Home » Racing & Sport » MR System – Try Use It To Your Advantage

MR System – Try Use It To Your Advantage

It actually helps punters

The result of Saturday’s running of the Empress Club Stakes has again induced some ridicule aimed at the much-maligned Merit Rating (MR) system.

The winner of the race went off at an SP of 66/1 and paid R51.20 on the tote, implying that everyone thought the winner had less than a two percent chance of winning the race.

Jay August writes in the Sporting Post Mailbag that the winner was MR ranked 11th of the 13 horses in the race and horses so ranked generally win no more than 4.6% of all races they enter.

That equates to odds of around 20/1 and so the long odds available for the winner were therefore not surprising, and perhaps even provided compelling value.

Camphoratus gets up under Gareth Wright to beat Devin Habib and Running Brave (Pic – Chase Liebenberg Photography)

Is the MR system really that bad and is this result evidence of that?

If the MR system is indeed poor at rating (handicapping) horses, especially in fixed weight or weight-for-age races, then one should see that, on average, the MR rankings provide no help to punters when deciding on the winner.

The facts are quite the opposite though and show just how far from reality any negative perception of Merit Ratings is. As I’ll show the MR system has a good ability to predict results.

The table below shows all races run this season from August 1, 2018, and up to April 11, 2019.

Shown are the winning stats for all MR Ranked horses from MR rank 1-13, thirteen being the field size of Saturday’s race.

The very first row shows the stats for horses with no MR and which are accordingly not ranked, and the rows thereafter show the ranking of the horses ranked MR first to thirteenth.

It should be quite clear by looking at the table that the higher a horse is ranked by the handicapper in a race the higher its chance of winning that race.

Horses ranked 1 to 4 have almost double the chance of winning a race compared to horses ranked 5 and above. That is a significant statistic in favour of the efficacy of the MR ratings. Put another way, horses ranked 1 to 4 have a 37% higher chance of winning a race than average while horses ranked 5 and above have a 30% lower chance of winning a race than average.

These stats are even more significant because most races from which these numbers are derived are handicaps which seek to impede the ability of the better horses.

The MR system therefore helps punters a great deal in assessing the likely winner of a race but makes no prediction of price (odds).

The punter is responsible for discerning value and weighing the multitude of variables which effect a horse’s ability to run to its rating. So why would one blame the MR system for a result which happens just over 4 times in 100 races and is normal under conditions of uncertainty?

Saturday’s result is a rare event, and in no way indicates a weakness in the MR system. Some horses can up their game when so required. The winner was one such horse.

Those punters that criticise the MR system are failing to take advantage of a very useful and free tool in assessing form and price. Perhaps they simply cannot calculate uncertainty very well and are blinded by rare events!

Track conditions and the MR System:

Have Your Say

Comments Policy
The Sporting Post encourages allcomers to feel free to have their say in the spirit of enlightening the topic, the participants and the originator of the thread. However, if it is deemed to be either offensive, insulting, personal, false or possibly unsubstantiated, the Sporting Post shall, on it's own assessment, alter or remove comments.

16 comments on “MR System – Try Use It To Your Advantage

  1. Ian Jayes says:

    There is an old saying in horseracing: Put yourself in the best company and your horses in the worst. Every owner, trainer and punter should also be a handicapper.

  2. Steve Reid says:

    Jay you give me a satisfactory answer as to how the clowns at the NHA saw it fit to drop Oh Susanna MR after Saturdays run and you will have a convert for life.

    Unlike the NHA, you are not allowed to use opinion and/or fairy tales in your reply.

  3. Mac Naidoo says:

    Does the merit rating take into consideration of the goings. I think in Saturdays like conditions it rarely works. It’s my opinion

  4. Jay August says:

    Mac, no, the table is all races without any filtering as I wanted to show that on the whole the MR system has efficacy. But that is an interesting filter to apply and one I had not thought of. I will take a look tomorrow and see if it yields any variation to the norm. I doubt the data for this year will be of sufficient size to make a proper inference though, as wet conditions are far outweighed by good conditions. I may have to extend the test to two or three years data to make a valid comparison and that may take me some time.

