Dog Ratings

[Benny]

How do I work out the beaten margin in dog races? Do I use 1.5 lengths or something different?

I have worked out some class ratings for each grade of race. They are only a guide and need more experimenting.

Benny

[good 4th]

1 length= 0.066 sec
So if the dog was beaten by 15 lengths.
15 x 0.066 = 0.99
So 15 lengths = almost 1 second
Add the beaten time on to the winner.
That should do it.
Then this will give you how far the dog was behind the winner in lenghts and time.
One thing is dont rely on time ratings alone when it comes to dogs, this will send you BROKE......

[Duritz]

I would have thought time ratings would be the "light and the way" in dog racing. Is this not so?

[BJ]

I would tend to agree. The faster a dog is, the more chance it has of winning the race.

I have an excel spreadsheet that converts last 3 starts to speed and converts it to current race distance. If a dog heads all 3 of my criteria for the race, it becomes a betting opportunity for me.
You would be surprised at some of the longer prices paid for some of these dogs.

[dingoboy]

Played arround with this dog thing the last few weeks in regards to this post, you are right BJ, some of the pays i got were stupid, best so far was 44 units for the win at lismore race 7 the other night, always keep your eyes open for options

best regards

[Duritz]

so how do you convert the times from previous starts to todays distance? Do you divide it by the distance raced over then and multiply it by todays?

[DR RON]

Duritz, I think the trouble with doing that would be that dogs llike horses would have their preferred distances and a sprinte rrunning over a longer distance might lead the race up to that distance and then not be able to go on with it.

[BJ]

I convert it to metres per second. I also have a list of the track records for every distance, so I know the decay of speed in relation to the distance. Generally they stick to the same distances so the decay is not too much of a factor.

To be honest it is sitting there not doing a lot at the moment. I am trying to work out how to get the form directly downloaded into my spreadsheet so I don't have to do anything. Currently takes 5 to 10 minutes to enter the data in for each race, and as I am still testing, it is a bit much.

As far as I am concerned, speed is the most important thing when it comes to the dogs. And I only bet on a dog that is a fast beginner.
A fast beginner coming out of box 4 or 5 is a good betting proposition, because they are generally underbacked.

I can't really imagine too much else to be considered about a dogs form.

Oh, and if a dog meets my 3 criteria but is jumping too high in distance, I will not bet.....

[DR RON]

B J

on the greyhound victoria sitethey have info on each dogs finishing position from each box. I find it can give you a clear picture in some cases, of when to bet on the dog in question. Especially good too if a favourite is coming from a box that they have failed at constantly before, you can go searching for some value about the others in the race.

[BJ]

Possibly another avenue I can add to my testing. I get about 1 betting opportunity per 3 races, so I could maybe check that out also... Might cut out some losers... Though box numbers generally receive no consideration from me....

[woof43]

Hi,
I use Time extensively in my handicapping and one of the first things one needs to do is find out how far is a Length, especially when only the winners time is recorded an the Judge just takes measurements of Lenghts beaten off the photo.

I have added a spreadsheet that gives some insight into the effect of speed on Lengths beaten an the distance difference at differing speeds.

[good 4th]

Great stuff Woof,
I to do the same thing but have been very unlucky selecting my top pick on speed alone and lenghts beaten.

Started of great and i thought here we go but as with most things???????
If a dog is comming down in distance for todays race what sort of penalty/ bouns do you use and what other methods could i be looking at in regard to dogs comming down/ up in distance.
Also different track types.
1,2,3 corners, slope, etc ....
Any other ideas would be welcomming.

[woof43]

If you just divide time into distance you and make no adjustments,then dog with the avg shorter distances raced would normally appear to be the fastest, so you do need to make adjustments from track to track.

A starting point would be to gather all the win times for each track an distance in the State for a set period or for a set number of winners,then find the avg Win Time an the Standard deviation for all those times.

Now to find the Fixed Standard for each track, multiply the Standard Deviation by 1.96, then subtract this from your average time, this will become your Fixed Point for this Track but the main problem now is how to identify/quantify the effect of Better Classes of runners from track to track.

