Women’s Euro 2022

[harry_rag]

Expecting an Austria win in the first game, perhaps 3-0 or 3-1 but be delighted to see N Ireland upset the apple cart and the odds. I think Billa will probably score but so do the bookies. 10/11 was tempting me despite a reluctance to back at odds on in that market, the trim to 5/6 is enough to stay my hand. Her goal minutes are 42 to buy which is more than double the price of the next player in the betting. Dunst was on my radar but, at 23/10, I'll give it a miss.

One to watch, hopefully enjoy, and try and get a handle on how I see the England/NI final game going.

[Torque]

12 minutes ago, harry_rag said:

Expecting an Austria win in the first game, perhaps 3-0 or 3-1 but be delighted to see N Ireland upset the apple cart and the odds. I think Billa will probably score but so do the bookies. 10/11 was tempting me despite a reluctance to back at odds on in that market, the trim to 5/6 is enough to stay my hand. Her goal minutes are 42 to buy which is more than double the price of the next player in the betting. Dunst was on my radar but, at 23/10, I'll give it a miss.

One to watch, hopefully enjoy, and try and get a handle on how I see the England/NI final game going.

I'm backing two boosts, again at 365 which are based on Billa scoring. Fingers crossed.

10pts Austria to win, BTTS and Billa 2+ Shots on Target @ 6.00

10pts Austria to win HT-FT, BTTS and Billa to Score @ 7.50

[harry_rag]

4 hours ago, Labrador said:

I've gone with over 1.5 Northern Ireland cards at 11/8 with Bet 365. 

Bought NI bookings at 9 on 55 minutes. Can't see them negotiating the rest of the game without drawing at least 1 card.

[harry_rag]

I'm surprised quite how big favourites England are for the game against Norway, even allowing for home advantage. I think this could be a bit of a test for England though one they will ultimately pass. A one goal victory would be my guess at the outcome. At the prices (around 2.3) I think Norway +1 AH isn't the worst bet. 7.6 for them to win the game might look big in hindsight. I'll stick with the Riem-inspired bookings bets though, pending any player bets once the line ups are announced.

[Torque]

41 minutes ago, harry_rag said:

I'm surprised quite how big favourites England are for the game against Norway, even allowing for home advantage. I think this could be a bit of a test for England though one they will ultimately pass. A one goal victory would be my guess at the outcome. At the prices (around 2.3) I think Norway +1 AH isn't the worst bet. 7.6 for them to win the game might look big in hindsight. I'll stick with the Riem-inspired bookings bets though, pending any player bets once the line ups are announced.

After washing out with my last two bets, I'll just take one in this match. As you say, it should be competitive.

10pts England to win, BTTS and Ellen White 2+ Shots on Target @ 7.50 365

[Labrador]

53 minutes ago, harry_rag said:

Bought NI bookings at 9 on 55 minutes. Can't see them negotiating the rest of the game without drawing at least 1 card.

Only saw the last 10 minutes or so when it was obvious the Venezuelan ref was not remotely interested in troubling to reach into her pocket despite a couple of agricultural hacks from the N Irish team. Total opposite of Swiss referee Esther Straubli who showed yellow and a red for most innocuous stuff. Definitely a market to play for interest with small stakes.

[harry_rag]

To be fair I only saw one hack that had me thinking I might get the card to spare my blushes and losses and even that, on seeing the replay, didn't seem that bad. Hopefully Riem isn't in a benign mood! Agree about Straubli, had already flagged her. Hopefully her next game is one with a bit of "juice" in it.

[harry_rag]

10 minutes ago, Torque said:

After washing out with my last two bets, I'll just take one in this match. As you say, it should be competitive.

10pts England to win, BTTS and Ellen White 2+ Shots on Target @ 7.50 365

I'm going to say I don't think that's a great price simply because the multiple odds for the 2 components at the best prices I've seen is over 10/1. England win and btts is 21/10 and White 2+ SoT is as big as 5/2 (and as short as evens). I find myself thinking the 5/2 for the 2 SoT might be decent but it's a market I'm inclined to stay away from.

