Lay - Trainers

[MobiusGrey]

So in my quest for value backing/laying I was having a look at some historical stats. Almost every trainer year on year shows a level stakes loss at SP.

I know this won't work but I'm not good enough with the maths side of things 'innit'. What I want to know if anyone could be so kind as to explain, why won't just laying a certain trainer's horses in every race over the course of year at SP not be profitable, is it just commission and if you could get 0% you'd show a win?

[verance]

Are you looking at industry SP or Betfair SP? Industry SP is about 15%-20% worse, but obviously you can't lay it unless you're a bookie.

Here's all trainers with over 700 runners in 2021:

Trainers.png

[MobiusGrey]

I don't think this will use BSP will it? Sorry the resolution on that image is awful, it's basically the TimeForm info you can pull on trainers, Michael Appleby is showing a loss of £75 so far this year to level stakes £1

[Derek27]

So in my quest for value backing/laying I was having a look at some historical stats. Almost every trainer year on year shows a level stakes loss at SP.

I know this won't work but I'm not good enough with the maths side of things 'innit'. What I want to know if anyone could be so kind as to explain, why won't just laying a certain trainer's horses in every race over the course of year at SP not be profitable, is it just commission and if you could get 0% you'd show a win?

Forget trainers. Laying every horse at ISP would be profitable, but unless you're a bookie you can't lay at SP, the exchange prices are always higher.

[MobiusGrey]

Ok thanks Derek this makes sense. Another question though, in Verance's image some of those trainers are showing a loss at BSP, would laying every one of their horses across the year show +ev?

[verance]

I don't think this will use BSP will it? Sorry the resolution on that image is awful, it's basically the TimeForm info you can pull on trainers, Michael Appleby is showing a loss of £75 so far this year to level stakes £1

Yeah, that's SP. Last 12 months: SP -£75, BSP +£224!

[verance]

Ok thanks Derek this makes sense. Another question though, in Verance's image some of those trainers are showing a loss at BSP, would laying every one of their horses across the year show +ev?

I think it's something to consider but probably won't be profitable on its own. Lots of randomness in any year, here's the top trainers since 2016. All tend toward 0 (this has commission set to 0 FYI)

Trainers2.jpg

[MobiusGrey]

Ok thanks Derek this makes sense. Another question though, in Verance's image some of those trainers are showing a loss at BSP, would laying every one of their horses across the year show +ev?

I think it's something to consider but probably won't be profitable on its own. Lots of randomness in any year, here's the top trainers since 2016. All tend toward 0 (this has commission set to 0 FYI)

Trainers2.jpg

No wonder you never see a poor bookie!! Ok thanks for this Verance, very helpful and certainly something to incorporate into the other data I'm looking at

[Andriy]

This is a situation where a month by month (or even a quarter by quarter) breakdown is probably more useful, since the trends of certain trainers happen season on season, and the racing media and then everybody else jump on the bandwagon, leading to over-reactions in the market.

Two current cases known to every man and his dog:

Paul Nicholls: most of his horses have their flu jab at the start of the year, and his horses seem to underperform in January into February... would this lead to his horses going off at bigger prices than they should, and the rare winner giving you an overall profit?

Conversely, Venetia Williams: horses are very quiet to about mid-November, then for 2 or 3 months, strike rate goes through the roof when the very soft/heavy ground comes into play... so would they be overbet, and a blanket lay be profitable?

I don't know the answers, just putting a line of thought in. Likewise, I can imagine some smaller trainers do better in the summer when they take advantage of the fact that the bigger stables are resting most of their horses...

[wearthefoxhat]

This is a situation where a month by month (or even a quarter by quarter) breakdown is probably more useful, since the trends of certain trainers happen season on season, and the racing media and then everybody else jump on the bandwagon, leading to over-reactions in the market.

