Last week Geegeez added Betfair Starting Price (BSP) to numerous areas of the site, writes Dave Renham. For me as a researcher and writer this is fantastic news. As we know, most punters do not bet at Industry SP (ISP) anymore. Some still stick solely with traditional bookmakers, but to improve their bottom line they will use Best Odds Guaranteed (BOG) where available, as well as early or ante post prices. Some will use the Exchanges, primarily Betfair, with BSP one option to be utilised. Others will try and exploit both the bookmakers and the Exchanges to hopefully gain maximum advantage.
In my personal betting I use BSP for around 40% of all my horse racing win bets, so when researching ideas it is very useful for me to see the BSP profit and loss column.
For this article I am going to examine data from UK racing over the last two full years (2023 and 2024). In the overall findings I will be including all race codes, i.e. flat, all-weather (AW) and National Hunt (NH). For BSP profits/losses I will be using 2% commission which is what we, at Geegeez, are using in our calculations.
When we compare ISP to BSP there is no contest – BSP wins hands down. To give an example, if we look at horses priced between 5/1 (6.0) and 6/1 (7.0) combining all race codes in the UK over the designated time frame we see the following:
To BSP a profit of £226.24 to £1 level stakes would have been achieved compared with an £1826 loss if backing to Industry SP. That is some eye-watering difference. Just imagine if we were using £20 stakes and not £1 ones!
Before delving into BSP in more detail I do want to talk very quickly about Best Odds Guaranteed (BOG). This option is still available with 12 main bookmakers on most UK races each day. Essentially this option is a no brainer for those betting with standard bookies. When using BOG, it gives punters the chance to take an early price, but if the starting price (SP) is higher, we get paid out at the higher odds. I am in the process of doing some initial research into potential BOG strategies and at this early stage it seems there is a sweet spot in terms of price – or at least the early price. Early prices around the 5/1 (6.0) to 7/1 (8.0) mark seem to offer the best value long term for BOG bettors. I will need to dig much deeper, but I am fairly confident I am in the right early price ballpark to utilise BOG to its max.
There can be issues though with BOG betting such as limits on stakes and occasionally the BOG option will not be available – normally for those people that are winning consistently using it. Working out potential BOG profit and loss figures based on past prices is not always clearcut because of these aforementioned issues. However, I do hope to be sharing some research on this at some point in the future.
Back to main focus of this piece. Earlier I mentioned that the calculations in terms of Betfair commission across the Geegeez site is 2%. For those who currently 5% and are regular bettors on the machine, then log in to your Betfair account and choose the 'Basic' plan on this page. Once this is done, you'll pay 2% only on net winning Exchange bets.
Paying 2% commission on winning bets rather than 5% commission is clearly preferably but I want to illustrate this numerically by using real data to see what a difference it can make long term. Let me compare the BSP profits (to £1 level stakes) of all horses that had an Industry Starting Price of 13/2 in terms of 2% commission versus 5% commission.
Over this two-year period the difference would have been £116.89. To £20 stakes the difference would be a very significant £2337.80. In the table below I will share some other ISPs in terms of this 2% v 5% difference:
A palpable difference across the board and, for 8/1 shots, as with the 13/2 shots, a loss with 5% commission has been turned into a healthy profit when applying a 2% commission.
My next piece of digging is in connection with the ISP and the average BSP price for that specific price – comparing the difference between the two. The first graph compares a selection of ISPs under 10/1 (11.0) with their BSP average counterparts. The graph uses decimal odds for ease of comparison:
Hence an Even money shot at ISP (2.0) has paid 2.14 on average at BSP (before commission); a 9/1 (10.0) shot has averaged at 12.61. It is just another indication of why ISP on its own is outdated for any serious punter.
Let me now look at a selection of some bigger ISP prices ranging from 11.0 (10/1) to 41.0 (40/1):
As the ISP prices goes to 20/1 (21.0) or bigger the gap to BSP starts to increase considerably. Once we get to 40/1 (41.0) the BSP average is moving closer to double that of ISP. I have always been a fan of backing big-priced outsiders because if I can find a horse with a percentage win chance akin to its likely ISP then I have excellent value.
My next comparison is with average BSP prices for handicaps versus non-handicaps at various ISPs. I wonder how many of us have assumed the average prices would be basically identical – well, within a hundredth of a point or two over two years’ worth of day at least. This is indeed the case for an ISP of Evens (2.0) where the difference is 0.02 of a point (2.13 versus 2.15), but as the prices get bigger, the gaps between the two start to increase. Once again, I’ll share two graphs, the first focusing on an ISP of 9/1 (10.0) or less:
The average non-handicap BSP is higher across the board than the handicap one with the difference between the two gradually increasing as the prices get higher.
Now I would like to examine the bigger prices:
With the bigger prices we see a similar pattern with the non-handicap BSP averages higher than the handicap ones and the gaps between the two once again increase as the ISPs get bigger. Looking at the ISP 40/1 (41.0) comparison we can see the gap between the two prices is close to 10 points (77.34 versus 67.87).
