Poisson Distribution - This Can Help You

cairod
Group 2 Class
Tips
Posts: 103
Joined: Fri Sep 13, 2013 11:00 pm
Location: Mow Cop

Re: Poisson Distribution - This Can Help You

Postby cairod » Wed Sep 06, 2017 7:37 pm

I stick to singles. Although I've got rid of one the Leeds v Burton bets and put a small stake on it all as an accumulator. Comes to over 2000-1 :shock:

makkad
Listed Class
Tips
Posts: 30
Joined: Wed Aug 10, 2011 11:00 pm

Re: Poisson Distribution - This Can Help You

Postby makkad » Fri Sep 08, 2017 12:28 pm

Here's a link to my Poisson Analysis for England Premiership, Championship, Bundesliga, Spain Primera Liga and Itlay Serie A matches.

It's possible to input bookie odds in order to identify 'value bets'

https://drive.google.com/file/d/0B9MTco ... sp=sharing
Last edited by makkad on Fri Sep 08, 2017 1:16 pm, edited 1 time in total.

makkad
Listed Class
Tips
Posts: 30
Joined: Wed Aug 10, 2011 11:00 pm

Re: Poisson Distribution - This Can Help You

Postby makkad » Fri Sep 08, 2017 12:30 pm

Here's a link to a summary of the Poisson predictions. I've left some previous matches analysis in order for the user to understand the outcome of the predictions

https://drive.google.com/file/d/0B9MTco ... sp=sharing

I recommend sorting on e.g. dcH (Double chance home - column AD) - then look at past results - which may indicate the outcome of this weeks forecasts

or, try Gd (column H) - which is the Poisson analysis basic goal strength difference for each match. It's surprising how this indicator can help differentiate with forecasts 1X2
Last edited by makkad on Sat Sep 09, 2017 4:50 am, edited 1 time in total.

davidg3907
At Stud
User avatar
Betting Blog Tips
Posts: 4854
Joined: Thu Jun 25, 2009 11:00 pm
Location: Oadby , Leicestershire

Re: Poisson Distribution - This Can Help You

Postby davidg3907 » Fri Sep 08, 2017 12:43 pm

I'm pleased to see that Cairod's and my figures are fairly close despite using somewhat different methods.

I have Liverpool at 27.98%, Watford at 56.08% and Stoke/Man Utd NO at 59.56%.

The first 3 rounds of Premier League matches produced a glut of goals in the first week and a dearth in the other two. With that inconsistency in mind, I do not expect the data I have for the first 30 matches to be very representative going forward, so will keep an eye open for some stability up to the next international break.

With an expectancy of about 2.80 goals per match, the current average of 2.433 is unlikely to be maintained. Similarly, the ratios of BTTS and under/over 2.5 goals are also out of tolerance. As the figures below reflect these trends, there is no great alarm due to the very small sample range.

Average Goals.
The top 15 figures have seen at least 2 goals in 13 of them, and at least 3 in 7 of them. the bottom 15 have 7 instances of under 1.5 goals, a further 4 of just 2 goals, and 4 over 2.5 goals.

Over/under 2.5 Goals.
Again this follows the same pattern, with 8 of the Overs in the top 15, and just 4 in the bottom 15.

Home teams winning to nil.
7 of the top 15 have won to nil, 4 of the bottom 15 have done the same.

Away teams winning to nil.
Only one of the top 8 has lost, 4 winning to nil. just 2 of the bottom 15 have won to nil, with none of the bottom 11 doing so.

BTTS Yes.
Although just 3 of the top 10 have produced a YES, 8 of the 9 BTTS YES matches have been in the top 21 and just 1 in the bottom 9.

BTTS No.
Obviously, this is the opposite of the above, so 8 of the top 9 predictions correct.

