Aug 2024 £3000 - ? Bank Challenge
-
No late goals picks but an overnighter in Titanic if anyone is burning the midnight oil. I won't be haha!
-
There's also a game in Ecuador although I'll be leaving that one well alone until I have more data.
-
It's gonna get quiet but there's a couple of potentias for tomorrow. Let's start off Sept in the green. I was going to say a bang but that will be pushing it haha!
-
@darri said in All Goals Are Equal But Some Goals Are More Equal Than Others:
Sometimes when it comes to trading and iv noticed it here people can massively over complicate things. I think your looking into sample sizes not big enough and just basing opinions rather than letting the data play out a bit more. For me if your basing a filter off of season stats then last seasons data should be void. Bit bullish to say that but so much changes, new signings, new tactics, new managers and new ambitions.
Each criteria for analysing data like this should have a larger sample size. Im not talking about the overall amount of trades but when you get granule and start looking at different criteria such as how many were start of season so between 0-10 games same with league data too you need to allow data to drive your decisions. Sometimes in the past you have jumped or reacted to certain things. 4 losses on the bounce and straight away your analysing just one example. Your doing fine mate. Dont do reaction decisions and dont use small data sizes let them breathe. Paper testing shouldnt be rushed. Only employ money on things once its become almost robotic like in selection and consistency. The late goals is something iv been trading for a while now and like to think im fairly well educated in them, make sure your looking at all the variables within the game not just the filters.
I'm still tracking them. I just think it should give me pause before putting money down while collecting the data. I know hat you mean the small data samples not being able to give you all the answers but I do think it could possibly act as a way to avoid costly mistakes until the data sample is larger in size. In fairness it's the way I worked on ELO's. What I want to avoid there is removing things after putting money on and rather keep my money in the wallet until I'm happy with certain areas.
-
Sometimes when it comes to trading and iv noticed it here people can massively over complicate things. I think your looking into sample sizes not big enough and just basing opinions rather than letting the data play out a bit more. For me if your basing a filter off of season stats then last seasons data should be void. Bit bullish to say that but so much changes, new signings, new tactics, new managers and new ambitions.
Each criteria for analysing data like this should have a larger sample size. Im not talking about the overall amount of trades but when you get granule and start looking at different criteria such as how many were start of season so between 0-10 games same with league data too you need to allow data to drive your decisions. Sometimes in the past you have jumped or reacted to certain things. 4 losses on the bounce and straight away your analysing just one example. Your doing fine mate. Dont do reaction decisions and dont use small data sizes let them breathe. Paper testing shouldnt be rushed. Only employ money on things once its become almost robotic like in selection and consistency. The late goals is something iv been trading for a while now and like to think im fairly well educated in them, make sure your looking at all the variables within the game not just the filters.
-
Still waffling on. Titanic had one qualifier yesterday. It was a loser but a extremely low odds loser @ 1.1. Prob could have got out even on 83rd minute when a goal came. It ended 1-2.
Not a massive issue but I know comeback results appear poor for some countries as I'm sort of excluding them on the ELO. Hungary is a country where I have one result and one loss in a years selections so if it's not happening on the ELO then I'll not post it for the Titanic either.
-
Just calculated my daily profit tracker for all of August and while it's a sickener to end on such a bad note I'm still 100% in profit for the 3rd month running. You could say it's 1st quarter in profit even since I started tracking again.
Not huge numbers so I'll share the results.
June £13.61
July £17.45
Aug £10.56Still on the right track.
Let's make it a good Sept.
-
Going to the cinema this afternoon but until then I'm going to update some stuff. I mentioned that following my deep analtical session on these I'd managed to right the ship and that continued yesterday. Have an overall yield of over 10% again now since I made those changes which is pushing closer to where I was on the last sheet at around 15%. Missed a decent one in Mexico as well although was a bit late.