Sorry that Billa didn't have such a great game though I must admit I wasn't keen on the need for NI to score.

[Torque]

Actually I will try another one. The odds are shorter, but given this could be a close match I actually think it's riskier. However, I like the chances of Hemp and Mead hitting the target and after that I'm just left with England to be winning at the half plus at the end. I'll take the chance,

10pts England HT-FT, Lauren Hemp 1+ Shots on Target and Beth Mead 1+ Shots on Target @ 5.00 365

[Torque]

5 minutes ago, harry_rag said:

I'm going to say I don't think that's a great price simply because the multiple odds for the 2 components at the best prices I've seen is over 10/1. England win and btts is 21/10 and White 2+ SoT is as big as 5/2 (and as short as evens). I find myself thinking the 5/2 for the 2 SoT might be decent but it's a market I'm inclined to stay away from.

Sorry that Billa didn't have such a great game though I must admit I wasn't keen on the need for NI to score.

Thanks for the analysis Harry. If it's 10 to 1 at another book then I've obviously missed a trick.

[harry_rag]

12 minutes ago, Torque said:

Thanks for the analysis Harry. If it's 10 to 1 at another book then I've obviously missed a trick.

I'm not saying you could get the combo at better odds, just comparing the multiplied odds of the individual components (which is what I used to do when considering these bets). 

Similarly with your second bet, with the individual components available at 2.35, 1.8 and 1.67 I'd have wanted better than 7 if I was going to consider it. They're not really related contingencies so (for me at least) if the odds you're offered are less than the best odds multiplied out then I'd tend to leave it alone.

I know you don't need me to pose as your metaphorical grandmother but I'd suggest it's extremely hard to beat these prop bets and I'd make sure I specifically recorded these type of bets so I could track their performance.

Bottom line, the vast majority of them will be at shorter than true odds. The only reason you would get value is if the bookies make a rick so big that the edge isn't eroded by the margin they apply or if they do it deliberately as a headline grabbing boost.

[harry_rag]

47 minutes ago, Torque said:

Actually I will try another one. The odds are shorter, but given this could be a close match I actually think it's riskier. However, I like the chances of Hemp and Mead hitting the target and after that I'm just left with England to be winning at the half plus at the end. I'll take the chance,

10pts England HT-FT, Lauren Hemp 1+ Shots on Target and Beth Mead 1+ Shots on Target @ 5.00 365

Off to a decent start. Time yet but my "close contest" take is looking a bit askew so far!

[Torque]

All fair points Harry. I think I'll bow out of the props while I'm ahead 

[Torque]

If Norway somehow score, my two bets could be 'Exhibit A' and 'Exhibit B' for examples of bad bets that win - mathematically, at least.

Edited July 11, 2022 by Torque

[harry_rag]

3 minutes ago, Torque said:

All fair points Harry. I think I'll bow out of the props while I'm ahead 

I don’t know mate, on today’s evidence I’d consider what I say and do the exact opposite! 

[GhostLetter]

Torque - looking at the Wimbledon Singles Finals that took place over the weekend, and now today’s Euro 2022 matches, you seem to be a bit obsessed with these prop. bet trebles at the moment. I’d totally agree with harry_rag and say that I can’t see any value in them whatsoever. You seem knowledgeable when it comes to sports and betting, but everything about these bets screams that they’re aimed at mug punters. I don’t understand why you’d go anywhere near them. They seem very random and over-complicated. I think you should look for simpler bets, rather than bets on three different things to all happen, in the same match. Feel free to prove me wrong in the long run, but none of your ten prop. bet trebles I have seen over the past few days have struck me as value.

Having said that, harry_rag, I think your calculations in simply combining the individual odds weren’t correct on this occasion. There were definitely related contingencies involved, which would therefore reduce the odds. If a certain player/certain players have a shot/shots on target, then obviously their team are more likely to score and also more likely to lead at half time or win the match. Some of the bets there are perfect examples of related contingencies. So, the odds aren’t necessarily much different to the individual odds - rather that the bets are a bit too complex and unlikely to come in. 