I shall post the top few in each category of this weekend's matches later today.
LAYING Taking it further
LAYING Delving deeper - NEW

FOOTBALL blogs under Blogs/Football
New RELEGATION diary in Blogs/systems

davidg3907
At Stud
User avatar
Betting Blog Tips
Posts: 4854
Joined: Thu Jun 25, 2009 11:00 pm
Location: Oadby , Leicestershire

Re: Poisson Distribution - This Can Help You

Postby davidg3907 » Sat Sep 09, 2017 9:10 am

The top percentages in the various categories this week are listed below. Matches involving Huddersfield, Brighton, and to a lesser extent Newcastle are based on a reduced amount of available data, and coupled with estimates.

Average Goals.
3.73 Man City v Liverpool 3.68 West Ham v Huddersfield 3.58 Arsenal v Bournemouth 2.80 Leicester v Chelsea 2.74 Everton v Tottenham

Average Goals.
1.90 Brighton v West Brom 1.90 Swansea v Newcastle 1.95 Burnley v Crystal Palace 2.42 Stoke v Man Utd 2.60 Southampton v Watford

Over 2.5 Goals.
68.95 Man City v Liverpool 67.88 Arsenal v Bournemouth 67.11 West Ham v Huddersfield 51.08 Leicester v Chelsea

Under 2.5 Goals. (The quoted figures are for OVER so deduct from 100)

29.38 Swansea v Newcastle 29.48 Brighton v West Brom 30.56 Burnley v Crystal Palace 42.53 Stoke v Man Utd
46.57 Southampton v Watford 49.83 Everton v Tottenham

Home teams winning to nil.
43.34 Arsenal 30.57 Swansea 25.27 Burnley

Away teams winning to nil.
36.34 Man Utd 32.31 West Brom 26.39 Tottenham 25.91 Huddersfield 23.01 Crystal Palace

BTTS Yes.
63.48 Man City 59.82 West Ham 50.64 Leicester

BTTS No. (The quoted figures are for YES so deduct from 100)
28.89 Swansea v Newcastle 33.27 Brighton v West Brom 35.20 Burnley v Crystal Palace 40.44 Stoke v Man Utd
48.41 Everton v Tottenham 49.07 Southampton v Watford 49.74 Arsenal v Bournemouth
LAYING Taking it further
LAYING Delving deeper - NEW

FOOTBALL blogs under Blogs/Football
New RELEGATION diary in Blogs/systems

cairod
Group 2 Class
Tips
Posts: 103
Joined: Fri Sep 13, 2013 11:00 pm
Location: Mow Cop

Re: Poisson Distribution - This Can Help You

Postby cairod » Sat Sep 09, 2017 12:59 pm

I've started cobbling together a spreadsheet based on number of shots per game for each team. So far it works out the likelihood of each team getting a finite number of shots but I'm not sure where I can take this now.

I was going to use this data to find out likelihood of specific number of goals based on how many goals each team scores per number of shots. I'm just trying to work out whether that will just give exactly the same results as the average goal method and therefore be a complete waste of time. My brains aching a bit!

davidg3907
At Stud
User avatar
Betting Blog Tips
Posts: 4854
Joined: Thu Jun 25, 2009 11:00 pm
Location: Oadby , Leicestershire

Re: Poisson Distribution - This Can Help You

Postby davidg3907 » Sat Sep 09, 2017 1:30 pm

Cairod,

That is a regular conundrum, and why I treat Average Goals and Over/Under 2.5 etc as largely unrelated.

The example I tend to use is two people playing a game of dice.

One has a standard die with the digits 1-6 on the 6 faces, while the other has a 0 on 5 faces, and a 21 on the other. Given a large enough sample set of throws, both will average about 3.50. One will be above average 50% of the time and below average 50% of the time. The other will be above average just 16.67% of the time and below average 83.33%.

Unless it is clear what drives the figures in such cases, the outcome can be misleading.

