In this article I will discuss another angle that can be deployed in our betting, and that is negative draw bias, writes Dave Renham. I think the phrase was coined in the late '90s by Russell Clarke when he used to write regularly in betting magazines. [He has since contributed an excellent eight part series on the betting markets here on geegeez, which can be read here].
What is negative draw bias?
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, one would potentially have a ‘horse to follow’ and worth backing soon afterwards when granted a more favourable position in the starting gates.
As with many things in racing, negative draw bias is not quite as simple as it sounds. There are potential issues with this idea – for example, once we have a ‘horse to follow’ we have the tricky decision of how long to continue supporting the horse in the future? One run? Two? Until it wins? What if it loses four or five races? There clearly is no ‘correct’ answer’ to this question.
We also need to think about under what circumstances we back the horse. Should we back it blindly? Or only under similar conditions? What if it is drawn poorly again the next few times it races?
A third question to consider is, "can we be completely sure the horse has actually run well against a draw bias?" If the horse has been beaten a neck over 5f at Chester from stall 14 then we can be as good as 100% certain. However, generally, races - especially big field affairs on a straight course - where one side of the draw seems to be strongly advantaged over the other. There is a case to say that biases that occur like this can be down to a pace bias (i.e. the fast horses were all congregated on one side of the track and therefore made that 'mini race' quicker) than a draw bias but, regardless of which, it is likely some form of ‘bias’ is in play.
Examples of negative draw bias
It's time to look at some examples of negative draw bias in action. I want to look first at a race at York over the 1 mile trip. This course and distance is in 7th position in my top 10 draw bias courses which I looked at in a previous article with low draws holding an edge over middle draws, and high draws at a big disadvantage. The race was run at the backend of 2021:
This race had a maximum field of 20 and, as can be seen, three of the five lowest draws filled the top three positions. Two middle draws in 9 and 12 came 4th and 5th and then the best of the high draws, Another Batt (drawn 20) and Ouzo (19) came 6th and 7th. This looks a solid example of Another Batt and Ouzo running well considering their negative draws. In fact, draws 19 and 20 are the worst of the lot being stuck ‘out in the car park’.
From a negative draw bias perspective, both Another Batt and Ouzo look to be horses to follow. So how did they fare after this good run? Well, next time out Another Batt went on to win at Donny:
He was joint favourite that next day, so clearly others noticed the good run at York from a poor draw. Even so, he won fairly comfortably and 9/2 are decent enough odds. Ouzo, meanwhile, has yet to run since but may be worth noting. He was bought for 62,000 guineas in the Newmarket Autumn sale and has moved to Jamie Osborne's stable.
Now, of course not all good runs from poor draws will produce next time out winners. So this goes back to the earlier question about what to do when you find one of these negative draw bias horses, and for how long do we support it, and under what conditions? I said earlier that there is no ‘correct’ answer. What we decide will simply be down to personal preference. From my perspective I tend to keep an eye on these horses for three or four more runs. That does not mean I will back them every time and, once they have won, I tend to cross them off my list. Why three or four runs?
Well, as mentioned, conditions in subsequent races will influence their chances. They may been drawn badly again; they may be in a highly competitive 20-runner race; the going may not ideal, and so on. Also, if they do not return to the track relatively quickly, as in the case of Ouzo, then that gives another potential cause for concern. So there are many factors that will make me think twice about backing the qualifying horse, even though sometimes I will miss a good winner by being more selective.
A system from the '90s
There is another reason I will keep the horses on my radar for a few subsequent runs and that is down to a system I used back in the 1990s. This system was based on negative draw bias and the optimum strategy for this particular approach was backing such runners on their next three starts, but stopping if/when the horse was a winner. It was very successful for a four or five year period, and it made me realise that these types of horses should not be immediately discarded if they ran poorly in the race following their negative draw bias run.
I mentioned at the beginning of this piece that big field races on straight courses can produce what seems to be a draw bias but may actually be a pace bias (which, I guess, is a sort of moveable draw bias). Ascot is one such course where this happens on a fairly regular basis. A good example can be seen in the Royal Hunt Cup of 2020:
High draws dominated this race as you can see in the result above. Maydanny, who finished 7th, was the only low drawn horse to finish in the first eight. Now normally your eye would not be drawn (excuse the pun) to a horse that had finished outside the top six. However, there clearly was a bias occurring here, and Maydanny was first home on the disadvantaged far side.
Maydanny did not follow the script next time when beaten into fifth as an odds on favourite. However, on his second subsequent run this happened:
From a plum draw (for a front-running type) in stall 1, he destroyed an 18-runner field at Goodwood, winning by five lengths at odds of 5/1.
Looking back to the Royal Hunt Cup, the in-running comments were insightful, too. Maydanny was the only horse to race on the far side out of the first eight finishers. Therefore, on a straight course especially, it is a good idea to look at the race comments in conjunction with the draw positions for the first few runners home in a race.
