Dan's Tipping Game

[RandoEFC+]

8 hours ago, Dan said:

Your style on this is the most intriguing for me because I think it's along the similar sort of statistical data I try and go for myself. I've gone up the corners route before but it can just be that bit too volatile for me. I'm trying to avoid in-play bets as a general rule but I've had the odd winner on corners on that.

Interested to know how you made that calculation on Girona too.

I have a two-layered algorithm effectively for goals, corners and cards. One layer of counting literally how often a match involving one of the teams has featured over 9 corners or whatever. For team goals, corners, cards, it's a combination of how often Team A gets that many goals, corners and cards and how often Team B concedes them. The second layer is a use of Normal Distribution, which I don't know if you know anything about, but basically measures the average number of goals/corners/cards expected and how consistently it happens to feed back probabilities. I then use an average of the two methods for my final probability and try to stick to betting on stuff that comes out as 70%+ in terms of likelihood.

Combining the two methods basically means that if Team A vs Team B comes out as averaging 9 corners a game but are consistently getting 8, 9, 10 corners and rarely more or less than that, and Team C vs Team D also average 9 corners, but that's because they usually get 12-13 corners and there's just been a couple of games where they've had weird matches with 3 or 4 corners, then Team C vs Team D, for example, will have a higher probability of Over 10.5 Corners than Team A vs Team B.

The calculation for results and margins is a lot more complicated and I've changed it a lot throughout the season. At first I was literally averaging how many Wins Team A had and how many Losses Team B had just like my goals/corners/cards probabilities. It doesn't really make sense though. E.G. If Man City win 80% of the time and Sheffield United lose 70% of the time, that would come out as a 75% chance of a Man City win but the probability is higher than that. Now my spreadsheet reads data and calculates the winning/losing margin for each team in the matches I've included and effectively pitches each recent home team performance against each away team performance. E.G. if a Girona performance saw them win 2-0 and Almeria lost 2-1, these performances pitched against each other would result in Girona winning by three goals (they were good enough to win by two and Almeria were bad enough to lose by one). However, if another match in the dataset saw Girona lose 1-0, this would come out as a draw when pitched against Almeria's 2-1 defeat. I do this for each team's last 10 games and their last 10 home/away games and add a double weighting to their last 5 in each category, so a full dataset would see as much as 30 Girona performances and 30 Almeria performances pitched against each other, so up to 900 combinations. To get an "expected result" of "Girona to win by 2" means when all of those combinations are calculated, there's roughly as many outcomes where they do better than that as there are where they do worse. 

I don't know if that makes any sense, there's a lot of steps to it but my document automates it all. Hard to explain without showing to be honest but those are the main principles behind it. Basically it gives me a record of however many matches I have time to copy data for into my spreadsheet and turns key statistics into a summary of what's statistically likely to happen in each match based on each team's recent performances, without me having to look manually. What it doesn't do is take into account injuries, managerial changes and stuff like that. Luckily, because I'm a Maths nerd, I enjoy the process of tweaking the spreadsheet and having it calculate the probabilities for me, and I enjoy the fact that if I'm bored and there's a football match on that's in my database, I can find a bet to put on for it for fun. I never stake too much. If I ever found a properly good method that seems reliably profitable then maybe I'd up it but that would be a massive bonus.

[RandoEFC+]

Ulm vs Unterhaching

Ulm Match Corners at Home: 10, 8, 11, 12, 11, 10, 12, 11, 12

Unterhaching Match Corners Away: 10, 12, 12, 11, 10, 10, 13, 12, 4

Ulm Corners For at Home: 5, 4, 6, 6, 5, 5, 9, 7, 8

Unterhaching Corners Against Away: 6, 6, 6, 4, 6, 9, 12, 6, 0

I'm surprised to get odds of 1.70 here as there's just one home game for Ulm all season where this bet wouldn't have won, and that one was only one corner away in both cases. Unterhaching's away games are good for this bet statistically as well. The standout exception where there were 4 match corners and they didn't concede any was against Dynamo Dresden who had a man sent off in the first half and were defending a draw (then a surprise lead). Their previous meeting this season had 9 corners although Ulm only had 2 at Unterhaching. My only worry here is that if you include Ulm's away form, there aren't quite as many corners, but Unterhaching still see plenty of corners home or away so I'm happy to take this at 1.70.