Two current cases known to every man and his dog:

Paul Nicholls: most of his horses have their flu jab at the start of the year, and his horses seem to underperform in January into February... would this lead to his horses going off at bigger prices than they should, and the rare winner giving you an overall profit?

Conversely, Venetia Williams: horses are very quiet to about mid-November, then for 2 or 3 months, strike rate goes through the roof when the very soft/heavy ground comes into play... so would they be overbet, and a blanket lay be profitable?

I don't know the answers, just putting a line of thought in. Likewise, I can imagine some smaller trainers do better in the summer when they take advantage of the fact that the bigger stables are resting most of their horses...

Good summary.

One approach, is to create your own tissue prices. Takes a little finesse, but rewarding when you get it working well. The ole debate of value Lays comes into play and of course is always a little subjective. Horse racing by its' very nature is incomplete information, so some guesswork, with late confirmation, is required. It boils down to combining probabilities and converting into a price for market comparison purposes.

In theory, Laying by beating the ISP is the aim. (Less than ISP)

Example.

L1.00.png

Looking at all those ranked 4 and higher.

Lay qualifiers
Tokyo Live - Tissue Price 12.11, Early Show 5.00. So I'd start laying no higher than the tissue price. ISP was 11.00
Mister Moratti - Tissue Price 36.33, Early Show 13.00. So I'd start laying no higher than the tissue price. ISP was 11.00

Preferably, a L2B is the trade, In-Play with a "let it ride" option.

The information used to create ratings/tissue price is key. Trainerform would be one of them, including course records, it's then just a question of which other variables are included. As said, a tissue price is essentially a combination of probabilities.

[Derek27]

Ok thanks Derek this makes sense. Another question though, in Verance's image some of those trainers are showing a loss at BSP, would laying every one of their horses across the year show +ev?

I've never really been into stats but it takes a fair bit of digging and investigating, the big mistake that many make is simply latching on to a trend. Andriy pointed out possible monthly variations but even over a year, you may find horses beginning with W have made a good profit at a certain track.

[MobiusGrey]

Good summary.

One approach, is to create your own tissue prices. Takes a little finesse, but rewarding when you get it working well. The ole debate of value Lays comes into play and of course is always a little subjective. Horse racing by its' very nature is incomplete information, so some guesswork, with late confirmation, is required. It boils down to combining probabilities and converting into a price for market comparison purposes.

In theory, Laying by beating the ISP is the aim. (Less than ISP)

Example.

L1.00.png

Looking at all those ranked 4 and higher.

Lay qualifiers
Tokyo Live - Tissue Price 12.11, Early Show 5.00. So I'd start laying no higher than the tissue price. ISP was 11.00
Mister Moratti - Tissue Price 36.33, Early Show 13.00. So I'd start laying no higher than the tissue price. ISP was 11.00

Preferably, a L2B is the trade, In-Play with a "let it ride" option.

The information used to create ratings/tissue price is key. Trainerform would be one of them, including course records, it's then just a question of which other variables are included. As said, a tissue price is essentially a combination of probabilities.

Do you use software to work out your tissue or just excel? I'm currently only looking at sectional times, pace cards, and efficiency ratings to compare one horse against the favourite in a race to find if there is any value there. I've considered pricing up a race entirely using a number of variables like you've mentioned but I've no idea where to start with that and there's little info online about it. Do you start with every horse at evens and the add/subtract a factor to this based on the variables until you get to a price?

Apologies if this is asking too much and it would erode your edge, I'm fully aware there's only so much information people can give in a public setting so appreciate any info you are willing to give.

[wearthefoxhat]

Good summary.

One approach, is to create your own tissue prices. Takes a little finesse, but rewarding when you get it working well. The ole debate of value Lays comes into play and of course is always a little subjective. Horse racing by its' very nature is incomplete information, so some guesswork, with late confirmation, is required. It boils down to combining probabilities and converting into a price for market comparison purposes.

In theory, Laying by beating the ISP is the aim. (Less than ISP)

Example.