[harry_rag]

@GhostLetter Ooh I love a maths puzzle, been thinking about this too long given the late hour! I was going to say that team to win and player to score was a better example of related contingency but I’ve ended up thinking the following is true:

The true odds for team A to win and their striker to score are simply the true odds of  the two outcomes multiplied together, they are not related in a way that makes them any less than that. So if England are 2/1 to win and White 2/1 to score the odds for both to happen are 8/1.

I’ll recheck my maths after I’ve had my weetabix as that goes against my long held assumptions but I think it’s true!

[GhostLetter]

7 hours ago, harry_rag said:

@GhostLetter Ooh I love a maths puzzle, been thinking about this too long given the late hour! I was going to say that team to win and player to score was a better example of related contingency but I’ve ended up thinking the following is true:

The true odds for team A to win and their striker to score are simply the true odds of  the two outcomes multiplied together, they are not related in a way that makes them any less than that. So if England are 2/1 to win and White 2/1 to score the odds for both to happen are 8/1.

I’ll recheck my maths after I’ve had my weetabix as that goes against my long held assumptions but I think it’s true!

From what I have seen, you are the most knowledgeable gambler on this site. I am therefore very surprised at what I am reading from you here!

England to win and White to score are obviously very much related! If White scores, England are more likely to win. If England win, it is more likely that White has scored. One outcome happening vastly increases the chances of the other also happening. Therefore, if you combined them both - at the 2/1 odds for each that you used as an example - you certainly wouldn’t get 8/1. As I said previously, that is a perfect example of a related contingency. 

You can check/confirm that on most bookies’ websites. Find an anytime/first goal scorer bet. Find the odds of their team to win. Then find the bet that combines them - is it called ‘wincast’ or something like that? - and the odds will be much less than simply multiplying the separate bets together. That’s not down to the bookies knocking a sneaky bit of value off - that’s down to it being a related contingency.

In statistical terms, these are dependent events, rather than independent events, so the calculation is not as simple as multiplying together the chances of each separate event. That is the calculation for independent events.

Edited July 12, 2022 by GhostLetter

[harry_rag]

7 hours ago, GhostLetter said:

From what I have seen, you are the most knowledgeable gambler on this site. I am therefore very surprised at what I am reading from you here!

You flatter me, I just think I look more at the logic behind the bets than most but there are those with a far deeper understanding of statistical probability etc. To be fair I’ve surprised myself here in terms of changing longheld preconceptions but, having slept on it, I still think I’m right.

I’ll be back later to try and convince you (or eat humble pie).

 

 

[GhostLetter]

23 minutes ago, harry_rag said:

You flatter me, I just think I look more at the logic behind the bets than most but there are those with a far deeper understanding of statistical probability etc. To be fair I’ve surprised myself here in terms of changing longheld preconceptions but, having slept on it, I still think I’m right.

I’ll be back later to try and convince you (or eat humble pie).

 

 

I can’t wait to hear this.....

I’m not sure where your logic has gone wrong, but I’m interested to see!

I studied statistics at university, so maybe I would fall into that category you mentioned, of having a far deeper understanding of statistical probability?! I don’t claim to be an expert, and much of statistics is beyond me, but I think what we are debating/discussing here are fairly basic statistical concepts.

[harry_rag]

18 minutes ago, GhostLetter said:

I’m not sure where your logic has gone wrong, but I’m interested to see!

 I’d propose a wager but there’s too much risk that we’ll fail to reach agreement!

All I’ll say is that we’ve probably both fallen into the same trap of assuming dependent events are also “related contingencies” (where the true odds of both things happening are not the product of the individual odds) when it’s possible for them to be the former but not the latter.

[GhostLetter]

11 minutes ago, harry_rag said:

 I’d propose a wager but there’s too much risk that we’ll fail to reach agreement!

All I’ll say is that we’ve probably both fallen into the same trap of assuming dependent events are also “related contingencies” (where the true odds of both things happening are not the product of the individual odds) when it’s possible for them to be the former but not the latter.

I’m more than happy to have a wager. We only have to check the odds that bookies offer, to see that I’m correct!

 

I think you don’t understand - or have temporarily got muddled - as to what a related contingency is.

