Impact Value (IV) tells you how much more, or less, often horses with a certain factor win compared with the average runner. An IV of 1.50 means they win 50% more often. It tells you nothing about whether the price makes them worth backing. That second part is where most people go wrong.
Here’s how it goes wrong. A trainer is winning at 25 to 35 per cent over the last fortnight. Their runners win about 1.5 times as often as the average horse on the card. The stat is true. You back the next one up, and you lose, and you keep losing across the season.
The stat didn’t lie. It answered a different question from the one you were asking.
The short version
- IV = the win rate of horses with a factor, divided by the win rate of all runners.
- Above 1.00 means the factor wins more often than average. Below 1.00 means less often.
- It measures how often, not whether the price is right.
- Check price with the A/E ratio before you treat any factor as a bet.
What is Impact Value in horse racing?
Impact Value is a frequency ratio. It compares how often horses with a given factor win with how often a runner wins on average. A factor is anything you can sort runners by: a number on a ratings page, a trainer’s recent form, or a yes/no like “has won here before”.
The sums are simple: the factor’s win rate divided by the overall win rate. That’s the IV.
How do you work out Impact Value?
It takes three steps.
- Count the winners in your sample that carried the factor, and divide by the total winners. That’s the percentage of winners with the factor.
- Count the runners that carried the factor, and divide by the total runners. That’s the percentage of runners with the factor.
- Divide the first percentage by the second. That’s the IV.
An IV of 1.00 means the factor wins exactly as often as a random runner. 1.50 means 50% more often. 0.70 means 30% less often.
How big should it be before you take notice? I use rough rules of thumb here, and they’re my own, not tested thresholds: a reading of 1.10 to 1.30 across several thousand runners is worth a look, and above 1.40 is strong. Sample size matters more than the number. Under about 100 runners with the factor, the reading is mostly noise. A few thousand is the point where I’d start to trust it.
What does Impact Value look like on real data?
Here is Race Advisor’s UK dataset, July 2013 to March 2026, 1,664,333 runners. The overall win rate is 9.71%, and every IV below is measured against that.
| Factor | Runners | Win rate | IV |
|---|---|---|---|
| Previous course winner | 40,164 | 12.02% | 1.24 |
| Ran within the last week | 88,735 | 11.53% | 1.19 |
| Off more than a year | 22,004 | 5.71% | 0.59 |
| Trainer at 0% over last 14 days | 643,855 | 6.91% | 0.71 |
| Trainer at 25 to 35% over last 14 days | 97,825 | 14.78% | 1.52 |
| Course and distance winner | 18,861 | 11.75% | 1.21 |
| No course or distance form | over 1.17 million | 9.17% | 0.94 |
Hot trainers really do have more winners. Long absences really do drag on a horse’s chance. These are past results, so they show what has happened, not a promise of what comes next.
Does a high Impact Value mean I should back it?
No. A high IV says the factor produces winners more often. It doesn’t say whether the price pays enough to cover the horses that lose.
Markets notice what works. If a factor has produced winners for years, prices on horses carrying it tend to shorten. You can be right about the winners and still lose on the bets.
What happens when I check the price?
Run the same factors through Betfair Starting Prices (SP) using the A/E ratio. A/E is actual winners divided by the winners the prices said to expect. 1.00 is fair. Above 1.00 means more winners than the odds implied.
| Factor | IV | A/E at Betfair SP |
|---|---|---|
| Trainer at 25 to 35% | 1.52 | 1.01 |
| Previous course winner | 1.24 | 1.00 |
| Ran within the last week | 1.19 | 1.02 |
These three are illustrative rows, picked because their IVs are high. They are not a full set of every factor I’ve tested. The pattern in them is the point.
The IVs look strong. The A/Es are all within 0.02 of 1.00, which is close to fair. The market had already priced the factors in. IV shows you where winners cluster, and A/E shows whether the price is wrong. The A/E ratio explained covers that second step. These figures are before exchange commission.
How can I use Impact Value today?
Use it for two jobs. First, sorting factors. If a factor’s IV sits close to 1.00 across a fair sample, it isn’t adding information, and you can drop it from a system. If it sits well above 1.00, it earns a closer look, and then a price check. A factor well below 1.00, like a long layoff, is different. It does carry information, just on the negative side, so it’s useful for striking horses off a shortlist.
Second, reading a race quickly. Course form plus a hot trainer plus a recent run on the same horse narrows a card down to a few candidates. It won’t give you a price.
Any combination you find by cutting the data is an idea to test, not a proven edge. Why backfitted systems fail explains why, and form figure systems shows factors being tested properly.
If you want to see factors like these next to the market price on today’s runners, the racecards are the place to start.
What this doesn’t tell you
IV doesn’t say which horse will win a race, and it doesn’t say whether any price is good. It also can’t tell you if two factors are really one factor in disguise. A hot trainer and a recent run often go together, so stacking them may count the same thing twice.
