Introduction
Impact Value (IV) shows how strongly a horse’s price came in during the race, compared with what is normal for a horse at that starting price.
| Score | Meaning |
|---|---|
| 0 | The price never went below the Betfair starting price (BSP). |
| 50 | The price came in to about the typical in-play low for that BSP. |
| 100 | The price traded at 1.01. |
Higher means more in-play market support. 50 is not an average horse. It means about what we usually see for a horse at this price.
Why compare with a typical move?
Winners usually trade at or near 1.01 at some point in-running, and well-fancied horses often trade much shorter than their BSP even when they lose. So simply asking whether the price shortened is not very informative: most fancied runners do.
Impact Value instead asks whether the price came in more or less than is normal for a horse at that price. That makes a 2.0 favourite and a 20.0 outsider comparable on the same scale.
How it is calculated
The typical in-play low
From a large set of historical races, we have fitted a curve that predicts the typical in-play low for any BSP.
For example, a horse with a BSP of 5.0 (a 20% chance) typically trades down to about 2.15 at some point during the race.
The score
The horse’s actual in-play low, the shortest price matched on Betfair during the race, is placed on a scale with three fixed points:
| In-play low | Score |
|---|---|
| Equal to the BSP (no shortening) | 0 |
| Equal to the typical low for that BSP | 50 |
| 1.01 | 100 |
Between those points the score rises in a straight line. Because the typical low sits at 50, the scale is in two halves: from BSP down to the typical low (0-50), and from the typical low down to 1.01 (50-100).
Example: a horse starting at 5.0
| In-play low | Score | Why |
|---|---|---|
| 5.0 | 0 | No shortening at all. |
| 3.0 | 35 | Came in, but less than is typical for a 5.0 shot. |
| 2.15 | 50 | Exactly the typical in-play low. |
| 1.5 | 79 | Came in well beyond the typical low. |
| 1.01 | 100 | Traded at the minimum price (usually a winner, or very close to one). |
At very extreme prices, a simple straight line from BSP (0) to 1.01 (100) is used instead.
Single run and form
Per run
Each run gets one Impact Value. Use it to see whether a particular run attracted strong or weak in-play support.
Rolling form
On horse, jockey and trainer profiles, the rolling Impact Value is the average across all their runs in the period:
rolling IV = total of Impact Values / number of runs
This shows whether they tend to attract in-play money, not just what happened once. Best, worst and trend figures work as described in the statistics overview.
Normalised Impact Value
The normalised score is a 0-100 percentile against all rolling Impact Values for the same entity type and period, for example all horse 1-year figures. A normalised IV of 80 means the rolling IV is better than about 80% of them.
It is not another version of the per-run 0-100 score. See normalised scores.
How to use it
- Use IV when you care about market behaviour in-running, not just the result or the starting price.
- IV, PD and BR all measure in-play price moves. IV is the only one that adjusts for what is normal at each price. See In-play price moves for how they differ.
- Pair it with BSP: a 10.0 shot that trades at 4.0 and a 2.0 favourite that trades at 1.8 can score very differently.
- For jockeys and trainers, rolling figures bunch together because many runs average out, so the normalised score is usually more useful.
- Winners usually trade at or near 1.01 and score close to 100, so a high IV goes hand in hand with winning often.