A harness racing barrier draw decides where each horse lines up behind the mobile: on the front row, numbered out from the pegs (the inside line of the track), or on the second row behind it. Form guides print the draw as Fr or Sr and a number, so Fr1 is the inside of the front row and Sr2 the second spot on the row behind.
The draw matters more than in the gallops because the mobile lets the field go at racing speed with every horse in its drawn spot, and the runners form two lines within a lap. In 937 random-draw mobile races at TAB meetings in Harness Racing Australia's (HRA) results for 1 July to 30 September 2026, barrier 1 won 17.8% of the time against a fair share of 11.6%. The draw is one of several things that make the trots their own code, all mapped in harness racing betting.
How harness barriers are drawn
How many horses fit across the front is the Chief Steward's call, and the rest start behind them (AHRR 130). The answer depends on the track. In the July to September 2026 results the widest front-row barrier was usually Fr6 at tracks under 900 m around, Fr7 at tracks of 900 to 1,100 m and Fr10 at Menangle, 1,400 m around on HRA's track pages. Standing starts line horses up by handicap mark instead, as mobile vs standing starts explains.
The draw itself is not always random. HRA's handicapping rules define a preferential draw as any predetermined method used instead of a random draw (rule 2.19). Its race conditions print RBD for a random barrier draw or PBD for a preferential one, followed by the criterion: PBD/NR orders by National Rating, PBD/SEX by sex, and others by recent stake money or lifetime wins. Harness Racing Victoria's guidelines list these options (clause 6.1). Of the 3,390 mobile races counted, 937 were random draws and 2,453 preferential.
Three rules can move a horse after the draw:
- a horse the stewards exclude from the mobile draw for misbehaving starts from the outside (AHRR 134), and a horse excluded after the draw goes to the outside of its line;
- after a scratching, barrier positions are filled when and how the controlling body decides (AHRR 131);
- an error in the draw stands unless it is practical to fix, and the fix is a random redraw (AHRR 132).
Front row against second row
Barrier statistics only measure the draw when the draw was random. The random-draw races give a clean comparison. A fair share is what each horse would win if the draw did not matter, 1 / field size, so each starter in a field of 9 counts 0.111 towards its group's expected wins:
| Draw (random draws only) | Starters | Wins | Win rate | Expected wins at a fair share | Wins / expected |
|---|---|---|---|---|---|
| Fr1 | 937 | 167 | 17.8% | 108.3 | 1.54 |
| Fr2 to Fr5 | 3,740 | 490 | 13.1% | 431.7 | 1.13 |
| Fr6 and wider | 1,646 | 110 | 6.7% | 178.3 | 0.62 |
| Sr1 | 729 | 57 | 7.8% | 78.4 | 0.73 |
| Sr2 and wider | 1,406 | 115 | 8.2% | 140.4 | 0.82 |
A second-row horse starts behind the front line, so it can only go forward once a gap opens, and it rarely gets the early spots. At the bell, Sr1 horses were in one of the four front spots (the lead, the pocket, outside the leader or one-out one-back) in 24.4% of their random-draw starts with a recorded position, against 68.7% for Fr1.
The second row costs less than the widest front-row draws. A horse on Sr2 or wider often follows a front-row horse into the race, while a wide front-row horse must either use its speed to cross the field or drop back behind it.
Barrier 1 against wide draws, and the race to the lead
From the inside, a horse with early speed can hold the front without crossing anyone. HRA's stewards' comments record each runner's position at the bell, and in the random-draw races the horse leading at the bell came from:
| Draw | Led at the bell |
|---|---|
| Fr1 | 239 of 867 (27.6%) |
| Fr2 to Fr5 | 510 of 3,453 (14.8%) |
| Fr6 and wider | 91 of 1,341 (6.8%) |
| Sr1 | 9 of 721 (1.2%) |
| Sr2 and wider | 21 of 1,395 (1.5%) |
A wide horse needs real gate speed to cross the field, and if it is matched by a horse inside it, it can end up parked outside the leader all the way.
The inside draw has a risk of its own: a horse on the pegs without the speed to lead can be held in with nowhere to go. Some tracks soften that with a sprint lane, which opens extra room on the inside of the home straight. HRA's track pages list one at Albion Park, Bathurst and Melton, among others, and none at Menangle, Newcastle or Penrith (checked October 2026).
Putting the draws together into a forecast of the first lap is the job of harness racing speed maps.
Does track size change the draw?
The usual argument is that small tracks punish wide draws, because the first turn comes sooner and a race has more turns. The random-draw results support part of that:
| Random draws, by HRA track size | Under 900 m | 900 to 1,100 m |
|---|---|---|
| Races | 278 | 551 |
| Fr1 led at the bell | 91 of 276 (33.0%) | 136 of 549 (24.8%) |
| Fr1 wins / expected | 1.60 | 1.62 |
| Widest front-row draw, wins / expected | 0.64 | 0.59 |
| Second row, wins / expected | 0.75 | 0.81 |
The inside horse led far more often on the smaller tracks, but each wide or second-row runner paid about the same price at both sizes.
