A Guide to Betting on Greyhounds


Please Gamble Responsibly


The Facts

  • 15 per cent of Australian adults (18 years+) gamble regularly (excluding those that only play the Lotto or “scratchies”). 
  • Roughly one in ten of those adults would be classified as ‘problem gamblers’, with an additional 15 per cent experiencing ‘moderate risks’. 
  • Estimates of the number of problem gamblers in Australia are around 125,000. Another 290,000 Australians are estimated to be at moderate-risk of developing a gambling problem. 
  • Australians lose more than $18 billion dollars in gambling a year. $12 billion is lost on poker machines each year. Governments took more than $ 4 billion dollars in tax from gambling revenue in 2007. 
  • Tax revenue from poker machines accounts for 13.1 per cent of Victoria’s Governments Tax Revenue. In NSW it is, 9.4 per cent and 9.6 per cent in Queensland. 

You've decided to have a night at the Doggies and want to enjoy a little flutter on the next race. The only problem is you want to bet small, have no idea how to place a bet, what all these numbers mean, and have heard of but never understood what a trifecta is. Oh, and you want to know how to select a winner! READ ON

  • If a greyhound is priced at, for example, $4.40 and you make a $1 bet for a win, it means that if the greyhound wins you will get back $4.40 (profit $3.40)
  • Simply tell the TAB operator “$1 on number 8 for a win in the next race Wentworth Park”
  • A place bet means if the greyhound you select comes 1st, 2nd or 3rd in the race you will win*
  • The odds for a place bet are usually much smaller than a win because your chances of winning are much higher
  • Simply tell the TAB operator “$1 on number 4 for a place in the next race at Wentworth Park”
  • If you want to make a bet for a Win AND a Place (as a type of insurance), simply tell the operator “$1 each way on number 7 in the next race”. NOTE that this will cost you $2 because it is essentially two bets.
  • boxed trifecta is when you pick the first three greyhounds over the line in any order. Simply tell the operator “$1 boxed trifecta on numbers X, X and X in the next race”. This will cost you $6 as there are 6 combinations for the three greyhounds you have chosen to cross the line.  You can add more greyhounds to the mix to improve your chances.
  • First Four bet gives you the chance to pick the first four greyhounds across the line in correct order. This bet usually pays between $200 and up to several thousand $$. You can pick between 4 or 8 greyhounds in your bet. "Boxing" four greyhounds that you think are the best 4 in the race will cost you $24 (4x3x2x1=$24). Boxing 5 would cost you $120 (5x4x3x2=$120)
Other Things To Note
  • Placing a $1 or $2 bet is fine. Don't feel embarrassed!  
  • Betting prices change and are not ‘locked in’ until the race commences. 
  • You can select 'fixed price odds' which guarantees your price when you make your bet (note that all on-course Bookmakers provide fixed price odds). Simply tell the TAB operator "fixed odds" when placing the bet. 
  • Select a winner from the racebook (sold at the front gate) or download the app thedogs which provides you with tips, comments, form guide, estimated prices, race replays and speed maps 

Selecting a Winner: Quick Guide

Lets start with a Speed Map (found in your racebook or on the smartphone app "thedogs"). This will tell us roughly where we think the greyhounds will be after ~100m. The start is one of the most important factors to assess because a good start can mean a greyhound will avoid a squeeze. Here is a snap shot of a Speed Map, overlaid with some comments by ourselves: 

Number 1 (the red rug) appears to have plenty of room here as numbers 2 and 3 appear to be slower starters based on recent race history. The number 1 box also produces the most winners (see below "stats"), so we just want to then look at recent race times at Wentworth Park over 520m compared to the other greyhounds. If we see 30.10 as a recent time compared with say 29.90 by the greyhound in box 5, we would probably still stick with #1 even though 1/10th of a second is equivalent to about 2 lengths (i.e. the length of a greyhound). Greyhound runner #8 also appears a fast starter, so check how he matches up against number 1. #6 also has somewhat a clear run, remembering the greyhounds will gravitate upwards of screen toward the rail/bunny. The race review (slide the screen across to the left on the app) will usually mention if any of these greyhounds are what we call a "widey" (likes to race wide and avoid the other greyhounds) or a railer (hugs the rail, which is of course the quickest way to the finish post). 

Racebooks will also report "1st Sectional Times" which is the input for the above Speed Map and is usually around 5 seconds. Study these closely as even a fraction difference can determine a race. If a dog ran a poor 1st sectional last start, check from which box it ran from. If it is box 4, 5 or 6 then you could forgive it! Look for best times and best sectionals, but discount anything more than 4-5 months old. 

Check how many races the greyhound has won and its placings. You will see numbers like 45-7-4-3 which means the greyhound has run 45 times of which it has won 7 times, come second 4 times and third 3 times. 