Yesterday saw 5 wins and 1 small loss in Sweden. The wins came from Korea, China, Norway, Slovakia and Tunisia. Don't always have a market for it but I've noticed Tunisia seems to be quite good. Slovakia I'm on the fence still but it's ok. Not a definite losing proposition like some.
I noticed yesterday that the match in Chile I thought won must have been ruled out as it would have in fact been a loser.
Splitting the sheet up like this does definitely seem to be the way to go for now.
-
So yesterday I went through all 70+ qualifiers to my over and out filter and this is what I found:
When literally 0 games had been played in the season the results were 5 wins, 2 losses most likely resulting in a small profit. One of these was 2-1 @ HT and if any HT scoreline isn't going to make the grade longterm this is one. With that in mind it makes the results 4 wins and 2 losses from the scorelines I'm currently doing. Again probably still small green as one of those losses had smaller odds to lay. Also one of those losses was Dordrecht yesterday and we'll come back to that one later.
When 1 game a piece has been played (I'm alsom counting it as a new season after some leagues restarted following months on the sidelines) the results are 6 wins and 2 losses.Smalll green again and one of those losses was yesterday with odds close to evens.
2 games a piece - 3 wins, 0 losses. (2 of those have HT scorelines I would presently find favourable)
5 games total - 2 wins
6 games total - 3 wins
8 games total - 5wins, 2 losses (although one of these was 2-1 @ HT and not a scoreline I think will make the final cut)
10 games total - 3 wins, 1 loss. One of those wins was 2-1 @ HT and the loss was Santos yesterday. As with Dordrecht we'll come back to that game later.
11 games total - 1 win
12 games total - 6 wins (5 in the scoreline brackets I like)
16 games total - 1 win
18 games total - 1 win
20 games total - 3 wins (1 HT scoreline not so favourable)
22 games - 2 wins (1 favourable HT scoreline)
Could go on forever there but I think we've reached the point at which it stops being helpful.
So this didn't tell me a massive amount to be honest other than there was no specific reasoning to change anything based on games played so far.
I already ruled that when one team has score in 80% or loss then the other team must have scored in more than that which ruled out a good 4 losers and 4 winners if I recall so a 50/50 split.
Because the other numbers are so different I've been wondering how changing the requirements would make any difference. I tried throwing all criteria together previously and looking for a total to see if anything would present itself from that and it didn't.
Last night I took the conceded criteria which tend to be lower on each team and added them to the % goals scored in H2 to create a total figure from that if you will. Turns out both Santos and Malmo were in the lowest numbers for these so became instantly unappealing. You could say that this is backfitting to a degree if you like but 1) I make rules based on logic and the logic stands here. Less conceded goals across 2 teams and less % of goals in H2 makes a H2 goal less likely.
Before I had been weeding certain games out by using the conditiona formatting colour scale. Sometimes it seems blatantly obvious when there are nothing but low numbers. Other times everything seems to fall nicely in the middle. Malmo was one of those games yesterday. I had reservations from looking at the numbers but couldn't quite put my finger on it and when the equaliser went in and inplay stats looked good I went for it.
It is much easier, certainly for me, to rule games out based on a digit. It either qualifies or it doesn't. There is no grey area at all.
Of the lowest totals when these numbers are added together there are 4 selections. Until yesterday there 2 selections and 2 wins. Now there are 4 selections and a 50/50 split.
A look at the top figures when these criteria are added shows the higher the numbers the more wins so I'm satisfied with the logic. Unfortunately somtimes patterns take a while to develop and intial results can bely the actuality of what willl happen later on down the line.
In addition here (and I don't simply want to remove every loss but find a reason if one can be found for excessive losses) Dordrecht on top of being first game in the season had a 50% splitfor FH and H2 goals. Only 2 selections like that so it's a definite outlier and possibly one I would look to remove going forward so that any game would need to be over 50% rather than 50% or over.
Stuff to mull over. Until later people....