Small tracks hit the draw in another way: they have narrower front lines. In the July to September 2026 results, 30.6% of starters at tracks under 900 m began on the second row, against 24.3% at tracks of 900 to 1,100 m.
Menangle ran only 66 random-draw races in that period, too few to read on their own. HRA's own table for Menangle over the past 12 months (checked 8 October 2026), which mixes all draw types, points away from the inside: barrier 1 won 50 of 792 starts (6.3%) and barrier 5 won 126 of 792 (15.9%).
Measuring barrier performance: wins against starters
You can run the same check on any track you bet. HRA's track pages give a head start, with a mobile barrier table of starts and wins for the past 12 months. Its barrier numbers carry on past the front line into the second row, and it mixes random and preferential draws, so check both before you lean on it. For a cleaner count, use HRA's published results, which print each runner's draw and the race's draw type:
- Take one track or one group of tracks, one start type and random draws only.
- For each barrier, count starters and wins. Win rate = wins / starters.
- Add up the fair shares: 1 / field size for every starter. Divide wins by that total.
- Allow for luck with the standard error of the win rate: the square root of win rate x (1 - win rate) / starters.
- Treat a gap from the fair share as real only when it is more than about two standard errors.
For Fr1 above, the standard error is the square root of 0.178 x 0.822 / 937 = 0.0125. Two of those either side gives 15.3% to 20.3%, well clear of the 11.6% fair share. For Sr1 it is the square root of 0.078 x 0.922 / 729 = 0.0099, a band of 5.8% to 9.8%, below its 10.8% share. A few dozen races would give bands too wide to say anything.
Then check the market. Turn each runner's starting odds into a chance (1 / odds), scale each race so the chances add to 100%, and add them up by barrier. In the random draws, Fr1 runners won 167 races against 162.1 their odds implied, and Fr6 and wider won 110 against 123.8, a gap luck can explain. The market knew about the draw.
Worked: adjusting a price for a second-row draw
Example: Illustrative numbers, invented for this example and not a current market. Runner 7 is a quick beginner. Last start it drew Fr3, and you priced it by weighing where it settles and how often it wins from each spot. Today it draws Sr2.
| Where Runner 7 settles | Its win chance from there | Chance of settling there from Fr3 | The same from Sr2 |
|---|---|---|---|
| Leads | 40% | 35% | 2% |
| Pocket or one-out one-back | 28% | 30% | 20% |
| Outside the leader | 16% | 15% | 10% |
| Further back | 10% | 15% | 53% |
| Three wide without cover | 6% | 5% | 15% |
chance from Fr3 = 0.35 x 0.40 + 0.30 x 0.28 + 0.15 x 0.16 + 0.15 x 0.10 + 0.05 x 0.06 = 0.266, a fair price of 1 / 0.266 = $3.76
chance from Sr2 = 0.02 x 0.40 + 0.20 x 0.28 + 0.10 x 0.16 + 0.53 x 0.10 + 0.15 x 0.06 = 0.142, a fair price of 1 / 0.142 = $7.04
The second row nearly halves Runner 7's chance (0.142 / 0.266 = 0.53), a bigger cut than the average second-row horse took above, because early speed is this horse's main weapon. At $5.50 a $10 bet is worth 10 x (0.142 x 5.50 - 1) = -$2.19 on your numbers; at $8.00 it is worth 10 x (0.142 x 8.00 - 1) = +$1.36. A late scratching can change the picture, since the horse in front of Runner 7 decides much of its first lap.
Risk: Betting involves risk. Every number in the table is your estimate, the draw is only one input, and Runner 7 loses about six races in seven on your own figures, even at $8.00. There is no guarantee of profit; see responsible gambling for limits and support.
Mistakes with barrier draws
- Reading barrier stats that mix random and preferential draws. In the preferential draws counted here, Fr1 won 276 of 2,453 (11.3%), about its share, which hides the 17.8% it won when the draw was random.
- Comparing barrier numbers across tracks. Fr6 is the widest front-row draw at most small tracks and a middle draw at Menangle.
- Backing the inside draw at any price. The starting odds already allowed for most of it, so Fr1 is a bet only when your price says so.
- Trusting a small sample. Under the standard-error rule a few dozen races leave bands so wide that almost any barrier can look good or bad.
- Reading the draw before the final field. Scratchings and outside-draw orders move horses.
Checking the draw against the price
Once you have adjusted for the draw, compare your price with the market. The Terminal shows prices from 40+ bookmakers across racing and sports, with exchange back and lay prices alongside, and on a harness race the best available price on each runner is picked out. Reading the draw is your job: B337 does not rate horses or pick runners.
Prices are read on a repeating cycle, so confirm the price in your bookmaker account before you bet. 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: see responsible gambling for limits and support. Barrier statistics describe past races, and past results are no guarantee of future results. A draw that looks priced wrong may simply be priced right by a market that knows more, and a bet above your fair price can still lose. For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au.