 

I’ll use Harry Kane playing for England as an example. Harry Kane has scored in 47% of his appearances for England. From your argument, this would mean we would also expect him to score in 47% of England wins, 47% of England draws and 47% of England defeats, across the matches he has appeared in.

OBVIOUSLY, that is ludicrous. He is far more likely to have scored in matches England have won, and vice versa, England are far more likely to have won in matches he has scored in. And so the stats show: In England wins, he has scored 62% of the time. In draws, he has only scored 24% of the time, and in defeats that figure drops to 18%.

The chances of ‘Harry Kane scoring’ and ‘England winning’ are clearly linked to each other and therefore group together more than if they were independent events. That’s why the bookies don’t give you the odds of multiplying the two separate bets together.

If your ‘Harry Kane to score’ bet is a winner, you now effectively have a one-goal head start on your other bet, ‘England to win.’ Therefore, presumably the actual combined odds for ‘Harry Kane to score’ and ‘England to win’ would be the odds of Harry Kane to score, multiplied by the odds of England (+1) on the handicap betting.

I’ve just checked this now, for the Germany vs. Spain match this evening. Bet365 have Germany to win at 15/8 and Lea Schuller to score first at 11/2. If you combine those odds (as if they were independent events/unrelated bets), by multiplying the odds, you would get close to 18/1. However, the odds offered for the ‘Wincast’, for Schuller to score first and Germany to win, are 9/1 - that is the same odds as if you combine Schuller to score first at 11/2, and Germany (+1) at 8/15. (Schuller scoring first gives your other bet a one goal head start.)

 

I hope that makes sense? Out of all the people on the forum, you would have been the last person I expected to be explaining odds and statistics to. I worry someone has hacked your account - or your brain!

[harry_rag]

Ok, let's try this although I feel it's as likely to flounder on my ability to articulate what I mean as for what I mean to be wrong!

My hypothesis is that the true odds for a team to win and one of their named players to score is the true odds for each event multiplied together, as opposed to it being a genuine "related contingency" where the true odds are different to that calculation.

Let's take the bet "England to win and Smith to score" where both components have true odds of 2/1. I'm not disputing the extent to which the events are related during and after the game, just how you calculate the probability before the event. Game kicks off and Smith scores, one leg is in and the odds for the remaining leg shorten. England score and the odds for that leg shorten though the odds for Smith to score will be largely unaffected (and drift with time). Opponents score and the odds for the win drift, pushing out the odds for our bet. 

If you ignore (for now, bear with me) the fact that we are specifying the result and a named scorer our bet boils down to the proposition that England will win and score at least one goal. The odds for that outcome are obviously not the product of England to win and England to score a goal, it's just the odds for England to win as they can't do so without scoring a goal. Someone has to score for the bet to land and each players' odds are simply whatever they are, 2/1 in the case of Smith. Smith to score first and England to win is different as the proposition of England to win and to score first will have different odds to simply England to win.

I contend that the specific maths are simply as follows. It doesn't matter who the named scorer is or what their price is, the principle is the same.

England to win and Smith to score = 3 x 3 = 9

England to win and Smith not to score = 3 x 1.5 = 4.5

If you dutch those 2 outcomes you get the win odds of 3. The only way the first can be shorter than 9 is if you think the second should be longer than 4.5 (for every named player). That England will score is factored as a given into the win price. Trying to name a scorer as well inflates the odds by whatever the probability for that scorer is, whether it's the 1.5 super striker, the 4.0 midfielder or the 21 defender. If England score their odds will contract to the same extent regardless of who has scored the goal. 

I've got a feeling that this will leave you unconvinced in which case I can only ask what would be your take on where my calculation is wrong?

On the subject of the bookies offering less than the multiplied out odds, these markets are catnip for mug punters; I contend that it's perfectly feasible that the bookies would tack a bit more margin on to maximise profit and protect themselves from shrewder punters who picked them off where they'd got the underlying prices wrong.