Other information is important such as the record of the Trainer and first-time starters at Wentworth Park, but overall the variables and inputs that influence a race start to shrink. Unlike the horses there is no need to check track conditions, weather, the jockey, whether the rail is out or if its a full moon etc. Minor considerations might include whether your selection is a bitch (female) and if two dogs (male) are placed either side of it at the starting boxes. Dogs are bigger than bitches (about 5kg) and can therefore crowd her style at the start if they are level starters. If the race book says "moody type" and the greyhound is in a late race and is boxed in an odd number box, discount slightly as odds numbers are loaded into the boxes first and then evens thereafter. 

There is a lot of science to picking a winner, and this makes the analytical exercise enjoyable for many, but the time spent is probably 25% of what you need to consider relative to assessing a horse race. 


Converting the betting odds into implied probabilities of winning may prove a more intuitive tool to use than viewing betting odds in isolation, especially for the novice.

To convert odds into implied probabilities, simply divide 1 by the decimal odds.

For example, a greyhound with a market price of $3 implies that the market believes its probability of winning is roughly 33% (1/$3.00 = 0.33). We say roughly because the sum of all probabilities must equate to 100%. However, because bookmakers and the TAB clip the odds, the market odds available are actually less than what would be available without an intermediary, thus elevating the sum of the field's probabilities above 100%. For example, if you and a friend bet on a coin toss, and the odds are $2 for heads and $2 for tails, then the winner will win $10 on a $10 bet. If the TAB were acting as intermediary, then the odds would likely be $1.95 for heads and $1.95 for tails. The winner would record a net win of $9.50 and the TAB would keep 50c (the two players threw $10 x 2 into the win pool, and the eventual winner received $10 x $1.95 = $19.50, a net win of $9.50). The $1.95 odds imply the probability of winning the coin toss is 51.28% which of course we know is not true (it is 50% on any coin toss). 

To overcome this, you can build yourself a "Wheel of Fortune" (WOoF) in excel like the one below. You'll note that the probabilities of each greyhound adds up to a figure greater than 100%. However, the wheel of fortune adjusts the percentages of each runner to accurately illustrate the probability of winning according to the market. Using the wheel, you can try and imagine which number you might back based on risk and return. Or alternatively, the next time you see a greyhound runner with odds of $10, the wheel can help you visualise your chances of winning (again, if the market is correct!). You can download the WOoF excel sheet here 

Box Number Odds Offered to win Odds translated into Probability of Winning
1 $10.00 10%
2 $5.00 20%
3 $2.10 48%
4 $8.00 13%
5 $12.00 8%
6 $20.00 5%
7 $26.00 4%
8 $16.00 6%



Don't forget that the market can be wrong, so whilst the betting odds might suggest a 59% chance of winning (odds of $1.70), punters might have artificially increased its probability of winning by following the market as opposed to their own assessment.  Fortunately the biggest players in greyhound betting markets do analyse the form independently, and novices are usually small time punters that seldom influence the market (as opposed to the stock exchange. Remember the dot-com boom anyone?)  

As another alternative, our weekly racebook tipster assigns estimated odds for each greyhound runner in each race. Whilst these odds are not translated into probabilities, you can simply do this translation in the spreadsheet provided above and write it in the racebook. 


The Edge and Outlay Risk Management (Getting Serious Now!)

Betting on the greyhounds should be a two stage process; Determining the probability of a greyhound winning a race and then determining the optimal outlay, if any at all. We discuss the latter which is seldom considered by punters 

A punter’s expected return from gambling should be zero or less.

For example, the expected return from betting heads at even money on 100,000 coin flips is zero. Without an 'edge’, the punter cannot raise the expected return above zero. The edge constitutes a greater amount of information that other betting participants have overlooked (insider information is also an ‘edge’ in stock markets and gambling, but obviously this is illegal). In many cases, such as the flip of a coin, the edge does not exist deeming any betting participation as futile. 

In greyhound racing, an edge might be gained from analysing the form guide at length (handicapping) or even employing a professional tipping service. On course, Wentworth Park subscribes to the tipping services of the curiously named Giddy-Up which is designed largely to provide theoretical prices for each runner (rather than clear cut race winners) and speed maps. If punters can secure better odds than their model predicts, they may have an edge. The service is located on screens in the bookmakers' betting ring.

If a punter believes a greyhound is paying overs (a price greater than their estimated price the greyhound should be running at), the punter should then calculate what percentage of their bankroll should be gambled on the runner. This is largely overlooked by most racetrack followers. It is what we refer to as the Risk Management process or Stage 2 of the betting process.

A system developed by John Kelly over 50 years ago is possibly the most popular calculation used by professional gamblers (and investment funds for that matter) to determine the optimal outlay based on the estimated probability of winning and prices available. Whilst the original Kelly paper detailing how the formula was derived will be too difficult for most readers to understand, the end formula is quite simply: 


p = probability of winning 

q = 1-p

d = decimal odds available (eg $4.50) 

We provide an example excel spreadsheet below and recommend readers experiment with inputs to get a feel for the model.

The main caveat with the system, and any modelling system, is that the data input (the edge) needs to be reasonably accurate. Garbage in, Garbage Out! 

