'Thy' Kingdom Come?

The Peter Moody trained Thy resumed at Caulfield on Saturday 17 August, 2013 and was certainly not expected to figure in the finish of the race, with a starting price of $21.00 (top fluctuation $31.00). She ended up coming from well back over the unsuitable 1200m trip, to win running away by a length. Puissance De Lune also resumed at the same race meeting and with some major spring targets in mind, has dominated post-race analysis and media coverage. We add something 'more…' to the Puissance De Lune analysis, at the end of this article.

Let's have a look at the full card of race results available as part of a Sectional Pro Form subscription.


The above table contains 200m increment split and sectional time information, along with benchmark data for the winner of each race. To enable a straight comparison of the various races (regardless of class or distance) the default view shows 'All Ave Benchmarks' and we can see that the best 'overall' (Fin) performance was recorded in race 1 (15 lengths better than the worst, which was race 5), whereas the best L600m performance considering all race winners on the day, was recorded by Thy in race 7 (-ve 12.7 lengths).

We can also look at the last 5 race starts for Thy (see table below) and see that her L600m benchmark was the best she has recorded to date. Also noteworthy; this is the first time she has returned the fastest L200m in a race. It is true that a slow early pace has enabled her to save energy and assisted in the setting of a career best L600m benchmark etc, but given the unsuitable sprint race distance (1200m, as opposed to an ideal 1600m to 2400m), we believe her performance may be an indication she is in for an unusually good preparation.



'More….' on Puissance De Lune

Our own preference (when looking at races in isolation) is to look at race pace and performance in terms of what would be expected for the 'class' of race. This is one area where we can add something 'more' to the analysis which has already been published on the merits of the return by Puissance De Lune.  By looking at the full race results for race 6 using Sectional Pro form ('Cls Bmark Var L' is the default setting for this particular view), we can see that Puissance De Lune went out 1.3 lengths slow to the 600m point and got home the last 600m 0.8 lengths slower than would be expected for a Group 2 race (resulting in an overall benchmark of: slow by 2.1 lengths). Of course, we should keep in mind that this was also first-up over what might be considered an unsuitable 1400m trip, given the remaining spring 'staying' race targets for this galloper. Certainly, at least a 'pass mark', and probably indicative of a carefully planned and well executed preparation?

We will be continuing to follow Puissance De Lune's spring campaign from the perpective of his sectional times.



punterjj posted this comment 08/29/2013 12:10:39 PM

Hi S,

Thanks for the great article.

I am really interested in gaining a better understanding for how the sectional data should be interpreted to find more winners. I have a few questions...

Thy section

1. Is there anything in the sectional data leading up to this race to indicate why Thy's SP was $31?

2. As an aside who sets opening SPs?

3. Why would an ‘expert’ deem (incorrectly as it turned out) the 1200m trip unsuitable?

4. How is the Benchmark data derived?

5. What does the '-ve' mean?

6. You write “…that her last L600m benchmark was the best she has recorded to date.” Trying to interpret this on the screenshot is confusing. Her ‘L6’ (not ‘L600m’ as written in the article) ‘Sectional Time (sec)’ was actually her second best time in her last 5 races. On checking the ‘All Bmark Var L / Leader’ columns I can see that her ‘L6’ results for this race are ‘-12.7/-6.2’. Are you saying that the difference between these 2 numbers is the most of her last 5 races or ‘-12.7’ is more than the next largest negative number i.e. ‘-7.0’?

7. In the instance of ‘-12.7/-6.2’ how should a user interpret these numbers? If the difference in the numbers is what counts than wouldn’t it be better just to display the difference?

8. Why is “her last L600m benchmark” more important or significant than her actual faster sectional time?

Puissance section

9. Why is your preference (when looking at races in isolation) to look at race pace and performance in terms of what would be expected for the ‘class’ of race?

10. How have you derived the ‘Cls Bmark Var L’ numbers?

Thanks again,


ABettorEdge posted this comment 08/29/2013 18:29:00 PM

Hi Punterjj,

Thanks for the feedback and I'm glad you found the article of interest. Answers to your questions (by number) follow:

1. Thy's SP was $21 (top Fluc was $31), so someone thought the $31 was value and she was bet down to $21. The shortest trip she had been exposed to previously was 1400m and as such, there was insufficient past sectional time data to be 'definitive' about how she would perform at the 1200m first up 'sprint' trip. Although, as a staying type and drawn barrier 12, it was fair to expect that she would settle well back and have to unleash a 'special' last 600m to figure in the finish.

2. SP is determined by averaging the final price offered by selected bookmakers on-course. So you could say the weight of money determines the SP and it does generally reflect a runners chance of winning.

3. Refer also to Q1 answer. The 1200m sprint trip would have been regarded as being a conditioning run by most punters. The trainer had never started her over 1200m previously, indicating he also believed she was more of a 'stayer … 1600m+)

4. The All Ave Benchmark data is derived by comparing the individual times run against average times for the course and distance with adjustments made for; prevailing conditions (wet/dry/wind/turf length etc) and the quality or 'class' of runs which have fed into the average times used (sometimes referred to as par times). For the sake of simplicity, the difference between the adjusted run time and the par time is then converted to a measurement in lengths and 'normalised' for the race distance in question.

5. Following on from Q4, the -ve just means less than, or faster than, the par time. The smaller (or more -ve) the number, the better the performance.

6. I am referring to benchmark figures (not the raw time in seconds). -12.7 is the most negative benchmark number for the Last 600m in her history. also see answer to 7. below;

7. As indicated in the drop down description there are two numbers shown: the first number which is in 'bold' font is the benchmark performance for the individual horse (-12.7 in this instance) and the second number (after the '/') is the Leader or Race benchmark (-6.2 in this instance). It is useful to see them side by side as on the one hand you are looking at the performance or pace figure for the individual horse, in comparison to the race pace as a whole (which only looks at the leader when passing the various distance markers…i.e. different horses can be leading at different distance markers).

8. Looking just at raw time (in seconds) is meaningless, if you are trying to compare runs at different tracks, over different distances with varying track conditions (wet/dry etc). Benchmarking is generally designed to make comparisons possible.

9. People will often say that the race was run at a fast pace compared to the average for that distance at that track as an explanation as to why the horse may have faded at the back end of the race. This is OK if trying to make comparisons of races across various race classes. When looking at races in isolation (not comparing to other races on the day) it is more 'correct' to consider what pace might be expected for the class of race. e.g. If you compare the pace set in a top tier Group One race to the course and distance average it will appear to be very fast …. but… compared to other Group One races over the course and distance, the early pace might be just 'normal' and therefore not explain a late fade-out by the horse being referred to. 10.'Others' generally don't have the facility to look at 'Class' Benchmarks because the numbers of races to base the benchmark on are too small to be statistically significant. We have done modelling of race time versus class and come up with some secret 'herbs and spices' which enables us to determine a 'Class' Par Time estimation from the more general 'All Ave' Par Time.

Regards, Shane.