 

Bristol City vs Watford

I'm sticking it safe here and just going for Over 8 Match Corners as there wasn't anything super compelling on corners for a specific team.

Bristol City Match Corners at Home: 8, 14, 12, 6, 15, 11, 12, 16, 13, 11

Watford Match Corners: 9, 8, 12, 9, 10, 10, 14, 12, 12, 9, 14, 12

Watford were a bit low on corners in away games earlier this season so I'm reassuring myself here by using their overall form. 11 out of 12 league games they've seen more than 8 corners just falling short once. Hopefully ignoring their away matches with 3-4 corners back in October won't come back to bite me.

[Stan]

Similar to a previous recent bet. Two teams at home who are very strong and high up in the table. 

Peterborough have scored 29 goals in 12 home games. Shrewsbury have conceded 21 in 13 away games. 

Real Madrid very strong at home as expected. 22 goals in 9 home games. They've scored 4 or more goals in 5 of their last 7 games. Almeria the worst away side, conceding 25 in 9 games. They conceded 3, 3 & 5 away at the other current top 4 sides. L

[Dan+]

Arsenal v Crystal Palace - Both teams to score @ 11/8 (2.37) - This is a quite peculiar pick because I think pretty much everyone's instinct for this game would be Arsenal to win to nil. Arsenal's defence bar the last few games has been pretty sturdy while Palace are pretty dour. But to be fair the stats actually back this selection up quite a bit and the price is really generous. Palace have scored in 9 out of 10 away games. Arsenal have scored in 9 out of 10 home games. I know Palace are missing Olise and Ayew, but I still think the odds are simply a bit too big here to ignore them.

Brentford v Nottingham Forest - Brentford to score a penalty @ 4/1 (5.00) - This is a lot more about intuition than any statistical data, but I feel the Toney return is totally set up for him to score, and even more so I can picture him scoring a penalty too. Forest also have a few out and I'd probably back Brentford to win the game anyway.

I need to be going a bit bigger at this point. £10 returns £118.75.

[RandoEFC+]

Just waiting for the Brentford penalty to come in and Palace to not even register a shot on target to ruin the double .

[Stan]

15 minutes ago, RandoEFC said:

Just waiting for the Brentford penalty to come in and Palace to not even register a shot on target to ruin the double .

Or Toney to miss the penalty

 

[Dan+]

Or Palace to win 0-4.

[Dan+]

I'm not gonna claim bad luck on that one to be honest. That was risky and Palace really are just a bit rubbish.

[Lucas]

Forgot to bet yesterday which is annoying as Coventry were 6/5 to win at Wednesday and would have been one of the selections along with probably Portsmouth.

Not much around but i'll go with this.

Bayern and over 3.5 goals 4/6

Juventus to win 4/5

£10 wins £30 dead on. Might even do that on Paddy P.

[Stan]

On 19/01/2024 at 16:32, Stan said:

Real Madrid very strong at home as expected. 22 goals in 9 home games. They've scored 4 or more goals in 5 of their last 7 games. Almeria the worst away side, conceding 25 in 9 games. They conceded 3, 3 & 5 away at the other current top 4 sides.

And Almeria are 2-0 up

 

[Lucas]

And Bayern lost.

At home.

To a team that haven't won away all season.

Who were 27-1 to win today

Fuck...

