Anytime Goalscorer System

[harry_rag]

1 from 3 "A" bets (Olmo) and 3 from 12 "B" bets (Sterling, Gundogan and Ramos) for a small profit since the last update.

175 winners from 545 “A” bets (32.11%) for a loss of 158.54 points
38 winners from 156 “B” bets (24.36%) for a profit of 14.69 points

[harry_rag]

10 points on Alvarez at 4.5 (68B)

[harry_rag]

10 points on Fernandes at 9/2 with Hills

[harry_rag]

Just now, harry_rag said:

10 points on Fernandes at 9/2 with Hills

(78B)

[harry_rag]

10 points each on Mbappe at 2.56 (78A) and Griezmann at 5.2 (77B)

[harry_rag]

Some idle musings on the notion of how best to bet on goalscorers (if the options are back, lay, buy or sell). 

In terms of selling on the spreads I've sold 769 player's goal minutes for a total profit of 4313.8 points. If you adjust that to 1 point level stakes (there were a few bets to lower limits) it's 4392 points and if you adjust it to best price (there were a few bets where I accepted 1 point below best price) it becomes 4440 points. So that's 4440 points profit at an average of 5.77 points profit per bet.

There were 192 sells in the price range 20-34, before I set 35 as my minimum price, for a profit of 1275 points (average 6.64 points per bet). I suspect that return represents a small sample that overperformed slightly (see below).

The 577 sells at 35+ returned 3165 points (average 5.49 points per bet).

Looking at the data for this thread I have 2187 players whose spread price had a midpoint of between 20 and 35. If you'd sold every one of those players you'd have made 8850 points profit (average 4.05 points per bet). Hence the earlier observation about the 192 actual bets at 20-34 overperforming a bit.

I also have 667 players with a midpoint between 15 and 19. Selling them would have returned 2347 points (average 3.52 points per bet).

Overall that makes for a data sample of 3623 players and a profit of 15637 points from selling them for 1 point at best price (average profit of 4.32 points per bet).

Might it be an understatement to say that, given the way this market is priced, buyers are up against it trying to find value and sellers have an outside chance of making a profit?

[harry_rag]

10 points on Keane at 4.5 (68B)

[Torque]

On 12/15/2022 at 3:41 PM, harry_rag said:

Some idle musings on the notion of how best to bet on goalscorers (if the options are back, lay, buy or sell). 

In terms of selling on the spreads I've sold 769 player's goal minutes for a total profit of 4313.8 points. If you adjust that to 1 point level stakes (there were a few bets to lower limits) it's 4392 points and if you adjust it to best price (there were a few bets where I accepted 1 point below best price) it becomes 4440 points. So that's 4440 points profit at an average of 5.77 points profit per bet.

There were 192 sells in the price range 20-34, before I set 35 as my minimum price, for a profit of 1275 points (average 6.64 points per bet). I suspect that return represents a small sample that overperformed slightly (see below).

The 577 sells at 35+ returned 3165 points (average 5.49 points per bet).

Looking at the data for this thread I have 2187 players whose spread price had a midpoint of between 20 and 35. If you'd sold every one of those players you'd have made 8850 points profit (average 4.05 points per bet). Hence the earlier observation about the 192 actual bets at 20-34 overperforming a bit.

I also have 667 players with a midpoint between 15 and 19. Selling them would have returned 2347 points (average 3.52 points per bet).

Overall that makes for a data sample of 3623 players and a profit of 15637 points from selling them for 1 point at best price (average profit of 4.32 points per bet).

Might it be an understatement to say that, given the way this market is priced, buyers are up against it trying to find value and sellers have an outside chance of making a profit?

Looks that way Harry. I'm not all that surprised though - goalscorers after all are just an extension of goal bets and as most people would rather bet on overs than unders it follows that selling goalscorers is likely to be more successful than buying them.

[harry_rag]

3 minutes ago, Torque said:

Looks that way Harry. I'm not all that surprised though - goalscorers after all are just an extension of goal bets and as most people would rather bet on overs than unders it follows that selling goalscorers is likely to be more successful than buying them.