I believe the reason we have these differences, and such differences are more pertinent to these bigger prices, is due to the shape of some non-handicap markets. I am talking primarily about non-handicap markets with a very short-priced favourite. Here is an example of such a race. It was the 6.30 at Southwell on 8th October 2024. It was a 5 runner 2yo novice race (non-handicap). Here is the result with the relevant ISPs and BSPs:
With a very short odds favourite in Shah at 2/13 (1.15) if we look at all the other BSP prices they are bigger than their non-handicap average price. The table below helps to illustrate this further:
Not only was the BSP comfortably above the average for all four of these runners, in the case of the two biggest priced runners, Something Splendid and Divot, the difference was huge (80 v 47.36 and 328 v 142.53). Of course, such huge outsiders in a race with a super-hot jolly win very rarely but when they do the BSP rewards handsomely.
In my two-year research time frame, there have been 64 races with a favourite priced 2/13 or shorter and 61 of those were non-handicap races. Therefore, having this type of market shape for handicaps is extremely uncommon. Hence, these higher priced outliers in terms of BSP will occur much more in non-handicaps, helping to push the average BSP upwards. Now my guess is that this is not the only reason for the big differences between the average BSP prices of bigger priced runners in non-handicaps versus handicaps, but more on that later.
Continuing the bigger priced theme as well as comparing handicap results to non-handicap ones, let me look at some more BSP data comparing strike rates and BSP returns. In the table below I have split ISPs into three groups – prices from 33/1 to 50/1 (34.0 to 51.0), 66/1 to 80/1 (67.0 to 81.0) and 100/1+ (101.0+).
When looking at horses priced 34.0 to 51.01 the win strike rates imply a small edge to non-handicappers and the returns show a clear advantage to that cohort, too. Once we get to 67.0 to 80.0 though, the strike rates have flip-flopped with handicappers winning nearly twice as often (albeit still very rarely) and with a huge disparity in the ROIs of around 35p in the £ in favour of said handicappers. This disparity just gets bigger once we hit those 101.0 or bigger shots. Although these 101.0+ handicappers have won on average just one race in every 175 they have seen a return of over 60p in the £ to BSP. Non-handicappers in this price bracket have won on average one race in 833 losing 45p in the £.
There are two reasons for sharing this handicap / non-handicap BSP data for bigger price runners, and I would like to clarify that it is not to suggest that we back all 100/1+ handicappers! The first reason is to show that with bigger priced runners the type of race does make a difference, as does the ISP or the likely ISP, in terms of win chance, likely BSP and potential returns. Secondly, this table might help to explain an additional reason for something I was discussing earlier in relation to why the average price of outsiders on Betfair is bigger for non-handicappers than for handicappers.
At this juncture it should be noted that BSP does not beat ISP 100% of the time. However, a BSP ‘win’ does occur 97.5% of the time (and therefore ISP has ‘won’ 2.5% of the time). It is this 2.5% subset of runners I want to look at next.
Given that we know the Betfair market is about as efficient as a betting market can get, when the ISP is higher than its Betfair equivalent, the expectation would be that this industry price ought to be very close to its ‘true’ price.
4486 horses had an ISP higher than their BSP during the two years in review, and if we had backed them at ISP, a profit of £241.69 (ROI +5.4%) would have been achieved. To BSP these runners would have lost us £77.63 (ROI -1.7%).
Of course, we don’t know the BSP or the ISP before the race starts, so you’d be forgiven for thinking this is a pointless piece of intel. However, for those punters who back late on Betfair, literally seconds before the ‘off’, knowing about this unusual state of affairs could offer a potential strategy.
The prices available very late on Betfair are going to be close to the eventual BSP, especially at the front end of the market. Technically, then, a strategy that may offer an edge would be to have both a Betfair live screen along with a couple of bookmaker live screens open on your computer, coupled with a live racing feed. If, a few seconds before the start of the race, the live Betfair price on a horse is lower than an available live bookmaker price, then back the horse with the bookmaker.
The chances are, regardless of the final ISP, that this will beat BSP with the price taken, or at least effectively beaten it after commission is considered. Not all of us have the time to watch live races on a daily basis and employ such a strategy but for those who do, I would be interested to see how this idea panned out over time.
Finally in this article, I want to examine the results when we use a BSP to ISP odds ratio. What I mean by that is if a horse has a BSP of 3.0 and the ISP was 3.0 the ratio would 1.0. If the BSP was 9.8 and the ISP was 7.0 the ratio would be 1.4 (9.8/7.0). I wanted to see if we could find anything useful out of looking at such ratios. To do this I have used ranges for the ratio and the table below shows my findings:
As expected, the strike rates tend to move in a positive direction as we move down the groups. In terms of returns, horses that have a BSP/ISP ratio of 1.01 to 1.24 have offered the best value. This again helps to illustrate how efficient the Betfair market is, especially at the front end of the market.
*
That’s all for this week. Any price-based research has flaws because as I have stated earlier, we do not know pre-race what the ISP or BSP will be. However, this type of overview analysis is important to understand. For those who never or rarely bet on Betfair I hope this article is enlightening. For those who do, then there should be plenty of new information and stats to be aware of which have the potential to improve one’s bottom line.
- DR
An excellent article as usual,with some thought provoking stuff. I just wondered though, as someone who sometimes bets each way whether the value lies on the exchanges or with the bookies. I’ve never seen an article that covered this aspect as it isn’t easy to research exchange place odds. Any views?
Yes it’s something I’ve never looked at and can’t remember reading anything on the subject either – I’ll have a look to see how easy it is to crunch some numbers.
Dave
Leave a Reply
Want to join the discussion?Feel free to contribute!