[Torque]

12 hours ago, GhostLetter said:

Torque - looking at the Wimbledon Singles Finals that took place over the weekend, and now today’s Euro 2022 matches, you seem to be a bit obsessed with these prop. bet trebles at the moment. I’d totally agree with harry_rag and say that I can’t see any value in them whatsoever. You seem knowledgeable when it comes to sports and betting, but everything about these bets screams that they’re aimed at mug punters. I don’t understand why you’d go anywhere near them. They seem very random and over-complicated. I think you should look for simpler bets, rather than bets on three different things to all happen, in the same match. Feel free to prove me wrong in the long run, but none of your ten prop. bet trebles I have seen over the past few days have struck me as value.

Having said that, harry_rag, I think your calculations in simply combining the individual odds weren’t correct on this occasion. There were definitely related contingencies involved, which would therefore reduce the odds. If a certain player/certain players have a shot/shots on target, then obviously their team are more likely to score and also more likely to lead at half time or win the match. Some of the bets there are perfect examples of related contingencies. So, the odds aren’t necessarily much different to the individual odds - rather that the bets are a bit too complex and unlikely to come in. 

I'm glad I seem knowledgeable - that's a relief. I'm not a big fan of prop bets and very rarely place them. Mathematically I'm sure they don't offer value - if they did it's unlikely they'd be promoted by the books - but betting isn't entirely about Maths. It's imprecise - mistakes get made. Prices are usually based on thousands of previous iterations of an event or a similar event, but none of those events can definitively approximate the event to come.

[harry_rag]

13 hours ago, GhostLetter said:

Having said that, harry_rag, I think your calculations in simply combining the individual odds weren’t correct on this occasion. There were definitely related contingencies involved, which would therefore reduce the odds. If a certain player/certain players have a shot/shots on target, then obviously their team are more likely to score and also more likely to lead at half time or win the match. Some of the bets there are perfect examples of related contingencies. So, the odds aren’t necessarily much different to the individual odds - rather that the bets are a bit too complex and unlikely to come in. 

I'd stand by my view on the specific bets I commented on. Where the premise can be boiled down to England will win and will have at least 2 shots on target (ignoring naming specific players) then it's not a great example of a related contingency. You won't get much bigger odds for England to win and have as few as 2+ SoT than the straight win. If England have a shot on target (as opposed to score) it won't particularly shift their win price.

I wasn't asserting that this was how to calculate the true odds so much as this; if you're getting significantly less for a prop bet than the best individual odds multiplied out (when those odds should already be worse than the true odds anyway) then it's probably not a great prop. It's a metric to consider IF you're going to make these bets.

But the whole disagreement about what is and isn't a related contingency and how much it impacts the true price obviously illustrates how hard it would be to beat the market on these bets anyway! 

[GhostLetter]

1 hour ago, harry_rag said:

Ok, let's try this although I feel it's as likely to flounder on my ability to articulate what I mean as for what I mean to be wrong!

My hypothesis is that the true odds for a team to win and one of their named players to score is the true odds for each event multiplied together, as opposed to it being a genuine "related contingency" where the true odds are different to that calculation.

Let's take the bet "England to win and Smith to score" where both components have true odds of 2/1. I'm not disputing the extent to which the events are related during and after the game, just how you calculate the probability before the event. Game kicks off and Smith scores, one leg is in and the odds for the remaining leg shorten. England score and the odds for that leg shorten though the odds for Smith to score will be largely unaffected (and drift with time). Opponents score and the odds for the win drift, pushing out the odds for our bet. 

If you ignore (for now, bear with me) the fact that we are specifying the result and a named scorer our bet boils down to the proposition that England will win and score at least one goal. The odds for that outcome are obviously not the product of England to win and England to score a goal, it's just the odds for England to win as they can't do so without scoring a goal. Someone has to score for the bet to land and each players' odds are simply whatever they are, 2/1 in the case of Smith. Smith to score first and England to win is different as the proposition of England to win and to score first will have different odds to simply England to win.

I contend that the specific maths are simply as follows. It doesn't matter who the named scorer is or what their price is, the principle is the same.