    Steve, I’ll get back to you tomorrow on that after I have looked at their numbers for OS more diligently. I have my own view based on my ratings but will leave that for discussion tomorrow. I also have an opinion which I think is statistically relevant, but you may just throw that out as purely opinion. I’ll attempt it on you anyway and see what happens!

  5. Jay August says:

    Mac, I’ve asked the editor to append a table with all races affected by rain so that you can compare that to the table for all data. As I expected the data sample is small..

    The trend in data is much in line with those days where the going is good or standard. The top 4 rated on rain affected ground win 95% more races than those rated from 5 and upwards.

    I can see no major reason why rain affected going, on average, is invalidating the MR system. Obviously some horses are effected more than others but the overall trend does not support a notion that the MR system becomes unpredictable when rain affects the going.

    1. Editor says:

      It’s been added at the base of the editorial, Jay / Mac


    2. karel says:

      Looking at Oh Susanna’s ratings with us (AR) on good going vs soft, her soft ground ones are the worst since she appeared on the big-race scene.
      You could conclude that soft ground isn’t quite the surface for her.

  6. Jay August says:

    Steve, I’ve not forgotten about your “challenge” on Oh Susanna. I’m rather mindful of falling into a trap of making observations on one horse and then having that extrapolated as an indictment on the entire MR regime.

    Once I’ve mulled over my thoughts I shall respond to your question more fully!

    1. karel says:

      Jay, Steve
      It seems to me that the question is whether Oh Susanna’s original high MR was deserved.
      After she produced it with her Met win, beating Last Winter,(and got Horse of the Year title because of that) she hasn’t reproduced that rating, on good nor soft ground.
      Chances are it was too high.
      The handicappers need to take rectifying action in that case, and do that gradually.
      The joys of over-rating!

  7. Jay August says:

    Karel, Steve – yes that is very apparent (high rating), but I wonder if it is also a question of the horse having reached maturity faster than the WFA scale anticipated, and therefore not having made the anticipated WFA improvement thereafter. It is also debatable whether she runs efficiently at 1600m. I’ve always opposed her at that distance.

    Karel as you mention in your other post. WFA is purely an expected average for a population of horses and has general application to that population, but can vary quite substantially by individual horse. I do not think most observers understand that point well.

    I rated her Met win much lower than the 119 she was awarded under the MR system. Even factoring in the NHA WFA scale and their kilos-lengths metric I still could not place that win at higher than 108.

    I rated Marinaresco at 113 in that race and earlier this year MdK was quite critical of Marinaresco getting a 118 rating in Dubai based on his SA MR. In the event he has now slipped all the way down to 112 in Dubai so perhaps a Met rating of 108 for OS was more likely.

    The problem with using line horses the way the official handicapper does is that you overrate the best horses more severely than you would lesser quality animals in lower class races and the discovery of that error, and retracing, is traumatic for everyone. That trauma is I think what Steve is now alluding to.

    The ratings in the top races are most visible to those who observe MRs from afar and unfortunately this is where I see the most errors. This gives an impression that the whole MR system is invalid which is not true. Lower down the scale where most races are run, the mean error is almost non existent. The error is mostly observed when you stretch further away from the median rating and it impacts relatively few horses.

    The error is also relative as it impacts all horses at the outer extremity of the right tail. So for instance I believe that Lady In Black is also overrated – she should be much lower but relative to the overrated OS the error is not so obvious. I’m not sure that this explanation will appease Steve or even make sense. It is a difficult concept to explain in a single comment!.

    PS – if I look back over the handicapping review comments in 2012 it appears that someone other than the handicappers insisted on a single line horse and that line horse not being further back than x position. When exactly did this fixation with a single line horse within x places manifest itself?