[RandoEFC+]

On 19/01/2024 at 16:14, RandoEFC said:

Ulm vs Unterhaching

Ulm Match Corners at Home: 10, 8, 11, 12, 11, 10, 12, 11, 12

Unterhaching Match Corners Away: 10, 12, 12, 11, 10, 10, 13, 12, 4

Ulm Corners For at Home: 5, 4, 6, 6, 5, 5, 9, 7, 8

Unterhaching Corners Against Away: 6, 6, 6, 4, 6, 9, 12, 6, 0

I'm surprised to get odds of 1.70 here as there's just one home game for Ulm all season where this bet wouldn't have won, and that one was only one corner away in both cases. Unterhaching's away games are good for this bet statistically as well. The standout exception where there were 4 match corners and they didn't concede any was against Dynamo Dresden who had a man sent off in the first half and were defending a draw (then a surprise lead). Their previous meeting this season had 9 corners although Ulm only had 2 at Unterhaching. My only worry here is that if you include Ulm's away form, there aren't quite as many corners, but Unterhaching still see plenty of corners home or away so I'm happy to take this at 1.70.

 

Bristol City vs Watford

I'm sticking it safe here and just going for Over 8 Match Corners as there wasn't anything super compelling on corners for a specific team.

Bristol City Match Corners at Home: 8, 14, 12, 6, 15, 11, 12, 16, 13, 11

Watford Match Corners: 9, 8, 12, 9, 10, 10, 14, 12, 12, 9, 14, 12

Watford were a bit low on corners in away games earlier this season so I'm reassuring myself here by using their overall form. 11 out of 12 league games they've seen more than 8 corners just falling short once. Hopefully ignoring their away matches with 3-4 corners back in October won't come back to bite me.

A winner at last. Bristol City vs Watford had 15 corners and Ulm just got their 5th corner vs Unterhaching (10th in the match) with 25 minutes left.

+£13.10

[Stan]

1 hour ago, Stan said:

And Almeria are 2-0 up

 

Thank you, Real! 

Another win. 

[Dan+]

I can't complain. I put £20 on Real Madrid to win at half time. You just knew they would.

For a team on so few points Almeria have put in two seriously good efforts against the top two.

[Stan]

17 minutes ago, Dan said:

I can't complain. I put £20 on Real Madrid to win at half time. You just knew they would.

For a team on so few points Almeria have put in two seriously good efforts against the top two.

Sounds like it was controversial in Real's favour but I have no complaints for now

 

[Dan+]

59 minutes ago, Stan said:

Sounds like it was controversial in Real's favour but I have no complaints for now

"You win some, you lose some" doesn't just apply to when you lose as far as I'm concerned

[Pyfish+]

My picks for this week - Cunha to score v Brighton anytime, and Leicester to beat Ipswich. A £75.60 return would be amazing.

[RandoEFC+]

I had a bet on Molenbeek vs Eupen in Belgium for Molenbeek to have under 5.5 corners. They were on 1 corner when the match was abandoned in the 86th minute and they've settled the bet as void. Surely that's BS?

[Dan+]

4 hours ago, RandoEFC said:

I had a bet on Molenbeek vs Eupen in Belgium for Molenbeek to have under 5.5 corners. They were on 1 corner when the match was abandoned in the 86th minute and they've settled the bet as void. Surely that's BS?

I think sadly that is just very unlucky. I came out on the right side of this a few weeks back - I had AZ Alkmaar to beat NEC Nijmegen in a four fold acca, NEC 1-2 up in injury time and Bas Dost collapses, game is postponed, bet is void and the other three teams all won.

[Pyfish+]

20 hours ago, Pyfish said:

My picks for this week - Cunha to score v Brighton anytime, and Leicester to beat Ipswich. A £75.60 return would be amazing.

Sadly not to be  

[Dan+]

On 16/01/2024 at 19:07, Dan said:

WEEK 24 - COULD HAVE GONE BETTER

WEEK 24 STATS
Winning bets: 2/6
Winning selections: 6/11
Weekly forum position: +£7.66 (Ranked 9/24)

---

OVERALL STATS
Winning bets: 42/165
Winning selections: 143/310
Total forum position: -£303.09

 

And we're back again. Not loads of change in the table and one of the most balanced weeks we've had in terms of overall performance. I'll work chronologically on what actually happened over the weekend rather than up and down the table as it's about the only way I'll not come to myself last and there's only so many ways I can address that at this point.