I could write a thesis/bore for England on the subject but I think there’s a clear linear progression.

Total goals isn’t that volatile a market, e.g. 2.4-2.6 for a typical league game. Maybe there’s a bias towards more money on the buy side but the price won’t be far out. I suspect that blindly selling would lose less than blindly buying but I doubt it would be profitable unless you were good at identifying wrong prices.

Total goal minutes is more volatile given the difference between, for example, 2 very early goals and two very late ones. It will attract more muggish buy money given the hope of a flurry of late goals and deter shrewder but cautious sellers because of the degree of randomness in respect of when goals are scored. I suspect blindly selling would be, at worst, close to break even.

Player goal minutes is likely to see a huge chasm develop between the amount of buy money and sell money. Selling match goal minutes at around 120 is one thing, selling a player at 20 feels like a much more dangerous proposition. I read one estimate that suggested 20 buyers for every seller. If you compare a team’s goal minutes to the total for their starting 10 outfield players the midpoints will be closer than they should be bearing in mind the % of goals that are own goals or scored by subs. I’d suggest that the price is typically pitched so that the sell price is more than 10% higher than the true value. Very much “buyers don’t bother” rather than “beware”.

Bearing in mind that not every goal has an assist, player assist minutes seem to be priced at even more of a premium than their goal minutes. I’d suggest that most punters have a worse understanding of this market than they do of player goal minutes (something borne out by my costly experience of fixed odds betting on players to get an assist). Assuming the market is pitched to make the firms the optimum profit it’s mind boggling that there punters wiling to buy at the prices on offer.

 

[harry_rag]

10 points on Almiron at 3/1 with 365 (84B)

Drifting now so you may get a bit better on BF.

[harry_rag]

0 from 1 "A" bet and 0 from 5 "B" bets since the last update.

175 winners from 546 “A” bets (32.05%) for a loss of 168.54 points
38 winners from 161 “B” bets (23.6%) for a loss of 35.31 points

[harry_rag]

1 hour ago, harry_rag said:

175 winners from 546 “A” bets (32.05%) for a loss of 168.54 points

Figures would be 174 winners from 540 bets and -132.46 points if you discount the bets I placed in error (there was a phase, after the current criteria came in, when I was getting confused by the flags in my spreadsheet, since rectified). That means that there has been a loss of 11.19 points on these bets since that criteria was adopted. 

I'm still inclined to persevere with the same approach here because of the overall figures. The current criteria would have produced 287 bets for a profit of 41.61 points with an ROI of 14.5% (to best bookies' price). My actual real money return on those selections shows an ROI of 8.28% (I wouldn't have backed some of the selections because of the criteria that applied at the time).

So sticking with the current criteria which would have made a decent profit had they applied from day 1, would have almost halved the number of bets and have been profitable where I did actually back them. I'm thinking a new thread for 2023 with a set bank and revised staking approach.

Not sure about the "B" bets, I'll review them tomorrow. The return on those is actually a bit worse if you ignore the bets I placed in error and we are talking about less likely scorers.

[harry_rag]

20 hours ago, harry_rag said:

Not sure about the "B" bets, I'll review them tomorrow. The return on those is actually a bit worse if you ignore the bets I placed in error and we are talking about less likely scorers.

Going to drop these from next year onwards. Not been profitable and no obvious trends leaping out of the data. Will focus on the "A" bets (more likely goalscorers). Should halve the amount of time I spend doing data entry clerk duties if nothing else!

20 hours ago, harry_rag said:

I'm thinking a new thread for 2023 with a set bank and revised staking approach.

New thread for the current "A" bet criteria from the new year. It's not been profitable since I adopted it on 23rd September (17 winners from 39 qualifying selections for a return of -2.55 points and an ROI of -4.55%) but:

• ROI is 14.5% if you'd backed all 287 qualifying selections from the data sample of 2198 players
• ROI is 8.28% on the qualifying selections I've actually backed based on the different criteria that have applied during the lifetime of this thread
• If I split my data up into 4 chronological quarters it's been profitable during each one (4.75 points then 19.2, 14.08 and 3.58).

