Draw Bias – Any Profitable Angles?

My favourite time of the year for horse racing research is primarily the winter as it is when I do all my number crunching for the next flat season. I have always preferred flat racing over jumps racing, and as I bet far less on the horses in the winter it gives me that extra time I require.

This article is part of our Complete Draw Bias Guide — covering every UK course and distance in one place.

For this piece I am going back to my original stomping ground of the draw. In the 1990s before draw bias became well known to the masses, the opportunities to exploit it were huge. For me, that period coupled with first few years of the Noughties, provided me with numerous great opportunities and consistent profits. However, as with many things, when a value method becomes more widespread and better understood, the edge starts to gradually disappear.

This is very much an example of ‘shelf life’ – good ideas gain an edge because the majority of people do not use them.

As time passes however, the betting public and the more importantly bookmakers become aware and catch up. As a result, the prices start to contract, and the edge or value soon disappears.

In general, this is what has happened with draw bias over time. Also, such biases are less prevalent now than 25 years ago, due to far better watering systems and rail moving.

Back in the first lockdown I shared with readers two draw articles comparing the bottom half of the draw with the top half at all UK courses (5 furlongs to 1 mile 10+ runner handicaps). For this article I am solely trying to find ways to profit from the draw. Is it still possible? I am going to look at three options. So, let’s see ……

Individual Stall Positions / Groups of Stall Positions

If we travel back in time again to the 1990s certain stall numbers (draw positions) did make a consistent ‘blind’ profit especially over the shorter distances. It occurred regardless of whether the race was run round a bend or run on a straight track. These tended to be either the very lowest draws or the very highest draws.

In those days right handed tracks had their numbers reversed to what they are now, with the very highest draw being closest to the inside rail. Left handed courses were the same as they are now with the lowest drawn horses closest to the inside rail.

As we know draw biases can occur on round courses simply due to the fact that horses on the inside should have the advantage of taking the shortest route. In terms of straight courses at that time, very low or very high draws often had a clear edge too, as racing close to either rail was far more advantageous at some courses then, than compared to now.

Nowadays, as I mentioned in the introduction, draw biases are less strong in general, especially on straight courses. Hence backing individual stall positions totally ‘blind’ at individual tracks is unlikely to prove anywhere near as successful as it would have done 20 to 30 years ago. However, before drawing a line through backing individual stalls I have collated some data to look at.

The races I have looked at are 5 and 6 furlong handicaps with 10 or more runners over the past 5 seasons. My focus has been on the performance of the lowest drawn horses. I have chosen these runners because on all of the round courses they are drawn closest to the inside (regardless of whether the course is left or right-handed), and also because on around ¾ of the straight courses they are positioned close to one of the rails.

Before looking at the data, it is important to note that in 95% of cases the lowest drawn horse is drawn in stall 1, but if there is a non-runner from that stall, then draw 2 effectively becomes draw 1.

On the very rare occasions when both stalls 1 and 2 are non-runners, then stall 3 effectively becomes stall 1, and so on.

The following courses have made a profit with their lowest drawn runner during this 5 year period. (Based on £1 win bets):

Course	Runs	Wins	SR %	Profit / loss (BSP)	ROI % (BSP)
Nottingham	80	12	15.00	+ £178.80	+ 223.50
Southwell	156	18	11.54	+ £133.49	+ 85.57
Thirsk	129	13	10.08	+ £104.82	+ 81.25
Ascot	61	9	14.75	+ £67.80	+ 111.15
Chepstow	54	6	11.11	+ £18.10	+ 33.52
Chester	46	10	21.74	+ £16.62	+ 36.13
Windsor	95	12	12.63	+ £13.50	+ 14.21
Ripon	71	9	12.68	+ £9.94	+ 14.00
Chelmsford	171	19	11.11	+ £9.13	+ 5.34
Hamilton	64	7	10.94	+ £3.75	+ 5.86
Pontefract	62	8	12.9	+ £1.20	+ 1.94

What we need to remember is that relatively small course samples like this will often get skewed by a big priced winner or two. It is also worth sharing with readers that I delved back into the results from 2011 to 2015 to see if any of the 11 courses were profitable in the 5 years previous to this.

Four courses were – Chester, Hamilton, Ascot and Southwell. I think this clearly shows that this approach is far riskier than it would have been in the dim and distant past.

Even then profits were far from guaranteed.

Having said all that, I do have another interesting stat to share regarding the lowest drawn runner in these sprint handicaps. If we combine all such runners at all courses, they have actually made a profit in the last 5 seasons.

Stall Pos (All courses)	Runs	Wins	SR %	Profit / loss (BSP)	ROI % (BSP)
Lowest	3027	303	10.01	+ £188.11	+ 6.22

If you are keen to back an individual stall, then one could argue that you are better off backing that specific draw position at all courses. It is a lot of bets, for sure, but at least you are hedging your bets. 

Interestingly it is not just the lowest drawn runner at all courses that has made a profit in the last 5 seasons (in 5 & 6f; 10+ handicaps). The second lowest drawn horse has made a ‘blind’ profit too:

Stall Pos (All courses)	Runs	Wins	SR %	Profit / loss (BSP)	ROI % (BSP)
2nd Lowest	3027	312	10.31	+ £157.88	+ 5.22

Very similar figures to the lowest drawn stats in terms of strike rate and returns.