It may actually take you as long to work out the difference in the data as it would to check the difference in the results. :D
LAYING Taking it further
LAYING Delving deeper - NEW

FOOTBALL blogs under Blogs/Football
New RELEGATION diary in Blogs/systems

cairod
Group 2 Class
Tips
Posts: 103
Joined: Fri Sep 13, 2013 11:00 pm
Location: Mow Cop

Re: Poisson Distribution - This Can Help You

Postby cairod » Sat Sep 09, 2017 2:56 pm

It may actually take you as long to work out the difference in the data as it would to check the difference in the results. :D
At this rate the season will be over by the time I've even developed the spreadsheet let alone checking any differences :D

I'm looking at working out the number of shots a team would usually require to score a specific number of goals. I'm just remembering that the Poisson distribution deals with finite numbers so anything with a decimal point is meaningless. So I was thinking of creating a combined likelihood of a team getting a range of finite number of shots which corresponds to the number of goals but I'm sure that's flawed because there'll be teams who only get 3 shots in a game and score 3 goals and on the flip-side there'll be teams who get 30 chances and not be able to score if they play until midnight (something I'm very familiar with having watched Forest for many years :D )

I'm abandoning this particular idea. I'm trying to dabble in the dark art of statistics without knowing enough about the maths behind it. Generally I enjoyed maths at school but statistics bored me to tears. I just wish I was into betting back then.

It's a shame bookies don't offer odds on the number of shots per game then the last few hours of my life wouldn't have been a waste of time :D

cairod
Group 2 Class
Tips
Posts: 103
Joined: Fri Sep 13, 2013 11:00 pm
Location: Mow Cop

Re: Poisson Distribution - This Can Help You

Postby cairod » Sat Sep 09, 2017 6:59 pm

OUCH! My whole theory is on it's arse at the minute!

cairod
Group 2 Class
Tips
Posts: 103
Joined: Fri Sep 13, 2013 11:00 pm
Location: Mow Cop

Re: Poisson Distribution - This Can Help You

Postby cairod » Sun Sep 10, 2017 10:32 am

This weekend's EPL and Championship Poisson predictions:

Man City v Liverpool, away win at 4.1, 32% likelihood;
Southampton v Watford, away win or draw at 2.25, 56%;
Stoke v Man Utd, no to BTTS at 1.9, 64%;
Fulham v Cardiff, away win at 3.6, 43%;
Leeds v Burton, less than 2.5 goals at 1.83, 75%;
Leeds v Burton, no to BTTS at 1.83, 75%;
Norwich v Birmingham, home win at 1.87, 69%;
QPR v Ipswich, away win at 3.8, 34%;
Reading v Bristol, home win at 2.45, 61%
Well that was a bit poor overall. I'm annoyed about backing Leeds v Burton <2.5. Leeds looked very good against Forest in their last game and it shouldn't have been a surprise that they took Burton to the sword. The Liverpool bet was always about the value so no real surprise there. Stoke scoring 2 was a surprise though.

I've been a little more cautious with today's and shown more due diligence with regards to checking team news and previous encounters.

Burnley v Palace, over 2.5 at 2.45, 56%. I know Burnley games aren't usually a goalfest (writing off the 3-2 v Chelsea because of the double sending off) but Palace away games were last season and they look very likely to concede at the moment IMO.

Deportivo v Sociedad, over 2.5 at 1.925, 58%.

Celta Vigo v Alaves, Away win at 5.4, 45%! I'm sure I've missed something here but I'll take the risk at those odds.

Villareal v Betis, home win at 1.909, 73%. Decent home team and did the double over them last season
Villareal v Betis, over 2.5 at 2.05, 59%. Plenty of goals in Villareal home games against lower placed opposition and plenty in Betis away games.

Udinese v Genoa, home win to nil at 4.2, 42%. Last season's home and away form against respectively higher and lower placed opposition speaks for itself and Udinese won this fixture 3-0 last time.

These last couple of rounds have reminded me that it's best to use the Poisson method as a guide and back your bets up with with knowledge, team news and gut instinct rather than just blindly following it.

davidg3907
At Stud
User avatar
Betting Blog Tips
Posts: 4854
Joined: Thu Jun 25, 2009 11:00 pm
Location: Oadby , Leicestershire

Re: Poisson Distribution - This Can Help You

Postby davidg3907 » Tue Sep 12, 2017 10:16 am

The results from the top percentages in the various categories this week are listed below.

