nails with the back of your saw. Every tool has a purpose.
I think this is a great tool but you can't drive nails with it. Thanks.
- firlandsfarm
- Posts: 2724
- Joined: Sat May 03, 2014 8:20 am
Hi Derek, I bring an olive branch and wave a white flag!Derek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
Just found this thread while catching up, sounds brilliant tool. Not got to the end and downloaded/played yet but the above post raises an unanswered question I have had for a while ... is there a close to the curve formula for calculating the decay (or 'boost', is that the right term when the price naturally increases as the time decreases?) as per SM and your tool?
I think SM calculates probabilities based on prices in associated markets while the Poisson distribution is purely based on goal expectancy. Assuming a match is 96 minutes, 15 minutes into the match the goal expectancy would be multiplied by 81/96 before being put through the Poisson function.firlandsfarm wrote: ↑Tue Jun 22, 2021 8:21 amHi Derek, I bring an olive branch and wave a white flag!Derek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
Just found this thread while catching up, sounds brilliant tool. Not got to the end and downloaded/played yet but the above post raises an unanswered question I have had for a while ... is there a close to the curve formula for calculating the decay (or 'boost', is that the right term when the price naturally increases as the time decreases?) as per SM and your tool?
https://en.wikipedia.org/wiki/Poisson_d ... s_function
I tested it in fast-forward mode with varying probabilities. Whether it's accurate is another matter but it was remarkably consistent with the movement of real markets, as you'd expect.
- firlandsfarm
- Posts: 2724
- Joined: Sat May 03, 2014 8:20 am
The content of that link is way above my pay grade! I recently spent ages researching a (to me) complicated probability formula just so that I could apply it and I think that has exhausted my aging brain! (The problem was not the calculation but establishing what each symbol meant and how I should expand them to get a computable formula.) Gotta hand it to you for embarking on such a task, I'll have a look at the downloads later this week.
-
- Posts: 1084
- Joined: Fri Nov 20, 2015 9:38 am
If your are measuring the time between goals that is actually the exponential distribution (which is implicit in Poisson).
Not sure if that makes things easier or not to understand but as you have a goal "rate" for each player it is possible to model individual players and then combine to get the "team" expectation.
https://en.wikipedia.org/wiki/Exponential_distribution
edit - and the issue with both distributions is that the events have to be "memoryless" - I think I can recall an academic article which found this assumption is not true for footy... but they do seem to be the models of choice for footy )
edit edit - and the rate needs to be constant, which it certainly isn't in real life but then it is just a model...
Not sure if that makes things easier or not to understand but as you have a goal "rate" for each player it is possible to model individual players and then combine to get the "team" expectation.
https://en.wikipedia.org/wiki/Exponential_distribution
edit - and the issue with both distributions is that the events have to be "memoryless" - I think I can recall an academic article which found this assumption is not true for footy... but they do seem to be the models of choice for footy )
edit edit - and the rate needs to be constant, which it certainly isn't in real life but then it is just a model...
When people knock Poisson for football because it underestimates the draw, goals aren't entirely independent of each other, etc. my feelings are that awareness of danger is more important than the danger itself. There's nothing dangerous about a car with worn-out brake pads - it's being unaware that they're worn out that's the danger.
Poisson has its uses and limitations for football but I'm sure somebody cleverer than I am will have worked out algorithms to adjust the second half curve where goals are more likely. I might have a go at some point in the future.
Poisson has its uses and limitations for football but I'm sure somebody cleverer than I am will have worked out algorithms to adjust the second half curve where goals are more likely. I might have a go at some point in the future.
Athletico PR were underdogs for this match and traded 60-70 pre-match for any other home win. Based on half a seasons results, which I know has shortcomings, I had them odds-on to win and 16 odd for AOHW. Final score 4-2 to Athletico.
You do not have the required permissions to view the files attached to this post.
-
- Posts: 9
- Joined: Thu Jun 13, 2019 5:36 pm
Hi very well made, been using it for a bit had some good resultsDerek27 wrote: ↑Fri May 28, 2021 6:32 pmI've upgraded the front-end database and Excel template. The setup now has a scoreboard for updating the current score and a countdown timer with a start/stop toggle button (first worksheet only). The prices/probabilities are now based on the Poisson mass function for the time remaining of the game. You can adjust the minutes remaining at any time or take a peek into a future point in the match, like Soccer Mystic. It's not quite the same as SM as the spreadsheet will be ignorant of current prices and purely based on goal expectation. Might be useful for matches that don't have the markets required for SM.
Just wondering have you made any further improvements to it
Thanks
James