PSL Altitude Advantage

Last updated: July 2026. A living page — we re-run the analysis at the end of each season.

Home advantage in the Betway Premiership is worth about 0.40 points per game. But when a coastal club travels up to the Highveld, it loses roughly another 0.30 points per game on top of that — and the same penalty applies in reverse when Highveld clubs come down to the coast.

That second number is the one nobody publishes. We measured it ourselves across 930 matches, and the method is below so you can check it.

The Sample

We took the complete results grid for four Betway Premiership seasons, computed every club's home and away record from scratch, and validated the computation before using it: for each season, every club's derived points total had to reconcile exactly with the official published final table. All four do.

Two exclusions, both stated up front.

Royal AM, 2024/25. The club's results were expunged from the record part-way through the season, which is why the other fifteen clubs played 28 matches rather than 30. Including their voided results inflated ten clubs' totals by exactly the points they took from Royal AM fixtures — that mismatch is how we caught it. Removing the club entirely reconciles the season perfectly.

The 2021/22 season, dropped entirely. We intended five seasons. Wikipedia's results grid for 2021/22 does not reconcile with its own final table: two fixtures are recorded as dates rather than scores, and four clubs' win/draw/loss records disagree with the published standings. We checked the page twice by two different extraction methods and got identical output, so the problem is in the source, not our reading of it. Rather than publish figures we could not verify, we left the season out. Four validated seasons beat five with a known-dirty one.

How Big Is Home Advantage in the PSL?

Two things worth noting. Home advantage is remarkably stable — between +0.32 and +0.49 across four seasons, with no trend. And the draw rate is high: 29.2% of all matches finished level, close to three in ten. That is well above what a bettor arriving from the Premier League would expect, and it is the single most useful league-level fact here.

Note also how close home wins (42.0%) and away wins (28.7%) sit to the draw rate. In the PSL the draw is not a fringe outcome.

The Altitude Effect

This is the original finding. South Africa's football map splits cleanly: Johannesburg sits at about 1,753 m, Pretoria around 1,312 m, Bloemfontein 1,393 m, with Polokwane and Rustenburg between roughly 1,150 m and 1,450 m. Durban is at 8 m, Cape Town 17 m, Gqeberha 60 m.

We classified each club by its home city, split every match into four fixture types, and compared.

Read the middle two rows first, because they are the control. A Highveld club hosting another Highveld club takes 1.51 points per game. A coastal club hosting another coastal club takes 1.55. Statistically indistinguishable — the difference is −0.03 points per game with a p-value of 0.82. Highveld home grounds are not inherently stronger, and Highveld clubs are not inherently better.

Now the outer rows. Against a coastal visitor, a Highveld host jumps to 1.81 ppg — a gain of +0.30 points per game over hosting a fellow Highveld side (95% confidence interval +0.03 to +0.57, p = 0.032). And a coastal host facing a Highveld visitor drops to 1.25 ppg, a fall of −0.30 against hosting a fellow coastal side (95% CI −0.59 to −0.01, p = 0.043).

The effect is the same size in both directions, and the control shows no baseline difference. That symmetry is what makes this credible: it is the journey across the altitude divide that costs the visiting team, not the quality of the clubs at either end.

The defensive column tells the same story. Highveld hosts concede 0.75 goals per game to coastal visitors, against 0.95 to Highveld visitors.

The honest caveats. These are two tests at p ≈ 0.03–0.04 — real, but not overwhelming. A chi-squared test on the raw win/draw/loss split for Highveld hosts gives 5.38 on 2 degrees of freedom (p = 0.068), which falls just short of the 5% threshold. So the points effect clears significance while the result-distribution test sits marginally outside it. Four seasons is a decent sample, not a huge one, and altitude is bundled with travel distance and climate — we are measuring the combined effect of the trip, not oxygen alone.

Four clubs are excluded from this comparison: TS Galaxy (Mbombela, 660 m) and Maritzburg United (about 596 m) sit between the bands, and Marumo Gallants and Royal AM changed home venues during the window.