    1. karel says:

      Your PS refers – this is from a reply to you a few weeks ago:

      In reply to Jay August.

      This is the best I can do (from a NHA press release dated 30 March 2010):
      “Following a meeting of the Handicapping Liaison Committee held on 20 October 2009, the National Board appointed a Sub-Committee comprising Messrs Robert Bloomberg, Tony Rivalland and Vee Moodley to draft amendments to the Handicapping Guidelines. These amended Guidelines were approved by the National Board on 24 March 2010.”

      There were subsequent updates, like this one which we reported on https://www.sportingpost.co.za/2016/04/nha-handicapping-policy-changes-some-questions-has-the-nha-lost-the-plot/
      I am not sure who drafted the 2016 amendments.

  8. Steve Reid says:

    Karel knows me too well it seems. The obvious problem is that Oh Susanna has not reproduced a single run to match the rating that she received after winning the Met. Overrating is indeed a massive problem here, because it takes months sometimes years for a horse to get back to a competitive mark. Who pays the price for this? You got it in one, everyone associated with the horse, while the handicappers tut tut and steadfastly defend their turf at the owners expense.

  9. Jay August says:

    Steve, the bigger damage is done with the male horses which are overrated. Last Winter is a good example as he has never run again, sent on a long journey as a result of his Met penalty, and from which he has yet to emerge! Would he have gone on that long journey had he got a more comfortable 113-114 rating?

    Oh Susanna with the black type of the Met and the ability to run and beat most fillies and mares over longer than 1600m was never going to struggle too much, despite the obvious bad run last week and her too-high rating. I’m less inclined to feel sorry for her connections as a result. But we will never have a Winx-like horse in SA if we continue to push our best over the edge.

    I wonder to what extent the handicappers are mindful of the fact that rating Grade 1 (and all graded races) races too low, raises potential problems with our status as a Tier 1 nation with the ICC? I’ve often wondered if the ratings I see handed out in Graded races are not in part an attempt to ensure that we remain within the ICC norms for such races.

    SA already has a huge problem in that it cannot meet the money conditions for most graded races and to drop the ratings scale could prove costly. I’m not at all an expert on anything breeding, but that is my from-a-distance perspective.

  10. Steve Reid says:

    Jay to precis your reply, she got dropped an arbitrary number after her latest disappointing run to continue the stabbing in the dark by the handicappers in attempting to correct her incorrect rating achieved after winning the Met. So 14 months later this horse still carries the penalty for achieving. This is the gist of my abhorrence of this system. It often penalises the achievers too harshly and takes too long to drop incorrectly handicapped horses. We have seen horses getting raised 20 plus points in a single run – what is the largest drop in MR on record?

    I continue banging my RF drum – much better system to horse and punter. Let those old school legal idiots linked to the NHA keep tut tutting and rather stick to divorce matters.

    Happy Easter to all.

  11. Jay August says:

    Steve. there are as many drops of 20+ as there are rises. It is just that one never sees these as they are not contentious. This season look at Ekuseni and Twin Peaks as two examples of that.

    We do not have MR35 handicaps where a lot of these drops would become obvious, but we do have lots of Graded races where rises are very visible. To be precise we see all the hindsight errors in the right tail but seldom see any in the left tail, because we simply are not interested in the poorer class horses.

    There is some science to keeping a population of horses handicapped and comparable from time to time, as all of Karel’s posts over time will attest. That science is not seen or understood by most and we therefore have endless debates around individual handicap marks.

    Ultimately the handicapper must bear some responsibility for this seemingly unhappy scenario. A little more published PR and deeper review of the MR system, quarterly or semi annually, by the handicapper, would go a long way to dismissing some of the obvious observational errors we all make when viewing individual horses in isolation.

  12. Steve Reid says:

    Single run adjustment Jay.

    This is my entire point.

Leave a Comment

‹ Previous

UK Racing Today

Next ›

Callan Murray (photo: JC Photos)

Second Singapore Winner For Callan

Popular Posts