Speed Ratings – 7

A problem is...

Just a solution waiting to be found!

Yesterday, in Speed Ratings - 6, I mentioned that I was going to have a problem working out how many milliseconds a head or a length was going to be.

I asked for your comments and you responded.

Thank you for your contributions; they've helped me begin to resolve this issue.

The Problem...

1). We know the Average Seconds per Furlong for the different distances and goings.

2). We know the Average Seconds per Furlong for individual races.

3). We know that the Average Seconds per Furlong for the winning horse in an individual race; obviously, it's the same as the Average Seconds per Furlong for that individual race.

But what about the Average Seconds per Furlong of the horse that came second? Obviously, that horse ran slower than the winning horse. But how much slower is what we need to resolve.

The only thing that we know is the distance between the winning horse and the horse that came second (and, all other horses in the race.)

So, let's say that the horse that came second, came second by a length. How much is that length worth in terms of milliseconds? We need to know that in order to quantify how much slower the horse that came second was, when compared to the winning horse.

We could say that a length is 8 feet (or whatever) which is 8/660 of a furlong and then just use the Average Seconds per Furlong for the race to work out how many milliseconds that length is worth.

However, that would assume that the final furlong is run at the same speed as the rest of the race. That's not really good maths. That's not good enough for us.

The Solution...

In the end, it was Michael (amongst others) who helped show me the way.

He sent me a link to Turftrax and their sectional timings.... Lingfield - 12th February 2014 - 15:05

Just in case that link doesn't work for you, I've taken a screen shot of it.

Click on the image to see it full size - it should open in a new tab.

In the race, Keep Kicking came first and Brave Helios came second and was 2.75 lengths behind the winner.

To make it easier to see what happened in the race, I've created a chart for us to use...

Click on the image to see it full size - it should open in a new tab.

The chart shows Keep Kicking's race.

As you can see, he came out of the stalls fast and ran the first furlong in 12.18 seconds.

The next 10 furlongs were run at a slower pace until, 4 furlongs from the winning post, things start to speed up.

The race gets faster and faster until, quite reasonably, the final furlong is slightly slower than the preceding one.

Now, I don't particularly care if this is the typical profile of a 2 mile AW race at Lingfield but I imagine that it comes close.

However, all that I care about is the speed of the winning horse at the end of the race so that I can work out the relative speeds of all the other horses in the race.

Now, we could just use the Average Seconds per Furlong for the race.

But, Keep Kicking ran this race at an Average Seconds per Furlong of 13.13 (shown by the horizontal red line.)

And yet he ran the last furlong in 11.36 seconds.

He ran the final furlong in 86.52% of the time of the Average Seconds per Furlong (for the whole race.)

Now, all I have to mathematically identify is whether this is the norm for a 2 mile AW race.

In other words, come up with an average finishing speed for all 2 mile AW races, based upon the Average speed of the race (which I know for every AW race.)

And then rinse and repeat for all other lengths.

And maybe discriminate between the different surfaces and tracks.

Going Forward...

So, what I need to do is go through a lot of races and note the finishing speed for the winner of every race and put that into a spreadsheet which lists track, going, surface, race date, race time and average Seconds per Furlong for each race.

The spreadsheet is easy to generate.

Filling it in will take some time (even though I need only add one number to each line.)

I reckon that, if I set things up properly (spreadsheet, sorted in track/date/time order, on one screen, web browser on another) then I should be able to do 100 races per hour.

I'll want about 1,200 races to give me a decent data set.

Thus, 12 hours work.

Maybe another 4 hours to crunch the numbers.

An Alternative Method...

Of course, there's always another way of doing things.

If you lock three computer programmers in a room, feed them enough pizza and diet coke, then they'll come up with 15 ways of solving the problem.

That's part of the great thing about being a computer programmer!

That's part of the worst thing about being a computer programmer; you always know that there is an alternative (and possibly better) way of solving the problem.

What I can do is download videos of the relevant races; load them into a video editor, chop off all the race prior to and following the last furlong (for the winning horse) and the resulting length of the video is the number of seconds of the final furlong.

That'll give me the same numbers but will take a lot longer to do.

Conclusion...

Now, the good thing is that I can kick off the process of creating the Horse Speed Ratings and let that chug away in the background whilst I am working on this.

That will save some time.

I'll keep you updated on my progress; a collateral benefit is that we may get some standard race profiles for the AW races for different distances.

As far as yesterday was concerned, Jill and I didn't actually make it to the coast.

By the time we'd got our arses in gear, driven North, got hungry, had lunch and done a bit of shopping (well, quite a lot of shopping), it was too late to go for a paddle.

In any case, it was far too cold and wet for all that stuff!

Stuart was kind enough to pass on a joke, yesterday (credited to Tony Hawkes, I'm Sorry I Haven't A Clue, 22 December on BBC Radio 4)...

A penguin walked into a bar and said to the bartender "Has my brother been in today?"

"I don't know", said the barman, "What does he look like?"

And, on that note, I'll bid you a good Sunday and wish you a good day's racing.

As always, your thoughts and comments are very welcome.

My kindest regards