I'll post some more later this evening, once racing has finished

[good 4th]

Thanks for that,
What do you think about using the weight of a dog.........

Yes ive been trying to identify/quantify the Better Class of runners from track to track but seems to me as it is a bit hit an miss at this time.
Cheers

[Duritz]

I would have thought you wouldn't need to quantify the different classes, just compare pure times. Why doesn't that work?

[Rock Steady]

No-one has talked about the most important factor in doing dog form - 1st sectional time and racing style.
The "fastest" dog does not always win the race. Many times the "fastest" dog does not get into the race because of what happens in the run to the first turn. I consistenly see dogs at Angle Park that in a solo trial would break 29.8 seconds yet get beaten in a race in 30.25sec so you are barking up the wrong tree if you think the dog with the best speed rating is the logical winner.

[Rock Steady]

Who is going to "control" the race at the first turn?
Be able to answer that question for dog races and you're nearly there.

I'm afraid there are no shortcuts.If you don't watch video replays and understand what every dog does in terms of running style and best and average first sectional, middle sectional and last sectional, I think you are facing an uphill battle.

[BJ]

I agree that track to track times can differ, be it shape of track or the surface, But, as far as class goes I am with Duritz. It all depends on the time.

If a maiden dog runs 31.00 seconds to win a race, and then in a free for all over the same distance, the winner also runs 31.00 seconds. To me the performance is equal, and the class of the race is irrelevant.

About lengths beaten and time, I don't at all agree that 1 length equals .066 seconds. In a short race with the dogs running at 18 m/s 1 length is less time than in a 700 m race when they are going at about 16 m/s. I call 1 length 1 metre.
eg. A dog wins a 520 metre race in 31 seconds. A dog that gets beaten by 10 lengths, I consider to have ran a 510 metre race in 31 seconds. Convert that to m/s and then back to the original 520 metres.....

My 3 criteria are :
The fastest speed in m/s (using the last 3 starts, and converting them to current distance).
The fastest average speed in m/s(3 starts).
The current race distance must be within 10% of the average distance of the last 3 starts.
I will add in a fourth criteria which may appear obvious, every dog in the race must have had at least 3 race starts (not trials).

The reason I do this is because I figure that the dogs singled out will be quick out of the box and less likely to get into trouble. A fall in any start is considered as a 15 length loss. I do not want to back a dog that gets midfield and into trouble.

As far as boxes are concerned, I think the middle boxes are underbacked so will make up for any bias. I always bet to return a figure rather than have a level stake.

The whole point of having a system, is to have it as simple as possible with as little work. While I am still learning about data input from the web, this is as challenging as I feel it needs to be.
I do however treat others ideas with merit, as long as people have the means and the knowledge to put it into practice.

[woof43]

Hi Duritz,
When you gather Win Times from track to track,each tracks average will be in isolation, cause you will have better dogs racing at one place and not the other so when the dogs from the lesser track race at the better track they will have been over compensated in time adjustment and will perform worse than expected.
You need to record those dogs that have then won at both tracks and then you will find the difference in Class from track to track.

An example would be say The two Victorian city clubs the time differences in avg times would be around .22 which is very similiar to the track record from both tracks this is because the same dogs race there from meeting to meeting, my actual Class difference is .03 with The Meadows having the better types so the difference is around .188.

Regarding 1st Sectionals these definately are most important, and of all times these are the easiest to be compared and quantified from track to track, all thats needed is the distance from boxes to the first measuring point.

Because most of my processes are automated, when i have dragged out the Formlines for each dog in Todays race they are ranked then on 1st Sectional data, now based on that ranking say its in the top two i then find the avg finish time for that dog whenever it has raced in the first2, if there is insufficent data i create synthetic data based on other dogs performances from my database, then i can move to my next step in the process.

A second method if you don't have 1st sectional data, depending if you have true and meaningful track adjustments. Once you have standardised the past formlines you then isolate the formlines based on distances ie, under 400 >400<500, and so on. Then find the average for each time bin, this can/will provide a pretty good idea of who the 1st Sectional contenders will be in Todays race.

