Latest news

Volatility at Lingfield

We’ve been having fun with analysing price histories at Lingfield before the off and in-play, to see how much variation in price there really is.

First, let’s take a look at the top 3 in the betting from the last race at Lingfield yesterday.

The bottom (x) axis shows the time in minutes up to the race beginning.  As you can see, not really much price variation 15 to 10 minutes before the off.  This is quite typical – things start to get interesting with 10 minutes to go.  Here you can see strong support for True Scarlet, and corresponding lack of support for Distant Goddess.  Bearing in mind all prices on Betfair will tend to 100% probability, the market confidence in True Scarlet equates almost exactly to lack of confidence in Distant Goddess, when up to that point the two were almost inseparable in the market.

As you can see, even with only one additional horse price added, Harmonious, the scale on the y axis is somewhat extended, as there is quite some variation in prices in a horse race.  If we were to add the rest of the field, including the rank outside at 1000, you would not see what is really going on at the head of the market.  Also, note that the small price movement in percentage terms, in Harmonious’ price, from 8.0 to 6.0 then back to 7.0, on a linear scale, is represented more than twice as  strongly in terms of movement than the percentage drift in Distant Goddess’s price, which is 2.5 to 3, and therefore similar in percentage terms.

Bearing in mind that during the inplay market, one horse ends up at 1.01 and the rest at 1000, we would see none of this detail.  So we’ll introduce another concept, relating to taking the log of a price briefly here, before looking at what happens to these three during the inplay market.

Taking the logarithm of a price enables us to compare percentage movements in price more easily, as well as fit more prices that are diverse on the graph, as we will see shortly.

First, here’s the above graph on a log scale.

It solves the percentage change problem.

Now, let’s look at the inplay situation for those horses at the head of the market, as log prices.

This time the x axis represents time from the start.   The distance of this race is 10 furlongs, and we can see that there is not much variation until about 90 seconds into the race, when things hot up.  In particular, in the last 20 seconds of the race, which is just after the final bend at Lingfield.  Looking at the top 3 in the market only, clearly punters see no reason to desert the favourite in running, whose price is steady and who is clearly going well.   But the second two in the market also attract a bit of interest before swinging dramatically up and down until of course drifting markedly as the favourite starts to dramatically shorten.    If we look at the log price volatility of all runners we can see this even more clearly.   Visually, in a race of 8 runners or more,  the horses always seem to land in a heap after the final bend, with a lot of uncertainty about who will come out best.

It’s a bit of a mess trying to represent the whole field in play, but it’s easy to see that each horse is more or less keeping its own price track (in other words not shortening or drifting to overtake the price of another horse) up until about 25 seconds before the end of the race – which equates to roughly the last 2 furlongs at Lingfield.  Whatever the distance, the horses always will always end up on the same stretch.   You can draw your own conclusions from the above if betting in play.

If you liked this post and would like to see more price analysis and graphs like these, please let us know in the comments above or on Twitter.