@CaaC (John) was the first player bust this week as he went for the Friday night clash between Burnley and Luton. Luton grabbed a controversial draw but it wasn't enough, but luckily Man Utd didn't win either so he wasn't cost, and is the only player to lose both picks this weekend.

@Stan picks up consecutive wins picking Adam Armstrong to net in Southampton's latest 4-0 win over a bottom half team, which I feel like is almost a fortnightly occasion at this point. This did come however after his other selection in Bolton v Cheltenham was void due to a spectator illness, of which said Bolton fan did actually end up dying from, which is a shame to hear. He was given the option of replacing it with a second pick but cashed in for an £11 profit instead.

@Pyfish gets back to winning ways. The Cole Palmer price to score at anytime was fantastic I have to see. He's the one plus in what's been another dreadful season for Chelsea and he scored another penalty to give them a 1-0 victory. Newcastle v Man City both teams to score, pays out after 35 minutes. Lovely. It's the biggest win of the week and largely responsible for the small profit we find ourselves in.

Sunday however was much less productive. I happened to notice that the Barnsley - Bristol Rovers selection had come in after Barnsley won 2-1, so I just simply needed Girona, almost flawless Girona, to beat a barely believably bad Almeria side and for over a goal to land in the game. We've done it to death in here but you all know what happened. The price on Girona to win actually got even bigger prematch which is always a worrying sign. I watched the game play out and honestly, just how did they know? Almeria were the better team. It's baffling stuff to me. The sort that makes you wonder how people could ever make consistent money off doing this. That's 24 bets and just 1 win. That's truly awful, almost impressive to pull off. If you had to pick 24 bets to lose I think you would do very well to get 23.

@RandoEFC moves to corners and suffers similar pain. Charlton v Peterborough defies the recent form and just 7 corners occur in the game. My gripe with betting on corners is it's just a bit too reliant on game-state. I'll give you an example here in that I had something on Derby v Burton last night via Bet365. For those unaware, Derby went 2-0 up so the Derby to win leg pays out, but I need Derby to get 4 corners as well. They had just the 1 corner as they went 2-0 up and this is a problem because there's less desire to attack. I got very lucky though, Burton came back to make it 2-2 and all of a sudden Derby had to attack them, and the corners flew in so I won. What's unlucky about this Charlton Peterborough pick for me is there was nothing in the game state that meant that should've lost, so it's just really unlucky. To rub salt into the wound, there were a staggering 21 corners in the Manchester United - Spurs game.

There was room for one last bit of bad luck though as recent league leader @Lucas came awfully close to landing another massive win. AC Milan v Roma was a 3-1 home win, and a game that has ultimately ended up finishing Jose Mourinho's tenure there. Having watched their last three games, I did think it was coming soon, they look really poor. The unlucky part though is that PSG beat Lens 0-2, so he was just the goal off and Lens did have a man sent off in the first half, so I think it's quite plausible that they would've scored without it. This would've left him nearly £200 clear of 2nd and quite honestly, the comfortable favourite to win the league.

We move on.

WEEK 25 - ESSENTIALLY BREAKEVEN

WEEK 25 STATS
Winning bets: 2/6
Winning selections: 6/12
Weekly forum position: -£1.82 (Ranked 10/25)

---

OVERALL STATS
Winning bets: 44/171
Winning selections: 149/322
Total forum position: -£304.91

 

The latest instalment arrives. A fair bit of change in the league table and a pretty ordinary week on the whole.

The week produced two winners. The first is @Stan who moves back up into 4th courtesy of two second half turnaround victories. Always what I find to be the perk of backing a team to win and over X goals is you're less pissed off with the inevitable notification to find they've gone 1-0 down. Shrewsbury took a shock lead at Peterborough before half time, but there was no surprise in the turnaround and it ended 2-1. Almeria went one better, taking a shock 0-2 lead at the Bernabeu before the inevitable crumbling. It would've been quite something to see Almeria stunt another of the La Liga runaway teams but it appears that's only reserved for my picks. That's Stan's third straight win, and winning bang on £25 means he's essentially won back his bonus stake.

