Harness racing positions describe where a horse sits in the run. A harness field settles in two lines: on the pegs (the marker pegs along the inside of the track) and one off them. The spots that decide most races are the leader, the pocket behind it, the parked horse outside it, the one-out one-back behind the parked horse, and anything three wide.
Each spot has a cost. In Harness Racing Australia's (HRA) results for TAB meetings from 1 July to 30 September 2026, horses leading at the bell, which marks the start of the last lap, won 1,355 of their 3,265 starts (41.5%). Parked horses outside the leader won 12.1% of theirs. Starts, draws and form decide who gets each spot, and harness racing betting ties them together.
The positions in a harness race
| Spot | Where it is | What it costs or saves |
|---|---|---|
| Leader | First in the inside line | Controls the pace and runs the shortest trip, but has nothing to shelter behind |
| Directly behind the leader, on the pegs | Shortest way round and sheltered, but it can be held in with no way out | |
| Death seat, parked or in the breeze | One wide, outside the leader | Extra ground on every bend and no shelter |
| One-out one-back, or the one-one | One wide, directly behind the parked horse | The same extra ground as the parked horse, but sheltered |
| Further back on the pegs | Third, fourth or later in the inside line | Saves ground, needs gaps to open |
| Further back one out | Behind the one-out one-back horse | Sheltered, but has to come wider to pass |
| Three wide | Outside both the pegs line and the one-out line | Twice the extra ground of one wide; with cover if it follows another horse, without cover if it leads that line |
Three more terms run through race comments. Cover is having a horse in front to sit behind. A horse is parked when it races one wide with no cover. A hand-up is the leader letting a challenger cross to the front and taking the sit behind it. Glossary entries cover the death seat and the one-out one-back trail in more depth.
How often each position wins
HRA's stewards' comments record where every runner sat at the bell. Counted across the 3,270 races in the July to September 2026 results where positions were recorded, with the market's view beside each spot:
| Position at the bell | Starters | Wins | Win rate | Wins the starting odds implied | Wins / implied |
|---|---|---|---|---|---|
| Leader | 3,265 | 1,355 | 41.5% | 852.2 | 1.59 |
| 3,264 | 420 | 12.9% | 373.9 | 1.12 | |
| Parked outside the leader | 3,164 | 382 | 12.1% | 493.6 | 0.77 |
| One-out one-back | 3,128 | 315 | 10.1% | 342.7 | 0.92 |
| Further back on the pegs | 8,246 | 242 | 2.9% | 460.2 | 0.53 |
| Further back one out | 7,176 | 499 | 7.0% | 619.0 | 0.81 |
| Three wide without cover | 394 | 37 | 9.4% | 55.2 | 0.67 |
| Three wide with cover | 174 | 13 | 7.5% | 17.4 | 0.75 |
The implied wins come from the starting odds HRA prints: 1 / odds for each runner, scaled so each race adds to 100%, then added up for every horse in that spot. Menangle's comments use the 800 m point instead of the bell, so its races are left out.
The win rates mix ability with position, because better horses tend to get the good spots. The last column takes some of that out. Parked horses were fancied, with an average chance of 15.6% on their odds, second only to the leaders' 26.1%, yet they won 382 races against 493.6 implied, about 23% fewer.
Leaders won 59% more often than theirs implied. Part of that is the position, and part is hindsight: the horse in front at the bell is often the one going best on the night. Nobody knows the bell positions before the start, so these figures re-rate past runs; pricing a horse's chance of leading is worked through in harness racing betting strategy.
Track size moved the pocket: in mobile races it won 11.1% of its starts at tracks under 900 m around, against 15.2% at tracks of 900 to 1,100 m.
Why sitting parked is so costly
A horse one wide travels further on every bend. A bend turns the field through half a circle, so a horse racing a width w outside another covers pi x w more ground on each one, whatever the size of the track:
extra ground per bend = 3.1416 x 1.5 = 4.712 m one wide, at an illustrative 1.5 m width, and 3.1416 x 3.0 = 9.425 m three wide
The one-out one-back horse pays the same 4.712 m a bend as the parked horse, so ground alone does not explain the gap in the table. The difference is shelter: the parked horse has no horse in front to sit behind, and it has to match the leader's speed to hold its spot. The leader, meanwhile, decides that speed. It can ease the tempo through the middle stages while the parked horse keeps working, then quicken.
Getting cover and timing a move
A horse wide without cover wants a back to sit behind. It can drop into the one-out line, follow a horse moving forward three wide or, if parked, wait for the leader to hand up. A parked horse can also be relieved when a horse from behind ranges up three wide and takes over its spot, letting it slot in behind.