Basically I think it deserves a longer run to see if ongoing results can match historical ones.

[harry_rag]

10 points on Mahrez at 4.3 (B bet)

[harry_rag]

10 points on Sakala at 3.2 (78A) and Tillman at 4.4 (84B)

[harry_rag]

10 points on Wood at 3.35 (73A)

10 points on Daka at 4.4 (60B)

10 points on Trossard at 4 (86A)

[harry_rag]

10 points on Salah at 29/20 with Lads (82A)

10 points on Nunez at 13/8 with 365 (74A)

10 points on Watkins at 4.4 (77B)

10 points on Dykes at 4.13 (or 3/1 365) (76B)

[harry_rag]

10 points on Pukki at 5/2 with Sky Bet (84A)

10 points on Saka at 3,25 (72A)

[harry_rag]

4 from 7 "A" bets (Wood, Salah, Pukki and Saka) and 2 from 5 "B" bets (Mahrez and Watkins) since the last update.

179 winners from 553 “A” bets (32.37%) for a loss of 113.96 points
40 winners from 166 “B” bets (24.1%) for a profit of 0.35 points

That puts the "A" bets 65.77 points in front since the current criteria came in (ignoring selections placed in error).

[Heisenberg68]

Hi @harry_rag How are you ?

 

 

 

 

Have you looked at the players score two or more ?

it can be interesting for some players who take penalties for example

Kane for example (I'm French, and he can miss sometimes lol) , Benzema , Salah , Mbappé , Lewandoski.

Edited December 27, 2022 by Heisenberg68

[harry_rag]

15 minutes ago, Heisenberg68 said:

Have you looked at the players score two or more ?

@Torque asked me a similar question back in May and the gist of the reply was that it's harder to find prices that appear to offer value in the "multi scorer" markets than it is in the straight anytime markets, possibly because the casual punter might think a price (say 33/1 for a hat trick) is "big" and worth backing when it should be more like 40s or 50s. I'll review the numbers I posted then to see if the position has changed at all in terms of what I'd expect to see.

From 1487 players, 562 scored one or more goals (37.97%). My actual bet strike rate lags that at 32.96%.

Now 2209 players with 828 scoring at least once (37.48%) and actual bet strike rate 32.37% so fairly consistent in terms of those headline numbers.

469 scored exactly 1 goal, 77 scored twice, 13 managed a hat trick and 3 have scored 4 or more goals.

Now 688, 118, 18 and 4

My average fair odds are 2.67. If they were accurate then you'd have expected 37.45% of the players to have scored one or more, or 557 so we're certainly in the right ballpark on that basis.

Still 2.67 suggesting 827 players should have scored one or more (compared to 828 in reality) so the sample appears to be even more accurate now.

Taking poisson as a guide, the expectations for 2 or more or a hat trick look like this:

Brace: 8.12% or 120 versus actual number of 93 - now it's 179 expected versus 140 actual

Hat trick: 1.22% or 18 versus actual number of 16 - 27 expected versus 22 actual

38 minutes ago, Heisenberg68 said:

Kane for example (I'm French, and he can miss sometimes lol) , Benzema , Salah , Mbappé , Lewandoski.

Given how popular those players tend to be with punters it's quite hard to get them at an attractive price to score anytime; my feeling is that their price to score 2 or more is usually poor value. (There was a time when weekly singles and doubles on the Messi and Ronnie hat trick was a bit of a cash cow but the bookies cottoned on a while back.)

Out of curiosity, I just totted up how much I've won or lost selling the goal minutes of the 5 players you mentioned (essentially betting on them NOT to score). It's 275 points profit from 144 bets with Salah at a remarkable +455 points (33 bets) and Lew and Mbappe costing me a combined 297 points (51 bets). 

(Not sure you wanted such a detailed response but I was interested to revisit the figures a few months on!)