Average Goals top.
5 2 3 3 3
Average Goals bottom.
4 1 1 4 2
Over 2.5 Goals.
Yes Yes No Yes
Under 2.5 Goals.
Yes No Yes No Yes No
Home teams winning to nil.
Yes Lost Yes
Away teams winning to nil.
Drew Lost Yes Lost Lost
BTTS Yes.
No No Yes
BTTS No.
No Yes No Yes No No

The figures are still fairly heartening considering the little data on which to base them early in the season. Some trends are continuing, but until the number of games with BTTS and Over 2.5 goals get closer to expected tolerances, nothing will be really clear.

Analysis.

Of top 24 for average goals, 13 have produced 3 or more, 7 with 2, and 4 with 1. No bore draws.
Of the bottom 16, only 4 have produced 3 or more goals, 5 with 2, and 4 with 1. All 3 bore draws fall into this category.

Over/Under 2.5 goals follow a similar pattern with 13 of top 24 producing 3 or more goals,

Home win to nil. Of the 11 with 32%+, 7 won to nil. Of the 16 with 23-32%, 6 won to nil. Of the 13 below 23%, only 2 won to nil.
Away win to nil. Of the 23 with 15%+, 8 won to nil. Of the 17 below 15%, just 1 won to nil.

BTTS Of top 22, just 8 have returned a YES. Of the bottom 17, only 4 have. This has a lot to do with the relative shortage of BTTS matches (12/40), where the 10 year average is virtually 50%.
LAYING Taking it further
LAYING Delving deeper - NEW

FOOTBALL blogs under Blogs/Football
New RELEGATION diary in Blogs/systems

davidg3907
At Stud
User avatar
Betting Blog Tips
Posts: 4854
Joined: Thu Jun 25, 2009 11:00 pm
Location: Oadby , Leicestershire

Re: Poisson Distribution - This Can Help You

Postby davidg3907 » Tue Sep 12, 2017 10:57 am

Retaining last week's format for ease of reference, the top percentages in the various categories this week are listed below. Matches involving Huddersfield, Brighton, and to a lesser extent Newcastle are based on a reduced amount of available data, and then coupled with estimates.

Average Goals top.
4.06 Chelsea v Arsenal 3.51 Liverpool v Burnley 3.01 Tottenham v Swansea 2.89 Man Utd v Everton 2.64 Watford v Man City
Average Goals bottom.
1.82 Huddersfield v Leicester 1.92 Crystal Palace v Southampton 1.95 Bournemouth v Brighton 2.34 Newcastle v Stoke 2.44 West Brom v West Ham
Over 2.5 Goals.
73.42 Chelsea v Arsenal 66.69 Liverpool v Burnley 56.25 Tottenham v Swansea 53.39 Man Utd v Everton
Under 2.5 Goals. (The quoted figures are for OVER so deduct from 100)[/color]
27.67 Huddersfield v Leicester 29.85 Crystal Palace v Southampton 30.26 Bournemouth v Brighton 40.96 Newcastle v Stoke
43.02 West Brom v West Ham 48.37 Watford v Man City
Home teams winning to nil.
43.95 Man Utd 41.32 Liverpool 38.19 Tottenham 37.20 Bournemouth 28.37 Newcastle 26.69 Chelsea
Away teams winning to nil.
36.19 Man City 34.34 Leicester 30.61 Southampton 21.76 West Ham 16.27 Stoke 14.69 Brighton
BTTS Yes.
63.60 Chelsea 51.15 Liverpool 48.48 Tottenham
BTTS No. (The quoted figures are for YES so deduct from 100)
27.01 Huddersfield 29.69 Bournemouth 34.29 Crystal Palace 42.37 Man Utd 45.13 Newcastle 46.31 Watford 46.93 West Brom
LAYING Taking it further
LAYING Delving deeper - NEW

FOOTBALL blogs under Blogs/Football
New RELEGATION diary in Blogs/systems

Who is online

Users browsing this forum: No registered users