England to win and Smith to score = 3 x 3 = 9

England to win and Smith not to score = 3 x 1.5 = 4.5

If you dutch those 2 outcomes you get the win odds of 3. The only way the first can be shorter than 9 is if you think the second should be longer than 4.5 (for every named player). That England will score is factored as a given into the win price. Trying to name a scorer as well inflates the odds by whatever the probability for that scorer is, whether it's the 1.5 super striker, the 4.0 midfielder or the 21 defender. If England score their odds will contract to the same extent regardless of who has scored the goal. 

I've got a feeling that this will leave you unconvinced in which case I can only ask what would be your take on where my calculation is wrong?

On the subject of the bookies offering less than the multiplied out odds, these markets are catnip for mug punters; I contend that it's perfectly feasible that the bookies would tack a bit more margin on to maximise profit and protect themselves from shrewder punters who picked them off where they'd got the underlying prices wrong.

Where you are wrong is your lack of understanding of the difference between independent and dependent events, in terms of probability. Or, your lack of understanding of what counts as a related contingency, in terms of gambling.

 

First, in terms of probabilities - if you wanted to work out the combined probability of there being rainfall at the moment when you step outside in the morning AND of you tossing a single, fair coin, and it landing on heads, before you leave, then you’d simply multiply the probabilities of those two INDEPENDENT events together. On dry days, you would expect to land 50% heads and 50% tails and on wet days you would also expect 50% heads and 50% tails. That is how independent events work. One event is not changing the distribution of the other. There is no grouping or lack of grouping between two outcomes - such as wet and heads, or dry and tails.

However, if you wanted to calculate the probability of there being rainfall the moment you step outside AND of you standing in a puddle on the way to work, then those events are clearly related, so we are looking at DEPENDENT probabilities. If you stand in a puddle on 5% of days, and it is raining on 20% of days, you don’t multiply these dependent events together and say, the chances of both happening on the same day are 1%. Hopefully you can see that doesn’t make sense? That would mean that the other 4 out of 5 times you stand in a puddle would be on dry days.

So, you would be 4 times more likely to stand in a puddle on a dry day, than a wet day! Again, I hope you can see the logical fallacy there? In this case, there will clearly be a grouping between wet days and the days you stand in a puddle. And also the opposite grouping - of dry days and the days when you don’t stand in a puddle.

So, that means we need a different formula, for calculating the combined probabilities of dependent events. The calculation for finding the chances of two dependent events both happening is P(A) x P(B|A). That is the probability of event A happening, multiplied by the probability of event B happening GIVEN THAT WE KNOW A HAS ALREADY HAPPENED.

 

So, that’s what I was saying earlier, in my Germany and Spain example. If you’re betting on a player to score, when you then  look to combine that bet, with another outcome in the same match, you don’t multiply by that second event’s original probability, you multiply by its probability given that the first thing has occurred. Which, if you were betting for a team to win the match, as in the example we used, would mean you are multiplying by the odds for that team to win, with a one-goal head start, (+1 on the handicap betting).

If you don’t understand the explanation in terms of probabilities, I thought the Harry Kane example and statistics I gave earlier, showed it quite nicely in terms of gambling and related contingencies. Those weren’t numbers I made up, those were the actual stats. That wasn’t a biased pick by me, that was the first player that came to mind. We could use Lukaku or Lewandowski and look at their international goals, or look at any player’s club goals - it would take much longer to gather all the statistics - and we’d always see the same pattern. Whatever percentage of matches a player scores in overall, over a reasonable enough sample size, that will ALWAYS be skewed towards a higher percentage of matches scored in when the team wins, to a lower percentage when the team draws, to an even lower percentage when the team loses. Those events are clearly related. How could they not be? It is ludicrous to say that if a player scores, their team’s chances of winning/drawing/losing the match are still the same as in matches when that player doesn’t score. That’s what you are claiming here, with your lack of understanding of independent events and dependent events.

 

I genuinely am baffled that we are wasting our time having this debate. If you still think the bookies have their odds so wrong on this, then you have just unlocked a huge gold mine, and are about to become very rich. And probably also very famous, as you have just changed some of the most basic theories of statistics!

We can drill down into this further if we need to, but I’m hoping at some point soon the penny will drop and you will see where you have been going wrong with this?!