Now, I am not advocating this as an approach for the future, but I think it does show that there is potential in focusing on individual stalls or groups of individual stalls.

Personally, my approach would be to look at horses drawn in specific stalls which I perceived to be advantageous, and then combine this with other key stats (recent form, running style, fitness, etc). From there I could make a more informed judgement about whether to bet the horse or not.

Exotic Bets – Perming ‘well drawn’ horses in forecasts, etc

My biggest ever betting win came in a sprint handicap back in 2004. I combined forecasts and tricasts in that race landing a return of just over 20K. In fact, many of my most successful bets have occurred when using an ‘exotic’ bet – more especially forecasts and more often than not draw based ones.

Forecasts are bets when you are trying to predict the first and second; tricasts are bets when you trying to predict the first, second and third.

Clearly such bets are going to have a low strike rate and from a purely mathematical standpoint forecasts and tricasts are generally poor value. However, I used to perm the best drawn runners at the most biased tracks knowing that actually the odds were then going to move more in my favour. 

I tended to back the best two drawn runners in a reverse forecast, meaning that it was effectively two bets – Horse A to beat Horse B or Horse B to beat Horse A. The key was to do this even if one, or both of the runners, were big prices. From experience there was nothing worse than ignoring a race because two outsiders were the two best drawn horses.

I disregarded a race for that reason over 5f at Sandown back in the 90s when high draws had a huge edge when drawn closest to the far rail. The horses in question were 20-1 and 25-1, and yes, you’ve guessed it, they filled the first two places. The forecast paid around £500 for a £1 stake, which for a reverse forecast effectively amounted to around 250/1. Gutted does not even come close!  

Moving back to the 2016 to 2020 handicap data over 5 and 6 furlongs, I thought it would be interesting to see what would have happened if you had done a £1 reverse forecast on the two lowest drawn runners at all the courses over this 5-year time frame.

I was expecting a fairly significant loss, and although it did show a loss, it was a relatively small one.

So, if you had backed the lowest 2 stalls all course in 2 x £1 reverse forecasts your outlay would have been £6054 (3027 races x £2). Your returns would have been £5828 which equated to a loss of £226.

In percentage terms you would have lost just 3.7% (e.g., the ROI was -3.7%). In reality it was only a few wins away from making a profit.

Negative Draw Bias

The third idea I would like to briefly discuss is another draw angle that used to get decent coverage ‘back in the day’ and that is negative draw bias. I have mentioned the idea before and in truth it is about time I wrote a whole article or two on it.

Negative draw bias highlights a horse or horses that have run well from a poor draw and hence in theory have run much better than their finishing position may have initially indicated. From there you would have a horse that seems to be worth backing soon afterwards.

Now there are potential issues with this idea – two of them in fact. Firstly, you can never be completely 100 per cent sure whether the horse has run well against a draw bias, and secondly you have the tricky decision of how long do you go on backing the horse in the future? One run? Two? Until it wins? What if it loses 4 or 5 races?

Deciding upon the ‘correct’ decision is basically impossible and is all down to personal preference. 

Having said all that, this method if employed logically can pay off. Let me give you two slightly different examples.

The first one looks at using a ‘pre-determined’ bias such as the one you still find at Chester. In races of 5 to 7f at Chester high draws are at a huge disadvantage due to the tight configuration of the track. Any horse finishing close up from a wide draw in these circumstances has invariably run well against this draw bias. On 20^th August 2020 Roundhay Park finished 2^nd at Chester from stall 10 in a 6f handicap.

For the record the 1^st, 3^rd, 4^th and 5^th home were drawn 1, 5, 3 and 2 respectively. Roundhay Park finished 2^nd again next time out at Ripon before winning his following race in the Ayr Bronze Cup at BSP odds of 9.4.

My second example is in a race where there was no pre-determined bias, there just seemed to be a bias on that day in that particular contest.

There are so many examples I could have chosen, but this one from Ascot in 2017 (13^th May) from the 7f Victoria Cup is very clear cut.

In this 24 runner race, the first 11 runners’ home were drawn 18, 23, 11, 26, 20, 27, 24, 25, 22, 13 and 19. Higher draws dominated the finish so we can assume that lower drawn runners were at a significant disadvantage. This is about as close to 100% certain as we can be in terms of determining whether there was a bias or not.

The horse who finished 12^th Withernsea was drawn 2 and he was the horse to take out of the race, even though he finished halfway down the field. He was also 2½ lengths clear of the next best low drawn horse in 15^th. Withernsea returned to the track 7 days later at Newbury winning at the tasty BSP of 14.95.

Now of course not all horses that have seemingly run well against a draw bias go on to win soon after, but many do.

So, there you have it – three potential methods from which one could still profit from the draw and draw bias. The first two come with risks attached, but very simple to employ; the third comes more down to judgement and will take much more of your time.

I do believe all three still have scope to make a profit, but you will have to pick and choose any bets carefully.

David Renham  

Featured Image: http://Image Source: Wikimedia.com