So there’s a football match coming up

you’re looking at all of the stats and home team’s won four out of four the way

teams lost four out of four so you’re sort of thinking oh maybe our favor this

team what stats should you use if you’re looking at a football match that’s what

I’m going to talk about in this video click here to subscribe to our YouTube

channel you’ll get instant notification of new videos as they’re released so if

you go to a website like the BBC Sky pretty much anybody now you’ll see at

the bottom of a match preview you know key stats it says and the key stats will

say something like so-and-so has won when they played away at this venue they

haven’t lost or they haven’t won or it’s been a you know a large number of draws

or they don’t score this many goals and all these key stats pop up at the bottom

telling you things that are relevant to that particular match but they’re not

because generally the stats are historical items that bear no relevance

to the match that’s in front of you matches between teams vary and interlock

in quite complex ways and looking at historic results such as you find on

websites or at the bottom of summaries and match previews and stuff like that

very often don’t bear any relevance to the outcome of the particular result so

you’ll be interested to know that I don’t use them it’s interesting really

because if you go back many years ago I started digging around inside football

interests and try to understand how football matches should be priced and

had the outcomes of matches occur and the thing that was surprisingly it was

like discovered all sorts of little interesting things and I wrote to a

magazine and said all you know would you like to do an article on this and I got

a letter back basically saying no because it’s incredibly dull a mean and

interesting a you look at stats today and they’re everywhere there’s so many

of them you’re sometimes sport for choice and there’s just hundreds of

sites detailing loads of stuff but unfortunately most of it is completely

irrelevant unfortunately and part of the reason for this is that most of the

stats that generated a based upon historical data they inherited bias in

the way that they’re interpreted but also the creation of these stats is

there to stimulate you to do something so typically you know people offer bet

stimulus services where they throw stats a house that will get people to actively