We believe the Giddy Up service provides adequate speed maps and pricing for each runner that punters can use as a starting point to help gain an edge. It perhaps works best when Giddy Up's second and third favourites are mispriced by the market because too much money is directed at the favourite.

An example of the Giddy Up screen is shown below. In this example, the fixed odds available for Black Magic Opal are slightly above the Giddy-Up estimate. The Kelly model would suggest ~13% of the bankroll be placed on Black Magic Opal at $1.90. 

For more optimal results, the system would likely work better if two parties manage it, one providing the probabilities of winning and the other (the Risk Manager) entering the data inputs and executing the bet. This is because the formula can produce outlay amounts that may not be intuitive, especially should your bankroll grow and the outlays for each bet become much bigger. The model will also produce wild swings in balances that can throw a punter's confidence in the system, setting back punters or producing significant short term growth that instils an unrealistic level of confidence in the system which subsequently distorts future inputs.

It will also produce many “no bet” recommendations (i.e. don’t bet on the race as there is no edge) when the betting odds on offer do not exceed the minimum required. When this occurs, punters can lack the discipline required and either ignore the recommendation or revise their probability inputs so that a bet can be made.

Alternatively, punters can lay the favourite using Ladbrokes. In this instance, if you estimate the probability of a greyhound winning is 35%, then your probability of it LOSING is automatically 65%. If the odds to lay the favorite (Favourite vrs Field in Ladbrokes) is greater than $1.54 (1/probability), you determine how much to outlay using the Kelly Formula and hit it! 

An example excel spreadsheet model of the Kelly Formula can be downloaded here.

In summary:

  • Nominate the amount you wish to start with (the bankroll) which represents the cash you are bringing to the track or punting at home with and set yourself a target return over a nominated period (anything from 1 day, 1 month, 1 year)
  • Analyse the form and determine the probability of a greyhound winning the race (not the price). For example, 27%
  • Input 27% into the spreadsheet, or communicate this information to the Risk Manager to input
  • Input the odds available into the spreadsheet (we usually stick to the fixed prices that are available)
  • The Risk Manager (or yourself) will place the bet based on the model. If 'No Bet' appears, then do not bet
  • Enter W or L after the race for Win or Loss. This will carry forward the balance to the next line
  • In many instances, the spreadsheet will produce a recommended bet that is not rounded (eg $7.52 to bet). Either round the formula in excel or adjust the probability slightly using "goal seek" function in excel (eg set cell to $7.50 by adjusting the probability cell). 

Disclaimer: Use at your own risk. The information above is of a very general nature and is provided as a guide only. The usual disclaimers regarding no acceptance or responsibility for accuracy apply, including the use of the sample excel spreadsheet. The John Kelly Formula is merely a tool amongst many that may be of use to some gamblers. Giddy-Up also provides no guarantee that their service will generate any return in isolation. Giddy-Up is a service we subscribe to and we have no affiliation with beyond that subscription. Whilst most people enjoy gambling for entertainment only and apply the “zero expected return” ethos, it is a problem for some. See here for more information regarding problem gambling.


Wentworth Park Statistics (based on 3650 races between June 2011 - August 2014)
  • Of the eight boxes allocated for each greyhound race*, Box One (red) generated the most winners with 18.5%. This is traditionally the case at nearly all greyhound racetracks simply because the box is the closest to the rail and therefore the quickest route to the winning post. The greyhound contends with only the box 2 runner at the start, thus mitigating a ‘squeeze’ at the beginning of the race.
  • 38% of all Wentworth Park TAB favourites over the past 3 years won their races (3.8 favourites win per race meeting) paying $2.39 on average for the win and $1.40 the place. This is greater than Randwick/Rosehill thoroughbred races that recorded 33.8% (2.7 per meeting) over the past 3 years
  • The average TAB win dividend was $5.94 and the average place dividend for these winners in isolation was $1.97
  • The average trifecta payout at Wentworth Park was $218 and for the First 4 was $1082. The highest payout was $3161 and $16,571 respectively. The largest dividend payout in the exotics was a Big 6 which paid $122,000 (average $7,700)
  • Box 7 is the least successful box at Wentworth Park, winning 9.2% of all races and falling to 8.7% if we observe 520m events only. In a perfect world it should be 12.5% (1/8). 
  • Box 2 provided the best return on average over the three year period to Aug 2014, Box 7 the lowest average return for race wins
  • Of all 3650 winners, 14.8% paid even money ($2),  23.9% between $5-$10, 10% between $10 - $20 and 3.9% of $20 or more. 
  • 54% of all winners from box 1 were the favorities for the event and 75% were either the favorite or second favorite. The average TAB dividend for these winners was $2.20. 
  • Whilst 54% of all Box 1 winners were favorites, boxes 5 and 7 were only ~28% (eg of Box 5’s 362 winners, 100 were favourites). 48% of Box 5 winners were neither 1st or 2nd favourites. Despite producing only 10% of all winners of 3650 races, this may suggest punters are penalising /discounting these boxes too much.  

* due to late scratchings and other reasons, some races may not have 8 runners. If less than 8 runners, a place bet pays on only 1st and 2nd placings.