The BTC Football Trading Thread
-
@lee-woodman said in The BTC Football Trading Thread:
@richard-latimer Saying ‘more data is better’ is just a fallacy really because bad data can confuse things and actually have a negative effect. I would encourage anyone who is learning to trade to learn how to use data and how to understand it’s pitfalls.
The league average surely contains somewhere between 97% and 99.9% of information which doesn’t even hit your filter….that absolutely cannot be relevant
Regarding regression to the mean, there is no way to establish what the mean is from that and also you don’t know what timeframe it would regress to the arbitrary mean figure anyway
If it helps you over a long period of time then great but I always just consider whether a pro would consider it relevant and I’m certain they would see it as too flawed
What say you two?
For me, it's more a means of watching the data unfurl. I will post up what I'm talking about later but since I noticed the correlation it has saved me a bunch of money!
-
@richard-latimer Saying ‘more data is better’ is just a fallacy really because bad data can confuse things and actually have a negative effect. I would encourage anyone who is learning to trade to learn how to use data and how to understand it’s pitfalls.
The league average surely contains somewhere between 97% and 99.9% of information which doesn’t even hit your filter….that absolutely cannot be relevant
Regarding regression to the mean, there is no way to establish what the mean is from that and also you don’t know what timeframe it would regress to the arbitrary mean figure anyway
If it helps you over a long period of time then great but I always just consider whether a pro would consider it relevant and I’m certain they would see it as too flawed
-
It's also not necessarily due. It's more......apply caution.
-
Not sure how mean average over a full season is irrelevant?
-
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer sorry didn’t realise you raised manually thought it was from all you’ve done over time.
These days there are ways of downloading from the internet
Be my guest. This isn't my area
-
@lee-woodman said in The BTC Football Trading Thread:
I think you really need to be careful how you use and apply this data. You can’t really say ‘regression to the mean is due’ because firstly the mean figure quoted isn’t relevant over such a small sample size. And even so it means nothing when talking about the league generally when 99% of the games won’t be on your filter!
It gives a snapshot and it continues to play out over and over again. All leagues are viable but when looking at 90% over 1.5 I've found only the highest scoring leagues are profitable.
Case in point Germany 2. It's one of my favourite leagues but the highest over percentages have been pish on my data.
To me it's logical. Some teams defy the league average....most don't!
If it stops me making bad decisions..... and it has already......it's far from meaningless.
I would say any data you collect is far from meaningless.
Singapore very high odds as we know. You see these averages you can understand why!!
-
@richard-latimer sorry didn’t realise you raised manually thought it was from all you’ve done over time.
These days there are ways of downloading from the internet
-
I think you really need to be careful how you use and apply this data. You can’t really say ‘regression to the mean is due’ because firstly the mean figure quoted isn’t relevant over such a small sample size. And even so it means nothing when talking about the league generally when 99% of the games won’t be on your filter!
-
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer said in The BTC Football Trading Thread:
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer Its meaningless without the number of games all it needs is one countifs formula copied down - apologies my day job taking over
No, because I collected the averages manually. 100 games+ a season wouldn't seem meaningless to me
But people looking at that report won’t know that without the number of games showing there is no context.
I can try and automate the file and make it easier for you going forward if it helps
Just realised, you have potentially misinterpreted my post. These aren't from my filters.
They are overall averages for the leagues. Anybody can obtain them the way I have. I just thought it would be helpful for people to see.
Mostly I got the from thestatsdontlie.com
-
@richard-latimer said in The BTC Football Trading Thread:
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer Its meaningless without the number of games all it needs is one countifs formula copied down - apologies my day job taking over
No, because I collected the averages manually. 100 games+ a season wouldn't seem meaningless to me
But people looking at that report won’t know that without the number of games showing there is no context.
I can try and automate the file and make it easier for you going forward if it helps
-
@lee-woodman said in The BTC Football Trading Thread:
There always seems to be one league doing the dirty on my filter….German 2nd Division is a shower of crap this month!
Rrrr don't talk to me about that league
-
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer Its meaningless without the number of games all it needs is one countifs formula copied down - apologies my day job taking over
No, because I collected the averages manually. 100 games+ a season wouldn't seem meaningless to me
-
@richard-latimer Its meaningless without the number of games all it needs is one countifs formula copied down - apologies my day job taking over
-
@chris-osborne said in The BTC Football Trading Thread:
@richard-latimer Nice work Richard - Sorry to be a pain, can you add additional columns to show the quantity of games per year please would give more context.
I know there not judging from how thorough you are but to others they could be over 30 games or 5 games .
They are all full seasons so minimum of 100 games a season unless cancelled for covid. Took me long enough to get that, sorry
-
@richard-latimer Nice work Richard - Sorry to be a pain, can you add additional columns to show the quantity of games per year please would give more context.
I know there not judging from how thorough you are but to others they could be over 30 games or 5 games .
-
Left to right in order of year by the way so left is the previous year. In bold is average of the last 3 seasons.
-
@richard-latimer said in The BTC Football Trading Thread:
I've been collecting 3yr averages of goals in most leagues. I'll post it in a bit.
Maybe this will be beneficial for a few. For me it shows the leagues where a high % over 1.5 is less likely awaiting an imminent return to the mean and those leagues where seriously high numbers are very unlikely to continue.
-
@keith-anderson said in The BTC Football Trading Thread:
SHG in the Union Berlin v Augsburg match. 2-2 at HT
Finished 2-2. Union hit the crossbar and the post in the second half
-
I've been collecting 3yr averages of goals in most leagues. I'll post it in a bit.
Trading Spreadsheet linked below
https://www.dropbox.com/scl/fi/3gnk1vku1krkoyfcq37y6/Trading-Bible.xlsx?rlkey=01eqhp8u73s61iy2s814sqfll&dl=0 -
SHG in the Union Berlin v Augsburg match. 2-2 at HT
le jeu avant tout
Am Yisrael Chai
When you got skin in the game, you stay in the game
But you don’t get a win unless you play in the game...