engage just like so-and-so has done this it’s like oh yeah so they have right

I’ll have a fiver on that and that’s where a lot of their stats come from

they’re not actually interested in the prediction of the outcome of the match

at all really and the fact is most people aren’t most people just can’t be

bothered to go through that pain and suffering that comes with accurately

trying to predict the outcome on a ferment they just want some simple

little stats and then they’ll jump on them and away they go now the fact is

that there is an underlying core that sits within a football match that allows

you to pick out any number of different types of interesting statistics from the

match so using two variables I can actually describe the chance of the home

team winning any of the correct scores the away team winning the chance of a

draw when their first goal is likely to be scored if that first goal is scored

by whom if the home team is one and a half time it will describe the chance of

the away team coming back into the match at some point there I had the number of

corners the number of likely shots on-target

yeah but ever they’ve of it just two variables describe the whole thing but

you very rarely see people actually talking about them in any great deal of

information but before I talk about those let’s go back to where we suffer

from in terms of the historical stats so let’s say we have an unlimited deck of

cards let’s say that we can just deal as many cards as to like infinitely into

the future there’s an infinite pack if we keep turning over those cards it’s

very likely likely at some point a sequence will occur so we’ll go black

red black black red red red black black black black black black black black

black red red red red red red red red but the fact is that every time that

sequence occurs when you turn the next card in an infinite deck of cards

there’s no increased chance of a black or a red it’s completely random as long

as you’ve shuffled the deck and there’s no funny business going on

now I’ll actually got a great follow-up video to show you how I get completely

random deck of cards can sort of be made under and ‘im by well I won’t go into

depth as to how it’s done but basically I guarantee you if I show you this

method you can win money from anybody despite the fact that the turn of the

next card is completely random and it’s all part to do with the way that people

mix up sequences but also you know the gamblers fallacy is a common thing

that’s described in mathematics and probability and it’s basically the

feeling that if you have a sequence of long periods of red or black that

somehow that influence is the outcome of the next turn of the card and the fact

is it doesn’t just the same as when you’re in the roulette and there are six

blacker blacks come up it doesn’t influence the outcome of the next red or

black and as a consequence the gamblers fallacy permeates permanently you know a

string of winning favorites people start betting on the eighth favorite a string

of losers will alter people’s behavior but the fact is that what’s happened in

the past does not necessarily predict what’s going to happen in the future but

a lot of people will bet on it and that’s where the stats on football do

lead you astray because they’re designed to stimulate you to perform an action

and that is because of the sequencing of patterns and trends within it when in

fact that’s an irrelevance in terms of where you’re going to predict the next

match so what are the two key variables that I use to predict everything within

a football match well it’s really simple it just boils down to goals how many

goals is the home team gonna score how many goals is the away team gonna score

because if you can accurately predict both of those things which you can do

then everything works out from within them so if the home team is going to

score X number of girls in their way team’s going to score Y if that if the

way team scores a goal then that metric is still present within the match and

then the home team is expected to score this many goals and therefore the chance

of them coming back into the match is going to be related to the difference

between goals between the two teams and how many goals have been scored and how

much time is left on the clock so you can actually predict all of that just

using then of goals that home team scores and the

amount of girls in the waiting school so when I do the football ratings that’s

what I’m doing I’m working through all of the detail in terms of how many girls

are likely to be allocated to the home team how many are likely to be allocated

the way team and how many going to be scored overall and from that key

information everything spins off of that so if you don’t believe me well I’ve

been doing it for god knows how many years and it works beautifully but the

fact is most stat sites are not focused on that at all so if you know how many

goals are likely to scored you can forecast nil-nil one or two nil three

nil you can group all of those scores together to create the match odds market

over and unders and then you can start to look at other things such as when the

time of goal is going to occur and if a goal occurs at this time what’s the

likelihood of another goal at this point until the end of the match so yeah if

you’re gonna research and look at football in great depth then I suggest

that’s where you put your focus in studying intensely how many goals are

going to be scored and who is likely to score them look at these scoring stats

ignore the results because the results don’t really tell you that much

information because it could have been played against any number of teams and

you can go many many levels deeper but when you look at football essentially

what I do is just focus on the number of goals that are likely to be scored and

who’s going to score them that tells me absolutely everything that I know from

that I can predict its time of first goal what happens if a goal goes in what

the chance of the other team coming back into it when the match is underway I

then start looking to see if my projection before the match is matching

what I’m actually seeing during the match because if it isn’t is whether if

it isn’t then I need to reassess everything that I’ve done and look at it

from a different angle if the match is turning out differently from what I

originally forecasted so yeah when you’re looking at football you can

pretty much ignore all sorts of other stats such as teams beginning with the

letter B tend to be teams beginning with the letter L because it’s an irrelevant

so the fact that so-and-so has won X amount of times in a row or hasn’t won

on this ground for 20 years what you need to do is focus right in on those

basics and that will give you everything that you need to know just excessively

fork what’s going to happen in a football

match please like or dislike the video or leave a comment below that will allow

me to produce better videos and more of them in the future

I feel like I may be missing somthing but… If past statistics don't give any indication of future performance, then why would past goals scored statistics be relevant to future outcome of number of goals scored? Or do you formulate your goals scored projections by other means – aside from how many goals a team has scored previously?

Great video Peter – You speak very well

You don't explain how you determine the number of goals that will be scored in a particular match. If you are not looking at past stats where are you looking to determine number of goals that will be scored?

Peter is saying that past results and history between teams doesn't matter. Of course past goals scored does matter. How you come up with goals likely to be scored varies wildly though, and therein lies the art. How many past games do you look at? Home or away? How many goals is gone ground advantage worth (avgs close to 0.4 for most teams), do key player outs influence goals scored? Time of day and weather?

How many variables you include to calculate home/away goals is where the models change. Get proficient using Excel…

But once you input the variables, yes, the home/away goals are the only figures that influence prices across a range of markets. Finding the correct numbers is the challenge..

This is contradictory Peter. As Joseph says, if the past is no guide, then its no guide to goals either. No-one can predict a football match. I think you can only trade a match on what you see in front of your eyes on the day, and even that can change by the minute.

Sorry Peter this makes no sense at all. I'm a massive advocate of ridiculing historical stats, but you are using historical stats by using past goals scored. The whole thing seems like a massive contradiction?

can somebody share a profitable strategy for trading soccer matches ?

Peter you should make a video showing an example how you calculate this not speaking in general idea of football ratings.

The historic stats would only have importance IF the teams were containing the same players or containing the most important players. For example, in the case of the recent Arsenal Chelsea (3-0) game, Chelsea had a different manager and some different players etc, so the historical stats would not be relevant IMO.

makes sense, but you dont explain how to work out the number of goals each team may score, a pretty crucial point has not been discussed

Hi, Peter. Thanks for the Information. It would be great stuff, if you explain this topic in more detail and how you calculate the possible outcome of a football match. I don't know how these calculations are made, but I am using sportingindex and spreadex for checking the odds, when I am in a trade pre match … have your calculations something to do with the numbers on total goals and home supremacy from these sites or do you create your own numbers? Kind regards

Hi Peter,can you do a video showing how and where you get you goal stats and other information from.

Bye the way Peter, great videos. Regards Keith

For everyone who is unsure about how to get from predicted number of goals, to the odds of the various possible scores:

You need to go and read up on Poisson Distribution.

or for Excel users %chance = POISSON(Predicted outcome, required outcome). So if you know a team 'usualy' scores 2 then the chance of them getting 4 is POISSON(2,4) …i think, or the other way round 🙂

Then as Peter says, From there you know all the odds on it being 1-0, 2-0, 3-0, 4-0, 5-0 etc … add them up and that's you Home win

ditto 1-1, 2-2 etc for the draw odds… you get the picture

When it comes to time of goal, if the 'usually' score 2, then in any given minute they 'usually' score 2/90 …so to score 1 in that given minute the chance would be Poisson(2/90,1) or Poisson(2/90,0) to not score in that minute. So, if my understanding is right…..the chance of scoring in the 30minute would be 29 lots of Poisson(2/90,0) and then a Poisson(2/90,1)

Poisson is the answer almost all your pricing.

This particular poisson is not a red herring (groan….someone had to do the gag) but I'm going to look like some sort of pilchard if Peter tells me I'm wrong. 🙂

Peter is spread betting total goals and supremacy information a good starting point for assessing potential total goals in a match?

I often say, 'results create stats, stats don't create results'

captivating again thank you Peter

Where do you get the scoring stash from. and how many years should you go back. surely a team who averages 2 goals a game since 1890 has no relevance as they may only average 1 goal a game since 1980. What's the ideal number of years to check the scoring stats.

Peter, can you show us in your next video how you shuffle a pack of infinite cards?

Hi Peter, Can you make a video on this method how to do it or direct me to the video if you already done it ? Thanks.

A "key stat variable" that is defined as "How many goals are the home/away team going to score?" is not a variable at all, it is the result of the prediction model… What are the real key variables?

liverpool v manchester city 7 oct 2018. the two most prolific goal scoring teams in the premier league. final result 0-0

Just look at the odds near kick off if you want to know the percentage chance of a team winning a match? :)? For instance, if 0-0 is deemed statistically unlikely, the odds for 0-0 will be much higher than the average you would expect. 0-0 on average is about 7.5% probability on an average match?

I been wrestling with Football stats for months now, with no particular "ureka" moments as of yet 🙁 lol

I am gonna take complete note of the advice in this video and look into it !

Far too many variables to considering in a sport like football, that's why I tend to stick to a much simpler sports such as boxing or tennis.

But how do you do it