The other winner was @RandoEFC who finally pays off with the corners. This absolutely sailed in to be honest, with Bristol City v Watford producing 15 between them, miles over the 9 required. The Ulm v Unterhaching game went similarly with the bet all but in by half time, as there were 14 corners in total, of which Ulm took 8 of them.

I went ambitious this week and didn't really get anywhere near a winner. It was always risky picking a Palace side in pretty dire straits, missing Olise to go to the side with the least number of chances created against them and score, but the odds were backable for me given the data. Toney had his moment, it just happened to be a free-kick instead of a penalty this time. I can't really whinge this time.

Who probably can whinge however is @Lucas. It amazes me some of the odds you get on Juventus wins. Always worth a go. But for Bayern to turn that result in at home to a team with no away wins all season is remarkably unlucky timing. He does however retain his place at the top of the table as @Pyfish double blanks in a late pick. No Cunha goal, no Leicester win. Not his week.

@CaaC (John) was the other predictor. His went according to the odds. The comfortable favourites of Arsenal winning heavily. The outsiders of Nottingham Forest losing, dropping him to a season low position of 8th following five straight losses.

On we go.

[CaaC (John)+]

 

[Dan+]

3 hours ago, CaaC (John) said:

 

I knew I forgot to include something in the round up and it was a reminder that you can either stick with these or go for something new, luckily you remembered this time

[Dan+]

On 18/01/2024 at 08:54, Stan said:

I've recently found that in-play bets work better for me when watching a game. Especially cricket. But recently for football you can get a good inkling on what's going to happen, especially in the interval bets that SkyBet do.

Soucek was 4/1 to have 1 shot on target the other day against Bristol City. Came in within a few minutes.

Bruno Fernandes to have 2 shots on target, or 1 or 2 shots on target from outside the box is usually a good return. Or similar players - works well with the likes of Ward-Prowse, Maddison (when fit), Bowen, Foden, De Bruyne etc. 

Cards can be a bit trickier, but corners is also a good bet in-play. Yesterday's Blackpool/Forest had about 17 (?) by the end of 90 mins. Already 7 by half-time and it was good odds for anything about 13/14 if I remember correctly. 

I've been trying to avoid having in-plays because I found myself quite often doing things like backing a team to have another corner with 80 on the clock when attacking and they inevitably scored and stopped putting the pressure on but I do think every now and then you get some good opportunities for it like that too. I don't tend to use Sky as much but they've always had the best variety of bet builders and long shots etc... tend to find for match odds they're pretty bad though. Have you tried this acca freeze? I keep meaning to do it, I keep seeing people freezing a team at stupid prices and boosting an acca to a ridiculous price. Only a matter of time before they tone that down I think.

 

On 18/01/2024 at 11:06, RandoEFC said:

I have a two-layered algorithm effectively for goals, corners and cards. One layer of counting literally how often a match involving one of the teams has featured over 9 corners or whatever. For team goals, corners, cards, it's a combination of how often Team A gets that many goals, corners and cards and how often Team B concedes them. The second layer is a use of Normal Distribution, which I don't know if you know anything about, but basically measures the average number of goals/corners/cards expected and how consistently it happens to feed back probabilities. I then use an average of the two methods for my final probability and try to stick to betting on stuff that comes out as 70%+ in terms of likelihood.

Combining the two methods basically means that if Team A vs Team B comes out as averaging 9 corners a game but are consistently getting 8, 9, 10 corners and rarely more or less than that, and Team C vs Team D also average 9 corners, but that's because they usually get 12-13 corners and there's just been a couple of games where they've had weird matches with 3 or 4 corners, then Team C vs Team D, for example, will have a higher probability of Over 10.5 Corners than Team A vs Team B.