-
@martin-futter said in All Goals Are Equal But Some Goals Are More Equal Than Others:
@richard-latimer personally don't like early season, just from the perspective of knowing teams are yet to hit full match fitness
My data is very much inconclusive on this end for now. Went through all results yesterday looking for patterns and while I managed to find some things to tighten up this wasn't really one of them. After a bunch of poor results I often need to find reasons to make me confident enough to go again if anything and making snap decisions from logical reasoning helps me to do this.
I think I'll continue to track the things I'm getting rid of for posterity and transparency but there are things about 3 of yesterday's 4 losses which made them outliers and I have no problem removing outliers for the greater good.
When I can get to computer later I will go into more detail.
-
@richard-latimer personally don't like early season, just from the perspective of knowing teams are yet to hit full match fitness
-
Monday is quiet which is good. For one thing it's bank holiday and time to spend some quality time with the family. For another I've been analysing the s**t out of my data tonight following 4 losses on the bounce in the 2nd half goal picks.
What I thought was a really good trade in Santos perhaps wasn't so. Data is inconclusive to remove first game of the season on these but Dordrecht may not have been right regardless. Not making any decisions tonight but will come back to it in the evening.
There's one potential ELO in China but I doubt I'll bother. Anyone who wants it, it's on the sheet.
Trading Spreadsheet linked below
https://www.dropbox.com/scl/fi/3gnk1vku1krkoyfcq37y6/Trading-Bible.xlsx?rlkey=01eqhp8u73s61iy2s814sqfll&dl=0 -
Finally, the 'Titanic' filter which is starting to produce a couple of qualifiers again now. The latest was Legia who went 0-2 down and pulled one back (just quick enough for me to take around 50% ROI.
-
2nd half goal picks for Sun:
Waiting for the next loss on these now. Doesn't seem as though I've had one for ages haha! As documented on the football thread I missed Gamba Osaka personally but it's another one for the data. Missed the goal in Paraguay by seconds. Just diddn't quite come down to my price before River Plate got a free kick, the market suspended and goal flashed up. Got luckier in Palmeiras after Bahia had a goal ruled out on 46 allowing Palmeiras to do the honours in the 77th. Still enough time for Bahia to get one at the death but by then my race was run.
-
Sun ELO:
Saturday was a bit of a mixed bag really. Not especially good, not especially bad.
Shandong I saw was around 1.78 to lay the away when they went ahead so ave loss.
Then Pyramids were higher than that initially but got quite low before they pulled it back.
Nacional in Uruguay (I should say, I don't have much from Uruguay but what I do have is good for these) went behind to Fenix and pulled that back. Odds there were 1.79 to lay Fenix when they went ahead.Finally it was another low odds lay in Paraguay although this time it didn't quite come off. I got 1.4 on River Plate.
So I suppose Pyramids negated the loss on Shandong and the loss on River Plate cancelled out over half the green on Nacional. Still, it was definitely green which after last weekend is good.
On another note U. Catolica just pulled off a comeback in Chile for the leagues I'm monitoring but doing nothing else with at the moment.
-
Something to consider moving forward. Not quite sure how I'm going to change things but I probably will to an extent. The lower the goal odds at ht the longer they take to come in further it would seem. Just missed the first since I've been doing it this way where the goal came post 65mins.
Odds of 6.4 at hT will probably be closer to 3.2 after 60-65 mins.
Odds of 3.45 didn't even get to 2.26 by the 67th/68th minute. -
Short priced lay away (if score 0-2):
-
Goal selection (H2 specifically):
-
We made it!!! The weekend is upon us and finally there will be some trades:
Have already split out the good ELO leagues (Won't Stay Down) from the 'Can't Get Back Up' ones.
The good ones.
The sus ones.
-
@richard-latimer said in All Goals Are Equal But Some Goals Are More Equal Than Others:
Going to hope for an opportunity to put this one in Uruguay into Action in a bit:
Need 0-0/0-1/1-1@HT to enter. Also decent enough odds and attacking threat/good enough inplay stats. Once I have this I will lay the next under to hopefully match around 65-70 mins.
1-0 and a red card so that ones out.