Here is another example, where I would argue the race comments are more clear-cut than the draw numbers. The first five home in the Britannia handicap at Royal Ascot in 2021 were as follows:
If we purely look at the numerical draw positions of the first five finishers, we can see that higher draws seem to have been favoured, but on first glance we may not think the draw bias was hugely significant. It may be a different matter if four of the first five home had been drawn 24 or higher and the other runner had been drawn 1; the numbers are shouting out as us in a case like that. However, if we read the ‘in running’ comments for this race we can see that fourth-placed Dubai Honour was the only one of the first five to race on the far side. The other four raced near side. This fact coupled with the draw positions make this look like a good run from a poor draw.
Dubai Honour was a horse that we could have added to our negative draw bias list and if we did, he would have rewarded our faith next time out, getting up to win by a head at 11/2. Indeed, he subsequently won a pair of Group 2's in France before running second in the Group 1 Champion Stakes back at Ascot on British Champions Day!
The next two home on the far side were Mithras (unraced afterwards in UK, renamed Turin Redsun and now racing in Hong Kong) and Qaader, who won at 8/1 two starts later.
Identifying negative draw bias horses (and a shortcut)
I have picked out three examples of negative draw bias but there are plenty more I could have shared with you. Not all will follow the winning script, but a reasonable proportion will win within three or four races.
Ultimately, to pick up on all potential negative draw bias qualifiers, we need to look at results on a daily basis and then keep a track of them, which can be done on Geegeez using the excellent Tracker tool. However, there is a possible shortcut for those of you who simply do not have the time to do that. It won’t likely be as accurate but it will be a quicker way to determine negative draw bias type selections.
What we can do is deploy a rule-based racing system. I discussed numerous racing systems in a recent set of articles so combining that approach with the draw provides some gratifying symmetry.
Here are the basic rules of the system:
- Last time out (LTO) race was a handicap with 10+ runners
- Horse must have been drawn 10 or higher LTO
- Horse must have finished 2nd, 3rd, or 4th LTO
This system is then to be used where the LTO course and distance was one of the following:
Now, for the eagle eyed reader, you may have noticed that my top 10 draw biased courses from 2016 to 2021 are in the list. In addition there are some of the 'near misses' I published with that top 10, as well as Dundalk over 5f. It is very difficult to win from stall 10 or higher at any of these course/distance combinations which is why I chose them.
I looked at results going right back to 2009 – essentially this was to get a bigger individual sample for each course and distance. Combining all of the qualifiers from all of those courses in their next starts we get these bottom line figures:
Considering this is a very raw type of system these combined results are impressive. It should be noted that I chose the course and distances before I checked the results so there is no back fitting here. Indeed, five of the 14 made a loss, so I could easily have manipulated the stats by ignoring those courses to improve matters – but that is not my style.
For the record, those that made a loss were horses that ran last time out at Chester 7f, Kempton 6f, Goodwood 7f, Goodwood 1m and Pontefract 1m. The other nine combinations were all profitable.
I then thought it would be a good idea to compare the strike rates of the negative draw bias system horses with ALL horses that finished 2nd, 3rd, or 4th last time out in 10+ runner handicaps (2009-2021); not just the win strike rate, but the placed strike rate as well (placed SR% being win and placed runners combined). Here is the comparison:
A better absolute strike rate of nearly 2% in terms of wins, which is almost 14% better comparatively; while the placed results show a similar pattern:
Over 3% absolute difference in the placed strike rates, and an 8% comparative improvement. It's satisfying to see increased strike rates in both groups, which adds confidence to the basic system concept and the results thereof.
This system approach should not be time consuming. There will be far fewer races to check over the course of a year compared with worrying about checking the results of all 10+ runner handicaps. Indeed you will only need to check the day’s results when one of these track and trip combo's has rnu a handicap with ten or more runners. Also the system is only concerned with the very next run which means once a horse has run again you simply strike it off the list.
Of course this method is easily adapted: for example, you may want to change last time out position of 2nd to 4th to a distance beaten figure (in lengths, or lengths per furlong - perhaps using the Geegeez Px coloured dots on the left side of the Full Form result rows); you may want to change draw 10 to draw 8; you may want to keep qualifying horses for more than one run, and so on. Ultimately, there is lots of scope to change the approach to suit your style.
Keep in mind (of course) that, as we know, a system is simply that – it is not a magic bullet and just because 2009 to 2021 produced a profit, it doesn’t mean results will continue to be positive in the future, or that there won't be losing runs. This system, however, does follow logical negative draw bias ideas so one would hope it has a sporting chance of repeating its past success in the near future at least.
I hope this article has sparked your interest in negative draw bias and please share your thoughts or personal experiences in the comments below. I'd love to hear from you.
p.s. The recent Victoria Cup, again at Ascot, saw the highest three (out of 27!) stalls combine for a £5208.10 trifecta dividend - keep an eye out for draw bias angles, both positive and negative!