Football Data (CSV, JSON) - UPDATED 16/08/17

[Euler]

To the forum's credit, we've actually ended up with a stunning set of data to play with.

[Sickmund]

if you are considering football data and analytics from a historical perspective = starting prices > i can advise that the edge is zero on the basis that

expectation of a goal is dependent on the current score added to the time of the opening goal in an individual league and as a result the individual strength of a team can be weakened / strengthened as time decays in a game <

if you model focus is starting prices then I can assure you that your edge will be zero

If I would use your model, the first (simpele) thing I would look at is what happened in the past when a strong favourite goes a goal behind. As you say; percieved strength can weaken after a goal (in my book they are not weakened but 1-0 behind). The interesting thing to look at are (short) starting prices for the team going behind.

But thats my 2 cents and I will leave it there.

[Euler]

The chance of a home team conceding first is quite high in comparison to their starting odds. But that also means that the chance of them conceding first then still winning is higher than most people suspect.

[Sickmund]

Some quick calculations, not to use as a system:
EPL seasons 2011-2016
Full time result after 1st Goal by Away Team in 0-20th minute vs Home Team with odds between 0.00-1.60 (39 matches):
Home win: 56,4%
Draw: 10,3%
Away win: 33,3%
+2.5 gls: 82,1%
BTTS Yes: 82,1%

Full time result after 1st Goal by Away Team in 0-20th minute vs Home Team with odds between 5.40-9.80 (40 matches)
Home win: 5,1%
Draw: 12,8%
Away win: 82,1%
+2.5 gls: 66,70%
BTTS Yes: 61,5%

Odds acquired by football.data.co

[spreadbetting]

Without knowing the odds after the goal was scored the data is pretty much pointless as I'd guess the odds after the goal would be very much in line with the actual results.

[Dallas]

So i had a closer look yesterday at this 'survival analysis' - for it to be true and more importantly useable as an edge the end result will have no correlation to their starting odds, in fact no matter how the starting odds are grouped the end W, L & D % should come out identical.

I can now confirm with some figures (what i knew already) and that is the end results differ widely and are connected to the starting prices.

For example; Home team opens scoring in first 10 mins
When starting odds 1.01 - 1.50 =
Result = W 90.32%, D 7.74%, A 1.94%

When starting odds 2.01 - 2.51
Result = W 65.65%, D 17.56%, A 16.79%

[deansaccount]

So i had a closer look yesterday at this 'survival analysis' - for it to be true and more importantly useable as an edge the end result will have no correlation to their starting odds, in fact no matter how the starting odds are grouped the end W, L & D % should come out identical.

I can now confirm with some figures (what i knew already) and that is the end results differ widely and are connected to the starting prices.

For example; Home team opens scoring in first 10 mins
When starting odds 1.01 - 1.50 =
Result = W 90.32%, D 7.74%, A 1.94%

When starting odds 2.01 - 2.51
Result = W 65.65%, D 17.56%, A 16.79%

So in your opinion 'survival analysis' does not exist?

[Dallas]

So i had a closer look yesterday at this 'survival analysis' - for it to be true and more importantly useable as an edge the end result will have no correlation to their starting odds, in fact no matter how the starting odds are grouped the end W, L & D % should come out identical.

I can now confirm with some figures (what i knew already) and that is the end results differ widely and are connected to the starting prices.

For example; Home team opens scoring in first 10 mins
When starting odds 1.01 - 1.50 =
Result = W 90.32%, D 7.74%, A 1.94%

When starting odds 2.01 - 2.51
Result = W 65.65%, D 17.56%, A 16.79%

So in your opinion 'survival analysis' does not exist?

Not in the way Jonnyg is suggesting.

ie, 50 goals have been scored by the home team on the 8th min, the home team goes onto win 25 of them games therefore the odds of the home team following a opening goal on 8mins should be 2.0 = survival analysis

[deansaccount]

So i had a closer look yesterday at this 'survival analysis' - for it to be true and more importantly useable as an edge the end result will have no correlation to their starting odds, in fact no matter how the starting odds are grouped the end W, L & D % should come out identical.