Home Advantage by Club

Across the same 930 matches, for clubs with at least 40 home games in the window.

TS Galaxy have the largest home advantage in the league at +0.71 points per game — which fits their mid-altitude position at Mbombela, awkward for Highveld and coastal visitors alike.

At the other end, Kaizer Chiefs and Orlando Pirates have among the smallest home advantages in the division. Both are Soweto clubs playing in front of large crowds, and both were better away than at home in 2025/26 specifically. Whatever the FNB Stadium and Orlando Stadium provide, it is not the edge their reputations imply.

What About Referees?

The brief for this page included referee tendencies. We are not publishing them, because we could not source referee-level data at a quality we would stand behind. Per-match officiating appointments with card and penalty counts are not available in a complete, verifiable form for these seasons from the sources we use. Rather than build a table from partial data and present it as analysis, we have left it out, and will add it if a reliable dataset becomes available.

What the Numbers Mean

Fixture direction matters more than league tables show. A coastal side travelling to the Highveld is measurably worse off than the same side playing a home-and-away pair against a fellow coastal club. Treating every away trip as equivalent hides a real and repeatable effect.

The draw is a bigger part of this league than of most. At 29.2% league-wide, and higher in specific fixtures, the PSL draw rate is high — a direct consequence of the low-scoring profile documented in our goals-per-game record.

Do not over-read a single club's home record. The per-club table above rests on 44–59 home matches each — enough to be indicative, not enough to be decisive. The league-level and altitude findings sit on far larger samples and are the more reliable numbers on this page.

And the limits. This is descriptive analysis of what happened in four seasons. It does not forecast individual matches, squads turn over heavily between seasons, and the altitude effect is a tendency across 175 matches, not a rule that applies to the next one.

FAQ

How big is home advantage in the PSL? About 0.40 points per game across 930 matches in four seasons: home sides average 1.55 points per game against 1.15 away. Home teams win 42.0% of matches, 29.2% are drawn and away teams win 28.7%.

Does altitude really affect PSL results? On our measurement, yes. A Highveld club hosting a coastal visitor takes 1.81 points per game, against 1.51 when hosting another Highveld club — a gain of 0.30 points per game (p = 0.032). Coastal clubs hosting Highveld visitors show the same effect in reverse, dropping 0.30 points per game (p = 0.043).

Could that just mean Highveld clubs are better? No, and this is the key control. Highveld clubs hosting Highveld clubs average 1.51 points per game; coastal clubs hosting coastal clubs average 1.55. The difference is statistically indistinguishable from zero (p = 0.82). The penalty attaches to crossing the altitude divide, not to the clubs themselves.

Which PSL club has the biggest home advantage? TS Galaxy, at +0.71 points per game between home and away — consistent with Mbombela's mid-altitude position, which is unfamiliar to Highveld and coastal visitors alike. Kaizer Chiefs and Golden Arrows have the smallest, at +0.17.

How many draws are there in the PSL? 29.2% of matches in our four-season sample, close to three in ten, and stable season to season between 25.7% and 32.1%.

Why does this page use four seasons rather than five? We excluded 2021/22 because the public results grid for that season does not reconcile with its own published final table — two fixtures show dates instead of scores and four clubs' records disagree with the standings. We verified the fault by two independent extraction methods before dropping the season.

Download the data

Every figure on this page is in the files below, so you can check the arithmetic yourself. Released under CC BY 4.0 — free to reuse, including commercially, with attribution.

• Home advantage by season — CSV, 5 rows
  Columns: season, matches, home_win_pct, draw_pct, away_win_pct, home_ppg, away_ppg, home_advantage_ppg
• Results by altitude-band fixture type — CSV, 4 rows
  Columns: fixture_type, matches, home_win_pct, draw_pct, home_ppg, home_goals_for_pg, home_goals_against_pg
• Home advantage by club — CSV, 14 rows
  Columns: club, altitude_band, home_games, home_ppg, away_ppg, home_advantage_ppg