The BTC Football Trading Thread
-
Thanks Martin much appreciated. I also have Ross County on my LTD
Current strike rate is 86% since July but had 5 draws last week, which Iām putting down to post World Cup syndrome. It did pick up towards the end howeverā¦ -
@stuart-wallace-0
I started matchbetting in January 22. When I got banned from every bookermaker I transitioned into trading with a Ā£200 bank left over from MB in April 2022. Started with Ā£1 stakes, now at Ā£10 stakes and I have grew my bank to just over Ā£1k which Iām happy with. So Iād call myself a noob/rookie . You ?
-
Anyone done any stuff on Managers stats ? Was thinking about it because I belatedly saw Pirlo was a boss in Turkey and if his style of play in Italy was anything to go by its an "open door" policy .. quick look at some stats from key managers would be as you'd expect career wise but interesting nonetheless especially to factor with any managerial changes.
Arteta For;Against 2/.82
Mourinho 1.41/.93
Guardiola. 2.44/.91
Ten Hag 1.88/1
Conte 1.73//1.3
Klopp 1.94/1.26Only surprise is I'd have expected Conte to concede less ? .. Thought the recent Spurs conceding was abnormal but perhaps not
Pirlo .. 2.15/1.98!!!... guaranteed nil nil now
-
@mark-maguire said in The BTC Football Trading Thread:
@stuart-wallace-0 Other than the two hours on busy days and an hour on other days I try and allocate specific quite/non trading days to take a completely fresh look at my spreadsheets.
At other times I'll get an idea.. often from someone posting something here and it will get me thinking so I will create a formula or two to explore the idea.. I think thats why I'd encourage anyone starting out to record everything .. game, date, match time, start odds, goal times, ratings, results gap between goals, total goals.. because later on when your confidence grows the ability to go back and interrogate that stuff, even stuff you never knew you'd need .. is invaluable
I have an idea to develop something for match odds but until I've cracked the over 1.5 and under 3.5 100% I'm not going to do anything with it. Too many pies in the air and you drop them all!
-
@stuart-wallace-0 said in The BTC Football Trading Thread:
How many hours a week do you think you guys spend on your data and analysis on a usual week.?
Hope you guys have got your spreadsheets backed up
Less now. More when I notice something that needs checking. Currently it's as I update.the days selections. If I hit a loser I need to check to make sure I would do it again.
-
Limited trades on the filters today ..
Arguing with myself about the impact of Transfer deadline day on Benfica away at Arouca and Fernandes maybe leaving..
Its on my LTD filter. Will look at team news before final decision. my odds say the draw should be 3.45 so will at the very least wait for odds which are at 6 for the draw to come in before considering a trade.. if I miss out I miss out no harm done.
O2.5
Karagumruk vs Besiktas (Split stake out after 60 mins if no goal)
Will also be FHG but won't go big stakes on it.
01.5
TNS vs Bala
-
@stuart-wallace-0 Other than the two hours on busy days and an hour on other days I try and allocate specific quite/non trading days to take a completely fresh look at my spreadsheets.
At other times I'll get an idea.. often from someone posting something here and it will get me thinking so I will create a formula or two to explore the idea.. I think thats why I'd encourage anyone starting out to record everything .. game, date, match time, start odds, goal times, ratings, results gap between goals, total goals.. because later on when your confidence grows the ability to go back and interrogate that stuff, even stuff you never knew you'd need .. is invaluable
-
@nick-allan said in The BTC Football Trading Thread:
@richard-latimer It's a very good point Richard, I think the difficult thing for me is finding those correlations. It's easier in my opinion to look at leagues as that information is static it's either performing in a league or isn't. But when it comes to other data analysis I do tend to struggle. What I've done in the past is ask myself āwhat am I looking for in this tradeā, if it's over 2.5 goals am I looking for a strong 2.5 G percentage for both home and away, am I looking at goal averages, am I looking at home performances, and I end up recording so many pieces of data but then when it comes to analysing it I'm none the wiser. And I think this is the issue with me is that I just don't know how to breakdown the data to filter out selections. It could also be a time issue of course, I work full time in quite a demanding job but I'm willing to put in the time. I donāt have any kids and Iāve been trading sports for over 3 years so im learning everyday
I always hated looking at leagues and then for a while I did and I found that I was removing more and more leagues as an excuse for something not working.
Now I only remove the most corrupt as detailed on the global corruption index.
The only other way I look at leagues now is purely on data alone. Ave goals in said leagues and I don't rule any out. I just make more allowances for some higher numbers in higher scoring leagues.
The other thing to say is that if I don't do this, I hit a loser which I find out I could have avoided it really annoys me. On the flipside if I come at it from every possible angle and still can't see a reason why I shouldn't trade it......and it loses......I can take it on the chin much easier. Even if I then have to reconsider it's longterm effect there was no reason not to trade it at the time.