L1.00.png

Looking at all those ranked 4 and higher.

Lay qualifiers
Tokyo Live - Tissue Price 12.11, Early Show 5.00. So I'd start laying no higher than the tissue price. ISP was 11.00
Mister Moratti - Tissue Price 36.33, Early Show 13.00. So I'd start laying no higher than the tissue price. ISP was 11.00

Preferably, a L2B is the trade, In-Play with a "let it ride" option.

The information used to create ratings/tissue price is key. Trainerform would be one of them, including course records, it's then just a question of which other variables are included. As said, a tissue price is essentially a combination of probabilities.

Do you use software to work out your tissue or just excel? I'm currently only looking at sectional times, pace cards, and efficiency ratings to compare one horse against the favourite in a race to find if there is any value there. I've considered pricing up a race entirely using a number of variables like you've mentioned but I've no idea where to start with that and there's little info online about it. Do you start with every horse at evens and the add/subtract a factor to this based on the variables until you get to a price?

Apologies if this is asking too much and it would erode your edge, I'm fully aware there's only so much information people can give in a public setting so appreciate any info you are willing to give.

It's ok with any question, some answers may be a bit vague.

I just use excel and input the data I need for each race. (takes about 5 mins per race). To produce a spreadsheet format that I was okay with, took a while. (the old VRR spreadsheet from Michael Wilding was a good starting point).

When I started to produce a tissue price, I used to create a rating for each horse, total them, and then subdivide each rating into the total to produce a 100% book. In my view, it was inconsistent and my interpretation of some of the data was poor. It took a while to find a combination of data sets I liked and that were more consistent.

The old VRR sheet is essentially a zero sum ratings/rankings, seperate to the tissue price calculation and helps define the stronger/weaker types in each race. The tissue is calculated separately. The starting point is based on the number of runners for that race. 7 runners, each has a one in 7 chance until the other factors are introduced and the tissue price is finalised.

I've got it to the stage I can price up any race if necessary, but also found certain race-types work better than others. The bookies will have their super computers churning out data in a similar way, but...they all seem to follow each other over a cliff and that's where the real edge is.

The best (free) information for trainers and such like, is on the Attheraces website. (Todays Card - Smart Stats)

https://www.attheraces.com/racecards

Todays Cards

TCards.png

Smart Stats

AyrSmart.png

[MobiusGrey]

Good summary.

One approach, is to create your own tissue prices. Takes a little finesse, but rewarding when you get it working well. The ole debate of value Lays comes into play and of course is always a little subjective. Horse racing by its' very nature is incomplete information, so some guesswork, with late confirmation, is required. It boils down to combining probabilities and converting into a price for market comparison purposes.

In theory, Laying by beating the ISP is the aim. (Less than ISP)

Example.

L1.00.png

Looking at all those ranked 4 and higher.

Lay qualifiers
Tokyo Live - Tissue Price 12.11, Early Show 5.00. So I'd start laying no higher than the tissue price. ISP was 11.00
Mister Moratti - Tissue Price 36.33, Early Show 13.00. So I'd start laying no higher than the tissue price. ISP was 11.00

Preferably, a L2B is the trade, In-Play with a "let it ride" option.

The information used to create ratings/tissue price is key. Trainerform would be one of them, including course records, it's then just a question of which other variables are included. As said, a tissue price is essentially a combination of probabilities.

Do you use software to work out your tissue or just excel? I'm currently only looking at sectional times, pace cards, and efficiency ratings to compare one horse against the favourite in a race to find if there is any value there. I've considered pricing up a race entirely using a number of variables like you've mentioned but I've no idea where to start with that and there's little info online about it. Do you start with every horse at evens and the add/subtract a factor to this based on the variables until you get to a price?

Apologies if this is asking too much and it would erode your edge, I'm fully aware there's only so much information people can give in a public setting so appreciate any info you are willing to give.

It's ok with any question, some answers may be a bit vague.