[GhostLetter]

1 hour ago, Torque said:

I'm glad I seem knowledgeable - that's a relief. I'm not a big fan of prop bets and very rarely place them. Mathematically I'm sure they don't offer value - if they did it's unlikely they'd be promoted by the books - but betting isn't entirely about Maths. It's imprecise - mistakes get made. Prices are usually based on thousands of previous iterations of an event or a similar event, but none of those events can definitively approximate the event to come.

The chances of the bookies having got all the parts of the prop bet wrong are surely tiny though? You’d be better isolating which specific part of the prop bet you think is value and focusing on that.

If anything, due to the profit margins that bookies build into their odds, the more parts you add to a prop bet, the more you are allowing the bookies to offset/cancel out any part of the prop bet they may have got wrong.

I think it’s a lazy market, aimed at part-time gamblers. And from what I’ve seen here, you’re far better than that!

[harry_rag]

1 hour ago, harry_rag said:

I contend that the specific maths are simply as follows. It doesn't matter who the named scorer is or what their price is, the principle is the same.

England to win and Smith to score = 3 x 3 = 9

England to win and Smith not to score = 3 x 1.5 = 4.5

If you dutch those 2 outcomes you get the win odds of 3. The only way the first can be shorter than 9 is if you think the second should be longer than 4.5

What would be your best take on where these specific maths are wrong then? If England win there are only 2 possible outcomes relating to Smith scoring or not and their true odds have to add up to the same probability as the England win. I've provided my specific numbers and why I think they're right. Show your sums for this specific scenario in the hope it may finally get it through to my thick head! 

3 hours ago, GhostLetter said:

If your ‘Harry Kane to score’ bet is a winner, you now effectively have a one-goal head start on your other bet, ‘England to win.’ Therefore, presumably the actual combined odds for ‘Harry Kane to score’ and ‘England to win’ would be the odds of Harry Kane to score, multiplied by the odds of England (+1) on the handicap betting.

It's just struck me that this scenario seems to support my contention more than yours. Let's imagine our England win = 3 and Kane = 3 scenario and throw in England +1 is 1.5 (fair based on Germany's odds tonight v Spain). My contention is that the true odds for England win/Kane goal is 9 whereas yours is that it would be more like 4.5. Kane scores in the first minute and England's win odds are cut to roughly 1.5 (i.e. they've got a real +1 now). I'm free to stick my Kane winnings on an England win at 1.5 which would pay me as a 4.5 double. You're suggesting the fair odds pre match (when we don't know whether Kane scores or not) are the same as they would be after Kane has actually scored. Obviously Kane scoring increases the chance of an England win but only if he does. I think you're putting too much emphasis on the 2/1 chance of him scoring in this example rather than the 1/2 chance of him not scoring. 

[Torque]

14 minutes ago, GhostLetter said:

The chances of the bookies having got all the parts of the prop bet wrong are surely tiny though? You’d be better isolating which specific part of the prop bet you think is value and focusing on that.

If anything, due to the profit margins that bookies build into their odds, the more parts you add to a prop bet, the more you are allowing the bookies to offset/cancel out any part of the prop bet they may have got wrong.

I think it’s a lazy market, aimed at part-time gamblers. And from what I’ve seen here, you’re far better than that!

In regards to the first paragraph I can see what you mean but then as you said some things are related such that if one part of a prop looks value there's every chance another part is as well as a result, although I take your overall point and agree with it. From a variance perspective it would be better to take separate bets if you thought a couple of legs in a prop were worth betting on.

For the second part of what you say, again you're correct. Having said that though, props and multiples have the potential to offer exponential value if they're priced wrong which is something that can happen even if rarely. It's a door that swings both ways - negative value on each leg of a treble amplifies the negative value, but positive value on each leg amplifies the positive value, albeit at the expense of a lower probability of success.

As for your last point, I agree as I mentioned before. If it's promoted by the books then that's because they want you to bet on it and as they're not in the business of charity it's fair to say that they consider these bets to be to their advantage.