[Heisenberg68]

1 minute ago, harry_rag said:

@Torque asked me a similar question back in May and the gist of the reply was that it's harder to find prices that appear to offer value in the "multi scorer" markets than it is in the straight anytime markets, possibly because the casual punter might think a price (say 33/1 for a hat trick) is "big" and worth backing when it should be more like 40s or 50s. I'll review the numbers I posted then to see if the position has changed at all in terms of what I'd expect to see.

From 1487 players, 562 scored one or more goals (37.97%). My actual bet strike rate lags that at 32.96%.

Now 2209 players with 828 scoring at least once (37.48%) and actual bet strike rate 32.37% so fairly consistent in terms of those headline numbers.

469 scored exactly 1 goal, 77 scored twice, 13 managed a hat trick and 3 have scored 4 or more goals.

Now 688, 118, 18 and 4

My average fair odds are 2.67. If they were accurate then you'd have expected 37.45% of the players to have scored one or more, or 557 so we're certainly in the right ballpark on that basis.

Still 2.67 suggesting 827 players should have scored one or more (compared to 828 in reality) so the sample appears to be even more accurate now.

Taking poisson as a guide, the expectations for 2 or more or a hat trick look like this:

Brace: 8.12% or 120 versus actual number of 93 - now it's 179 expected versus 140 actual

Hat trick: 1.22% or 18 versus actual number of 16 - 27 expected versus 22 actual

Given how popular those players tend to be with punters it's quite hard to get them at an attractive price to score anytime; my feeling is that their price to score 2 or more is usually poor value. (There was a time when weekly singles and doubles on the Messi and Ronnie hat trick was a bit of a cash cow but the bookies cottoned on a while back.)

Out of curiosity, I just totted up how much I've won or lost selling the goal minutes of the 5 players you mentioned (essentially betting on them NOT to score). It's 275 points profit from 144 bets with Salah at a remarkable +455 points (33 bets) and Lew and Mbappe costing me a combined 297 points (51 bets). 

(Not sure you wanted such a detailed response but I was interested to revisit the figures a few months on!)

it's clear ! thank you

[harry_rag]

10 points on Collins (Derby) at 11/4 with Sky Bet (79A)

Has 7 goals from 17 league starts this season (none in 5 Cup starts). One brace among those goals. Only scored 3 times for Cardiff last season, 14 for Luton the season before which included 2 hat tricks.

My fair odds are roughly 2/1 anytime and 15/1 2 or more so I could make a case for the brace at 20/1 but I'll swerve it and trust to the anytime "grind"!

[harry_rag]

10 points on Piroe (Swansea) at 9/4 with 365 (75A)

10 points on Pulisic at 3.8 (56A)

[harry_rag]

2 hours ago, harry_rag said:

10 points on Pulisic at 3.8 (56A)

That was actually a "B" bet. 0-3 so far today, 3 more to hopefully avoid a famine to follow yesterday's relative feast.

10 points on Rashford at 15/8 with 365 (78A)

10 points on Fernandes at 3.45 (81A)

10 points on Rodriguez (Burnley) at 15/8 with 365 (89A)

[alexcaruso808]

20 hours ago, harry_rag said:

That was actually a "B" bet. 0-3 so far today, 3 more to hopefully avoid a famine to follow yesterday's relative feast.

10 points on Rashford at 15/8 with 365 (78A)

10 points on Fernandes at 3.45 (81A)

10 points on Rodriguez (Burnley) at 15/8 with 365 (89A)

Thanks Harry.  I got on Rashford and made a nice profit. ?

[harry_rag]

36 minutes ago, alexcaruso808 said:

Thanks Harry.  I got on Rashford and made a nice profit. ?

Should’ve just done him & no-one else myself yesterday! 

[harry_rag]

10 points on Moffi (Lorient) at 7/4 with PP (72A)

10 points on Castellanos at 3 (75A) - 2/1 if you can get on with Boyles

[harry_rag]

10 points on Morris (Luton) at 3/1 with 365 (76B)

[harry_rag]

10 points on Kalimuendo (Rennes) at 4.4 (54B)