I just use excel and input the data I need for each race. (takes about 5 mins per race). To produce a spreadsheet format that I was okay with, took a while. (the old VRR spreadsheet from Michael Wilding was a good starting point).

When I started to produce a tissue price, I used to create a rating for each horse, total them, and then subdivide each rating into the total to produce a 100% book. In my view, it was inconsistent and my interpretation of some of the data was poor. It took a while to find a combination of data sets I liked and that were more consistent.

The old VRR sheet is essentially a zero sum ratings/rankings, seperate to the tissue price calculation and helps define the stronger/weaker types in each race. The tissue is calculated separately. The starting point is based on the number of runners for that race. 7 runners, each has a one in 7 chance until the other factors are introduced and the tissue price is finalised.

I've got it to the stage I can price up any race if necessary, but also found certain race-types work better than others. The bookies will have their super computers churning out data in a similar way, but...they all seem to follow each other over a cliff and that's where the real edge is.

The best (free) information for trainers and such like, is on the Attheraces website. (Todays Card - Smart Stats)

https://www.attheraces.com/racecards

Todays Cards
TCards.png

Smart Stats
AyrSmart.png

This is great thank you. I already use ATR but only for the pace and sectional data and currently nothing else. I normally compare all horses against each other running over the same distance and ground conditions (ideally the same track if possible) and then make a decision based on this as to what offers value for either a back or lay. Most days I'll only find up to about 3 races to get involved in.

I'll start looking at VRR to be honest it's not a term I'm familiar with.

[wearthefoxhat]

Do you use software to work out your tissue or just excel? I'm currently only looking at sectional times, pace cards, and efficiency ratings to compare one horse against the favourite in a race to find if there is any value there. I've considered pricing up a race entirely using a number of variables like you've mentioned but I've no idea where to start with that and there's little info online about it. Do you start with every horse at evens and the add/subtract a factor to this based on the variables until you get to a price?

Apologies if this is asking too much and it would erode your edge, I'm fully aware there's only so much information people can give in a public setting so appreciate any info you are willing to give.

It's ok with any question, some answers may be a bit vague.

I just use excel and input the data I need for each race. (takes about 5 mins per race). To produce a spreadsheet format that I was okay with, took a while. (the old VRR spreadsheet from Michael Wilding was a good starting point).

When I started to produce a tissue price, I used to create a rating for each horse, total them, and then subdivide each rating into the total to produce a 100% book. In my view, it was inconsistent and my interpretation of some of the data was poor. It took a while to find a combination of data sets I liked and that were more consistent.

The old VRR sheet is essentially a zero sum ratings/rankings, seperate to the tissue price calculation and helps define the stronger/weaker types in each race. The tissue is calculated separately. The starting point is based on the number of runners for that race. 7 runners, each has a one in 7 chance until the other factors are introduced and the tissue price is finalised.

I've got it to the stage I can price up any race if necessary, but also found certain race-types work better than others. The bookies will have their super computers churning out data in a similar way, but...they all seem to follow each other over a cliff and that's where the real edge is.

The best (free) information for trainers and such like, is on the Attheraces website. (Todays Card - Smart Stats)

https://www.attheraces.com/racecards

Todays Cards
TCards.png

Smart Stats
AyrSmart.png

This is great thank you. I already use ATR but only for the pace and sectional data and currently nothing else. I normally compare all horses against each other running over the same distance and ground conditions (ideally the same track if possible) and then make a decision based on this as to what offers value for either a back or lay. Most days I'll only find up to about 3 races to get involved in.

I'll start looking at VRR to be honest it's not a term I'm familiar with.

It was an old spreadsheet created about 11 years ago. If interested take a look. Its' design is essentially based on 3 inputs, the spreadsheet does the rest. The "hardwork" is to find the sort of information you need to input to get the best results. As said, credit to Michael Wilding for the original workings.

VRRLite.xlsx