Timing is a trade between ground and time. A move three wide around one bend costs 9.425 m, as much ground as a full lap one wide. A horse that goes three wide at the bell on a 1,000 m track and stays there for both bends gives up 2 x 9.425 = 18.85 m. One that waits for the home turn gives up 9.425 m but has less time to use its speed.
Three-wide runners in the results above won 9.4% without cover and 7.5% with it, both below what their odds implied. Forecasting who will be where before the start is the job of harness racing speed maps.
Worked: equal mile rates from different positions
Example: Illustrative runs, not real races. Three horses win separate 2,200 m races at the same track, each in 1:57.0, so each race takes 117.0 x 2,200 / 1,609.344 = 159.94 seconds. They won from different spots, with an illustrative 1.5 m for each width.
| Run | Extra ground over the pegs line | The same effort on the pegs |
|---|---|---|
| Horse A led all the way on the pegs | 0 m | 1:57.0 |
| Horse B sat parked for four bends | 4 x 4.712 = 18.85 m | 117.0 x 2,200 / 2,218.85 = 116.0 s, about 1:56.0 |
| Horse C sat one-out one-back for three bends, then came three wide on the home turn | 3 x 4.712 + 9.425 = 23.56 m | 117.0 x 2,200 / 2,223.56 = 115.8 s, about 1:55.8 |
The form line shows 1:57.0 for all three. Horse B ran about a second a mile better than Horse A, with no shelter at any stage. Horse C covered even more ground, but it sat behind another horse for three of its four bends, so its run was easier than the number suggests. At Horse B's own speed of 2,218.85 / 159.94 = 13.87 m a second, its extra 18.85 m took 18.85 / 13.87 = 1.36 seconds.
The race shape matters too. A fast lead time makes the parked horse work hard early, while a slow one lets the leader save energy for the sprint home. That makes a parked win after a fast early pace the hardest kind of win to land.
Risk: Betting involves risk. Re-rating a run from its position can point you at a horse the market has underrated, but the horse may not get the same run next time and can lose for many other reasons. There is no guarantee of profit; see responsible gambling for limits and support.
How positions show in the form guide
HRA's form guide puts the position in the stewards' comments that close each past-start line. Its own sample lines include "outside leader at bell", "three wide without cover at bell", "3 wide and 1 back at bell, 3 wide late with trail" and "5 back on the pegs at bell". On HRA's results pages the same comments appear as codes, explained on its stewards' comments page:
| Code | Meaning at the bell |
|---|---|
| 1 | Leader |
| 2 | Outside leader (parked) |
| 3 | Behind leader (the pocket) |
| 4 | One out, one back |
| 5 | Three back on the pegs |
| 6 | One out, two back |
| W1 | Three wide without cover |
| W2 | Three wide and one back |
Wording varies by state. Queensland's comments add "death seat" and count horses on the pegs by the fence, so "3rd fence at bell" is the third horse in the inside line.
At Menangle, HRA's results give positions at the 800 m and 400 m points instead. In race 1 there on 3 October 2026 the winner's comment reads "800 4 400 W1": one out, one back at the 800 m, then three wide without cover at the 400 m. Harness Racing NSW's sectional report for the same race puts it 11.6 m behind the leader, one horse off the pegs, at the 800 m, then 10.6 m back and two widths out at the 400 m.
Mistakes with positions
- Reading a beaten parked horse as a bad run. In the results above, parked horses won 0.77 times what their odds implied, so a parked run that finished close was a strong one.
- Crediting a leader with a soft win. A leader that set a slow pace owed much of the win to the race shape.
- Reading the bell position as the whole race. It is one snapshot, so read the rest of the comment too: 3 wide late with trail is a different run from three wide without cover at bell.
Using positions when you price a race
Position reading changes how you rate a horse's last run and where you expect it to settle next; the price is where that view gets tested. The Terminal shows prices from 40+ bookmakers across racing and sports, and harness races share its racing board with the gallops and the greyhounds. Replays and race maps stay yours, since B337 does not rate horses or pick runners.
Once your read of where a runner will settle gives you a price, check the bookmaker's own price in your account before you bet, because the board reads prices on a repeating cycle. A free account opens a limited view of the Terminal with live odds. Bookmaker names are trade marks of their owners. B337 is not affiliated with them.
Risk: Betting involves risk. A horse you expect to lead can be crossed, a strong parked run does not promise another, and a price above your estimate can still lose. For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au. See responsible gambling for limits and support.