[good 4th]

Hey BJ,
Winner in 31sec over 520 mtrs...
equals 16.77 meters per sec

Other runner
Beaten 10 meters/lenghts
equals 16.45 meters per sec
so it's time is 31.61 sec

Winner 31.00 sec
Runner 31.61 sec

Am i on the right track ???

If you use 0.066 for one length or 0.07 it gets almost the same result.
Beaten by 10 meters
10*0.066 = 0.66
10*0.07 = 0.7
New time 31.66
New time 31.7

Just trying to understand your handicapping rules, if im wrong could you post a work out.
Cheers

[Duritz]

Aah I see - one query though - how do you get their lead sectional? Doesn't the GRV onlypost the lead sec for the leader?

[moeee]

Just bear in mind that a greyhounds brain is about the size of a peanut!

[Rock Steady]

Duritz,
I work out the first sectionals from the video tapes. On my video recorder there are 25 slow motion "clicks" to the second so it easy to work out the sectionals. Time the leading dog to the post the first time and record that sectional then count the number of "clicks" between that greyhound and the rest of the contenders. I can also say with confidence that the times in form guides are not always accurate. Double check them by timing yourself.

moeee,
My POT is running at 12% since we last "talked" on this forum. Casino on Fridays has been a bonanza whilst Angle Park is slightly in front. The Gold Coast has been consistent with a POT of 14%. Dogs brains might be small but they are creatures of habit. A wide runner will remain a wide runner and a fence crusher will remain exactly that. Dogs do not have jockeys on them that cost them the race and they also go flat out from start to finish so "pace of the race" is not an issue as it is in thoroughbreds. Only ever eight runners maximum too.

[BJ]

Hey BJ,
Winner in 31sec over 520 mtrs...
equals 16.77 meters per sec

Other runner
Beaten 10 meters/lenghts
equals 16.45 meters per sec
so it's time is 31.61 sec

Winner 31.00 sec
Runner 31.61 sec

Am i on the right track ???

If you use 0.066 for one length or 0.07 it gets almost the same result.
Beaten by 10 meters
10*0.066 = 0.66
10*0.07 = 0.7
New time 31.66
New time 31.7

Just trying to understand your handicapping rules, if im wrong could you post a work out.
Cheers

You look like you have understood correctly. The difference is when you have a 300 metre race, they run alot quicker so the distance becomes greater.
.066 seems to be a reasonable sort of estimate, but it is easy in excel to be more precise. The difference between 31.61 and 31.66 is nearly 1 length.
An estimate I have made which I cannot confirm is that 1 length equals 1 metre? Obviously some dogs are bigger than others but until somebody can give me better information, that is what I will continue to use.
I always convert all speeds to the current race distance before finding the fastest, the fastest average etc.
As I said, I convert using the track records as my base. From lowest distance to highest distance there will be noticeable pattern. I am not on my own computer so shall give more info about the formula at a later date...

Please remember, I am not claiming anything, just the process I use and believe in....

[Rock Steady]

BJ,
Quote:
"If a maiden dog runs 31.00 seconds to win a race, and then in a free for all over the same distance, the winner also runs 31.00 seconds. To me the performance is equal, and the class of the race is irrelevant."

This is where I completely disagree. A maiden dog can run fast time in low class but put it into a third or fourth grade event and it will struggle to repeat the fast time. It's called "PRESSURE".

The other point I would be interested to get your answer on is track bias.
How do you account for track bias in your database. The Angle Park track will vary by up to 1 second in times from one meeting to the next with similar class dogs running on both nights. If a dog has had 60 career starts and won 10 races at the track in the following times - 30.2, 30.25, 30.33, 30.25, 30.35, 30.40, 30.21, 30.25, 30.37, 30.36 and 29.85, do you assess this dog as a 29.85 dog or something else? You might find that the night it ran 29.85 the track was on fire with all winners on the night breaking 30.00 seconds.
In my opinion it would be a costly mistake to rate this dog as a 29.85 dog if racing against GENUINE sub 30sec dogs.

[Rock Steady]

Another query for those dog followers that don't do videos:

How do you know if a dog "chases' or not if you're not watching videos. The form guide does not tell you this important factor. You need to "see it" for yourself.

