Racing data

Do horses dropped in class win? We tested 1.2 million runs

The belief is one of the most trusted in the game: a horse dropping in class, or down to a mark it has won off before, is "well-in" and ready to pop up. It sounds obvious. Across 1.2 million runs the data says close to the opposite: class droppers win less often than class risers, not more.

Every day the tipping columns flag the "well-handicapped" dropper as the bet of the day. We can test whether that reputation is earned. We took every UK and Irish runner where we could compare the class of today's race with the class of the horse's previous run, in competitive fields (five runners or more), with a settled result and a valid starting price: 1,197,170 runs. Here is what happens.

15.5% vs 9.5%

Horses stepping UP two or more classes win 15.5% of the time. Horses dropping two or more classes win just 9.5%. The direction of the move that actually wins races is up, not down.

The higher a horse climbs, the more it wins

Line the runners up by how their class changed since last time and the win rate rises smoothly from the droppers to the risers. That is the reverse of the folklore. Notice, though, the last column: every group loses at starting price and every A/E sits near 0.90, so the market prices the whole spectrum fairly evenly. The signal is in the strike rate, not in a blind bet.

Class move since last run	Runs	Win rate	ROI at SP	A/E
Dropped 2+ classes	60,811	9.5%	-21.8%	0.90
Dropped 1 class	197,565	11.1%	-20.3%	0.90
Same class	669,605	11.7%	-18.7%	0.90
Rose 1 class	218,142	11.9%	-17.8%	0.91
Rose 2+ classes	51,047	15.5%	-13.9%	0.91

Why the myth is backwards. A horse usually drops in class because it has been losing: the drop is the handicapper reacting to poor form, not handing out a gift. A horse usually rises in class because it has just won or is clearly progressive. So "dropped in class" tends to flag a horse going the wrong way, and "stepped up" tends to flag one going the right way. The market reads this too, which is why none of it is a free bet.

The same story in the ratings: "down to a winning mark" is not enough

Handicappers have a sharper version of the idea: back a horse now rated below the mark it last won off, because it is "in at the weights." We tested that directly, comparing each rated runner's mark today with the mark it last won from. Again the well-in horses win less, not more. A horse racing at or above its last winning mark wins 12.9%. A horse a stone or more below it wins barely 9%. The deep drop is usually a horse that lost its form a long time ago.

Mark vs last winning mark	Runs	Win rate	ROI at SP	A/E
At or above last winning mark	468,990	12.9%	-14.5%	0.91
Exactly at the mark	46,712	12.3%	-18.6%	0.89
1 to 5 lb below	92,700	9.4%	-21.0%	0.87
6 to 10 lb below	47,963	9.2%	-20.6%	0.88
11 lb or more below	63,047	9.1%	-17.7%	0.89

A/E (actual over expected) is wins divided by the wins the starting prices implied. Below 1 means the group won less often than its odds suggested. The well-in bands sit around 0.87 to 0.91, so they are not even quietly overperforming their price.

What this actually tells you

"Dropped in class" on its own is not a positive. In the raw numbers it is mildly negative, because a drop is normally a symptom of lost form that the handicapper, and the market, have already spotted. That does not mean a class dropper can never win: some are unexposed types being dropped to find a winnable race, and those are the ones worth finding. The skill is separating the genuinely well-placed dropper from the far larger group that is dropping simply because it is out of form. The label alone will not do that for you.

That is the whole idea behind the gap between what a horse is worth and what the market is charging. Most of the time the price already reflects the class move, in either direction, and the sensible thing is to judge each dropper on its own merits rather than backing it because it is dropping.

See today's class droppers and risers, with our rating on each

Our free Class moves page lists every horse changing class today in UK and Ireland racing, and by how many levels, with The Overlay's own read on each. Or go straight to the horses dropping in class today. It is data-driven analysis, not a tipster, and it tells you when to sit a race out.

Open today's class moves

Method, in plain English

We used every UK and Irish run in our database that had a settled result and a valid starting price (non-runners excluded), in fields of five or more, where the horse had a previous run whose class we could read: 1,197,170 runs. Class is the official race grade, from Class 1 (highest) to Class 7. The class move is today's class minus the previous run's class. The ratings table uses our record of each rated runner's current mark against the mark it last won from. Win rate is wins divided by runs. ROI is level-stakes return at starting price, staking one point per run. A/E is wins divided by the sum of 1 over decimal SP. Figures are historical, for information only, and are not a guarantee of future results.