-
@mark-maguire said in The BTC Football Trading Thread:
@connor-garland agreed .. what I have done over time Is tweak filters .. for example adding relative ELO differences.. in those cases there is a reference point in my analysis to compare the filter pre and post tweak. I've never actually had to drop a filter .. but have tweaked the weakest. Ironically in one case I did this to the point where the number of selections is relatively small but the SR is amongst the best of my filters.
I've scrapped more filters than I care to think. What has changed it for me is being able to filter out specific info I want to see. Too many trees and you can't see the woods.
I've still made the odd mistake actually deleting data that I then think I should be collecting again and so shoot myself in the foot but the game change for me is subtotals in excel and learning how to use.
I take a good old look at the rabbit hole even if it's just a few fixtures which have had a negative impact. I take them out of the firing line with their respective data, watch and wait.
I have never had to put them back in. They almost never come up or they do and I save myself another loser.
The more data I get the more robust my system becomes.
My core data is ave goals, home ave, away ave, home ave conceded, away ave conceded, h2h over/under %, h2h 0-0 and when they last happened, last 10 over/under %, last 10 over/under and 0-0, when they happened.
I combine some of the individual related stats as home ave 1.1 may be good enough but not if the away ave is only 1.1 as well. Same with high numbers, on their own great but combined with other high numbers...hmmmm too much!!
-
@martin said in The BTC Football Trading Thread:
@connor-garland your spreadsheet is a work of art!
Thank you, Iāve definitely spent far too many hours on it but itāll all be worth it one day
-
@connor-garland agreed .. what I have done over time Is tweak filters .. for example adding relative ELO differences.. in those cases there is a reference point in my analysis to compare the filter pre and post tweak. I've never actually had to drop a filter .. but have tweaked the weakest. Ironically in one case I did this to the point where the number of selections is relatively small but the SR is amongst the best of my filters.
-
@richard-latimer couldn't agree more .. rabbit holes are aa really important place to visit for me..obvs not in my personal life .. I deny any such accusations
-
@Martin wow don't know what I've had done without the software .. its an interesting one for me because mentally I need that process and procedure because by the time I have done the stats on each game I feel I have created some perspective on the various opportunities and their value .. if it was all automated I fear I would lose some of that edge if that makes sense?
-
@andy-donnelly said in The BTC Football Trading Thread:
For me itās all about striking a balance with regard to analyzing and then tinkering your strategies. There is a danger that over analyzing and over-tinkering can lead you down rabbit holes and effectively mean your are back fitting as well. I had this problem last year on an over 1.5 filter which Iāve now scrapped.
Canāt wait for the football back testing because if itās anything like the horse racing it will be a game changer and make it so easier to test any changes you make along the way .
I 100% agree. I like to think that I look down the rabbit hole without falling in. I don't think back fitting really applies to what I'm talking about as if I can't find the slightest logical reason to remove it it stays in. If all the data falls in the middle of my existing dataset it stays in. It's when some numbers are on the outside so clearly a rare breed. Too high or too low. It works for me and despite the time I've put in to get to this point, I miss countless bad eggs because of it. I miss plenty of good stuff too but until I've got the data to back it up I'll cope with that.
It's an ongoing and fluid process. 50% of my selections stay on the watch and collect data list but it's getting less over time. In the end I'll want to be at 100% with a super tight filter that let's only the best stuff in and can be checked in minutes after copying lines over to my sheet.
-
@andy-donnelly said in The BTC Football Trading Thread:
For me itās all about striking a balance with regard to analyzing and then tinkering your strategies. There is a danger that over analyzing and over-tinkering can lead you down rabbit holes and effectively mean your are back fitting as well. I had this problem last year on an over 1.5 filter which Iāve now scrapped.
Canāt wait for the football back testing because if itās anything like the horse racing it will be a game changer and make it so easier to test any changes you make along the way .
I used to do this a lot. Chopping and changing a filter every 1 or 2 weeks and results werenāt consistent, but neither was my behaviour.
I decided to leave it and let variance take its course. To make my current strategies, I have backfitted but not made it overly complex and it seems to be working on the forward testing.
Granted Iām only 2 weeks in to the forward test but 15 points up whilst underperforming strike rates when compared to the backtest isnāt too shabby in my eyes.
-
@mark-maguire I'm the same, Saturday and Sunday mornings are probably my busiest times! Used to be even crazier before the software, I would go through every Saturday and Sunday game on soccer stats on a Friday (I had that day off work) would be a whole day's work, sometimes the next day too.
-
@connor-garland your spreadsheet is a work of art!