Lastly, and in response to your final remark about my competency, it's clear that you also know a thing or two about this subject and so it's interesting to read what you post. I only wish what you posted was a little more friendly. I think you'll do well to find anyone who reads what you post who wouldn't say that you come across as incredibly arrogant and patronising. There's just no need for it and it makes no sense to me.

[GhostLetter]

1 hour ago, harry_rag said:

What would be your best take on where these specific maths are wrong then? If England win there are only 2 possible outcomes relating to Smith scoring or not and their true odds have to add up to the same probability as the England win. I've provided my specific numbers and why I think they're right. Show your sums for this specific scenario in the hope it may finally get it through to my thick head! 

It's just struck me that this scenario seems to support my contention more than yours. Let's imagine our England win = 3 and Kane = 3 scenario and throw in England +1 is 1.5 (fair based on Germany's odds tonight v Spain). My contention is that the true odds for England win/Kane goal is 9 whereas yours is that it would be more like 4.5. Kane scores in the first minute and England's win odds are cut to roughly 1.5 (i.e. they've got a real +1 now). I'm free to stick my Kane winnings on an England win at 1.5 which would pay me as a 4.5 double. You're suggesting the fair odds pre match (when we don't know whether Kane scores or not) are the same as they would be after Kane has actually scored. Obviously Kane scoring increases the chance of an England win but only if he does. I think you're putting too much emphasis on the 2/1 chance of him scoring in this example rather than the 1/2 chance of him not scoring. 

Your maths is wrong because you are still calculating them as independent events, rather than dependent!

In your example, you said that Smith fails to score in 2 matches out of 3, (overall). There is no way that someone who fails to score in 2 out of 3 matches overall also fails to score in 2 out of 3 matches when their team wins. That doesn’t make sense. For a start, you’ve just removed all matches when the team didn’t score at all, from the equation. We are now looking at a completely different set of probabilities - those related to England wins only - and that significantly decreases the chances/increases the odds of Smith failing to score. 

In my Harry Kane statistics - taken from his actual England career - he scores in 47% of matches and fails to score in 53%. However, in matches where England win, he scores in 62% of matches and fails to score in 38%. So that’s the significant jump I was talking about. You wouldn’t get the same anytime scorer odds for matches where England win, that you would for England matches as a whole. Two different sets of probabilities, equates to two different sets of odds.

You could just as easily flip it the other way. In Harry Kane’s England career, they have won 62% of matches he has appeared in. However, that figure jumps to 82%, for matches he scores in. These events are clearly and indisputably linked.

 

I’m struggling to follow what you’re asking, or where the confusion lies, in your second example. Yes, the odds would be around 4.5 for the double, the same as the odds for Kane, doubled with England (+1). And yes, that would be the same effect as having the single bet on the goalscorer, that bet coming in within the first minute and then reinvesting that on an England win.

 

I think I’ve thought of a different way to solve/explain this. Correct me if I’m wrong, but what you seemed to be suggesting to Torque yesterday, was that the bookies had just ‘removed’ some potential value from these prop. bets - the value had effectively vanished into thin air, or into the bookies’ back pockets, more likely! If you look at the Germany vs. Spain game tonight. Again, Lea Schuller to score first is 11/2. The match odds are Germany, 9/5; draw, 12/5; and Spain, 29/20. You were saying that these odds should be combined, if you were trying to form a double. However, if we look at the double odds, for Schiller to score first and each of the three potential outcomes, it’s 9/1 Germany; 25/1 draw; and  40/1 Spain. Those aren’t the calculations you suggest. Germany are much shorter than you suggested they should be, but draw and Spain are both now much longer. That is because, as I suggested, the Schiller first goalscorer odds have been combined with the Germany (+1) odds, of Germany, 8/15; draw, 3/1; and Spain 4/1.

Hopefully that makes more sense, to think of it from that point of view. If Spain are 29/20 to win the match, surely you can see that those odds have to be significantly higher, for a match when a Germany player scores first? If you still disagree, then you’re about to become very rich. If you think the bookies have removed a lot of value from the player to score and their team to win market, then I’m bringing you great news - by your definition, all that value now lies on the other side, on a player to score, and their team to draw or lose!

Edited July 12, 2022 by GhostLetter