The calculation for results and margins is a lot more complicated and I've changed it a lot throughout the season. At first I was literally averaging how many Wins Team A had and how many Losses Team B had just like my goals/corners/cards probabilities. It doesn't really make sense though. E.G. If Man City win 80% of the time and Sheffield United lose 70% of the time, that would come out as a 75% chance of a Man City win but the probability is higher than that. Now my spreadsheet reads data and calculates the winning/losing margin for each team in the matches I've included and effectively pitches each recent home team performance against each away team performance. E.G. if a Girona performance saw them win 2-0 and Almeria lost 2-1, these performances pitched against each other would result in Girona winning by three goals (they were good enough to win by two and Almeria were bad enough to lose by one). However, if another match in the dataset saw Girona lose 1-0, this would come out as a draw when pitched against Almeria's 2-1 defeat. I do this for each team's last 10 games and their last 10 home/away games and add a double weighting to their last 5 in each category, so a full dataset would see as much as 30 Girona performances and 30 Almeria performances pitched against each other, so up to 900 combinations. To get an "expected result" of "Girona to win by 2" means when all of those combinations are calculated, there's roughly as many outcomes where they do better than that as there are where they do worse. 

I don't know if that makes any sense, there's a lot of steps to it but my document automates it all. Hard to explain without showing to be honest but those are the main principles behind it. Basically it gives me a record of however many matches I have time to copy data for into my spreadsheet and turns key statistics into a summary of what's statistically likely to happen in each match based on each team's recent performances, without me having to look manually. What it doesn't do is take into account injuries, managerial changes and stuff like that. Luckily, because I'm a Maths nerd, I enjoy the process of tweaking the spreadsheet and having it calculate the probabilities for me, and I enjoy the fact that if I'm bored and there's a football match on that's in my database, I can find a bet to put on for it for fun. I never stake too much. If I ever found a properly good method that seems reliably profitable then maybe I'd up it but that would be a massive bonus.

That sounds like a pretty good system to be honest and I'd imagine over the long run that is likely to pay out, which once again brings me back to my point that this thread is probably a quite bad reflection of the bigger picture because you're only doing two selections and if you do two they're accumulated anyway, which I'm finding is a bit of a problem (my overall profit this season is lower than if I'd just done £10 on a single selection of everything I'd had). I mean it's a bit of fun ultimately but I remember years ago when I ran it I was almost hoping for good methods etc... to come out of it.

I think what you point out at the end is half the issue as well. If you start consistently beating or outthinking them they will just ban you. I know a couple of methods on how you can pretty much guarantee a win over time but I've not put it into enough practice. Basically finding odds above what an exchange (real odds) will offer, although quite how you consistently find those without software I'm not really sure.

My sisters ex boyfriend told me he used to have some bit of software that would not only find the bets like this but actually place it for him as well  

[Stan]

50 minutes ago, Dan said:

Have you tried this acca freeze? I keep meaning to do it, I keep seeing people freezing a team at stupid prices and boosting an acca to a ridiculous price. Only a matter of time before they tone that down I think.

On the odd occasion. Kind of indifferent to it, haven't used it enough to profit majorly from it i.e picking the right times to freeze and I guess it's dependent on which team it's freezing on. Like you mention it could be a 100/1 shot you freeze it on.

I saw someone on Twitter get banned from using SkyBet because he used 3 freezes on similar bets, but T&Cs state you can only use it once per week. The guy had won thousands off the 3 bets he used it on because it included the same game in each bet. I don't even think they paid out the full amount he should have got which is snidey as fuck, but when it comes to it a) it was their glitch that allowed him to put it on more than one bet and b) to ban him from the site is a bit shit.

I've also noticed a way to get extra free bets. Sometimes Skybet will offer £5 free build-a-bet offers. Can claim that fine, but if you log out the offer is there as a £1 build-a-bet offer, so you can click on that, log in and then claim that. So you get two separate ones to use any time you like before expiry.