[woof43]

Hi Rock Steady,
I have mentioned a number of times about the importance of the Standard Deviation in our Handicapping, what you have mentioned re "chasers" gives a prime example of how one can catergorise types of runners by just using the STDev with one other Factor.
A typical "chaser" has a smaller then normal lengths beaten from winner avg(will then be overplayed by the public), but a larger spread of Race Win Times thus a wider STDev.
(So knowing this type of runner/s are in todays race gives an edge once known.)

A sidelight to these types of runners and what one should do when modelling these runners is this, what chance has this dog of running times, 2 deviations slower then his avg time then running 2 deviations quicker, in fact this type gets nowhere near the 2 deviations quicker mark maybe not even halfway.

[BJ]

BJ,
Quote:
"If a maiden dog runs 31.00 seconds to win a race, and then in a free for all over the same distance, the winner also runs 31.00 seconds. To me the performance is equal, and the class of the race is irrelevant."

This is where I completely disagree. A maiden dog can run fast time in low class but put it into a third or fourth grade event and it will struggle to repeat the fast time. It's called "PRESSURE".

The other point I would be interested to get your answer on is track bias.
How do you account for track bias in your database. The Angle Park track will vary by up to 1 second in times from one meeting to the next with similar class dogs running on both nights. If a dog has had 60 career starts and won 10 races at the track in the following times - 30.2, 30.25, 30.33, 30.25, 30.35, 30.40, 30.21, 30.25, 30.37, 30.36 and 29.85, do you assess this dog as a 29.85 dog or something else? You might find that the night it ran 29.85 the track was on fire with all winners on the night breaking 30.00 seconds.
In my opinion it would be a costly mistake to rate this dog as a 29.85 dog if racing against GENUINE sub 30sec dogs.

Are you saying to me that a dog that wins a maiden will run slower in its next race?
I don't care how many starts a dog has ran in its career as long as all dogs in the field have had at least 3. I am only interested in the dogs last 3 races, I do not care about its fastest ever time. Therefore a dog with a fastest recorded time of 29.85 would not be rated as such, unless it ran that time in each of its last 3 races over the current distance.
I am looking for standout dogs, with the fastest speed rating, the fastest average speed rating, from its last 3 starts. This is to target dogs that don't get into trouble, generally dogs that are at the front at the first corner.
In races with quality fields, chances are there will be no betting opportunities because of the closeness of their ratings. As I said, there is about 1 betting opportunity every 3 races. Some of these dogs are odds on, some of them pay $20.
I do not have a database of previous starts. All I have at my disposal, is track records, and form of the dogs for last 3 starts. That is all I want. An easy system, that returns >100%.
I do agree that track bias would play a part, however, you would need to keep too much information up to date to put this to use. All statistics from each track in the country would need to be recorded and maintained, which does not interest me in the slightest.

[Rock Steady]

BJ,
Are you talking POT of >100%? If you are and you genuinely can do that over a long period, then you would be the most successful punter I have ever heard of. Good luck to you. I am very happy with a POT of around 12%.

How on earth do people do form without watching race replays? Whether it be dogs, trots, thoroughbreds, tennis, NRL, AFL etc etc., if you don't see them how do assess their ability and style?

[BJ]

Rock Steady,
I am talking about Returns of greater than 100%(100 being break even). You are saying 112% for yourself, which I would be more than happy with also.

I previously have said that I am not yet backing it up with money, but so far have been satisfied with the purely statistical results. At this stage I am trying to work out how to automate it somewhat, so I can check daily results without any manual labour. Time constraints, and PC problems are delaying any further production.

As far as Video replays, I would like to be able to study that way, but am only looking to have an automated system (as automated as possible). Watching videos would possibly lead to smart punting, but not a system of any sort. But I do tend to agree with other sports, a greater knowledge is needed. The greyhounds, I tend to think in the majority of races, are flat out for most of the race. They are not hindered by human strategy....

I have also stated that this is what I believe, and cannot guarantee any results, and will not provide any stats regarding win/loss.

I can totally understand that you are satisfied with your method, and good luck to you. If you find something that works, stick with it. I am trying to find something that works, and this is how I am going about it....