I can now confirm with some figures (what i knew already) and that is the end results differ widely and are connected to the starting prices.

For example; Home team opens scoring in first 10 mins
When starting odds 1.01 - 1.50 =
Result = W 90.32%, D 7.74%, A 1.94%

When starting odds 2.01 - 2.51
Result = W 65.65%, D 17.56%, A 16.79%

So in your opinion 'survival analysis' does not exist?

Not in the way Jonnyg is suggesting.

ie, 50 goals have been scored by the home team on the 8th min, the home team goes onto win 25 of them games therefore the odds of the home team following a opening goal on 8mins should be 2.0 = survival analysis

Thanks for the info Dallas. Do you trade football often?

[Dallas]

Thanks for the info Dallas. Do you trade football often?

I do but only the main European leagues, but its still Horse racing and Tennis that get my main focus

[LinusP]

I have come to a similar conclusion (so far) but to give him credit he has given me the kick to look into football.

[Dallas]

Y, TBF to Jonnyg his posts have prompted some interest in 'goal timings' and even if just one person finds something of use they can use then that's another positive

[jonnyg]

I have certainly raised the football data abad analytical bar > as far as I am aware I am the only entity to offer in play early goal data and analysis which is at a very high level > if I say so myself

in terms of survival analysis > certainly not my fault as I have explained this many times < I can see that people have struggled to understand what I am saying < so one more time for this thread >

I discovered that the betting industry do not understand survival analysis which is looking at how a team in an individual league hold onto their advantage < the mainstream believe that a team that concede are more likely to score more goals ... you can check the research via google >

I found this is incorrect and you need to look at individual leagues >

in terms of the Premier league and early home goal < Dallas talks about starting prices but has failed to understand 2 key areas >

1. my focus is on survival analysis in terms of expatiation of the pathway `< will it go to 1-1 ? > 2-0 ? or stick on 1-0 ??

the starting price is the red herring which is a key point < so for example when Burnley opened the scoring at home to Liverpool > early home goal < we had Liverpool at favs to beat Burnley and Liverpool at around 2/5 to score goal 2 < in my in play commentary i advised why > 2-0 would be just as likely >

2. is a massive key point which very few people have grasped >

expectation of a goal is dependent on the current score added to the time of the opening goal in an individual league as time decays <

which can weaken the perceived strength of a team > for example have a look at Man City away when losing 1-0 HT game state < have a chronic historical record


if you advise that survival analysis does not exist > you would need to detail what you understanding is of survival analysis >


naturally people do have a different interpretation of football data but my in play analysis is so strong that my ideas and analysis has become very difficult to brush aside

[jonnyg]

So i had a closer look yesterday at this 'survival analysis' - for it to be true and more importantly useable as an edge the end result will have no correlation to their starting odds, in fact no matter how the starting odds are grouped the end W, L & D % should come out identical.

I can now confirm with some figures (what i knew already) and that is the end results differ widely and are connected to the starting prices.

For example; Home team opens scoring in first 10 mins
When starting odds 1.01 - 1.50 =
Result = W 90.32%, D 7.74%, A 1.94%

When starting odds 2.01 - 2.51
Result = W 65.65%, D 17.56%, A 16.79%

will post some data in a moment to show you where you have gone wrong >

what price were Burnley to beat Liverpool before the game started ? and what price were liverpool when Burnley opened the scoring ?????

[jonnyg]

in summary

dont listen to me or Dallas or anyone else on here > get started with your own research < forget automating your betting / trading till you have found some strategies that give you an interest and an edge >

I can say with 100% expectation that randomly backing / laying the draw < backing a further goal on 80 in a random game , backing the draw in a game that is 1-0 HT when a team winning is Evens - 5/2

is not going to get you anywhere >

i am aware that accrington last 28 home games in league 2 when opening the scoring in 0-10 minutes time band and

22-2-4 < this is not enough for me < I discovered that a draw in league 2 is very rare when the home team open the scoring 0-10 < still not enough < I then look at why that would be >>>>>>>>>>

best of luck >