Does tournament environment change favourite win rates?

Favourites win most ATP matches. That is not news. What is less obvious is whether the place—how high the court sits, how humid the month usually is, whether the roof is closed—shifts how often those favourites actually close.

We joined 8,254 ATP matches with bookmaker odds (2022–2025) to a curated venue table (altitude, region, indoor/outdoor) and to monthly climate normals for outdoor sites: average temperature and relative humidity for the calendar month of the match at that latitude/longitude. Across the full sample, favourites won 70.3% of matches (5,804 / 8,254; 95% Wilson interval 69.3–71.3%). The environment cuts that matter are not “Asia vs Europe” in the aggregate. They are altitude on clay and, on outdoor hard, humid mild months vs mid-humidity mild months.

This is descriptive association, not a causal claim that thin air causes upsets. It is also not match-day weather: we use multi-year monthly normals, not the thermometer on the afternoon of the match. Indoor events are kept as their own bin for climate cuts.

Published: July 30, 2026

---

How we measured environment

Every match in this study has a tournament name, a date, a surface, two decimal odds lines, and a winner. We drop rows without usable odds. The favourite is the player with the lower decimal price.

We map each tournament name to a venue record with approximate elevation in metres, region, and whether the event is typically indoor. For outdoor venues we attach climate normals for the match’s calendar month: mean 2 m temperature and mean relative humidity from a multi-year archive window (2010–2019), aggregated by month. Venue join coverage is 8,254 / 8,260 matches with odds (99.9%); the six unmatched rows are non-tour league noise. Outdoor climate coverage is 7,353 / 7,353 (100%). Indoor matches (n = 901) are excluded from temperature and humidity bins.

Bins used throughout:

• Altitude: sea 0–200 m, mid 200–800 m, high 800 m+
• Temperature (monthly avg): cool <18°C, mild 18–26°C, hot >26°C
• Humidity (monthly avg): dry <55%, mid 55–70%, humid >70%
• Region: Europe, North America, Asia, South America, Oceania, Middle East
• Indoor vs outdoor (venue default or indoor surface label)

We report sample size n, favourite win rate, upset rate, and a Wilson 95% interval on every cut. Cells with n < 100 are flagged and treated as exploratory only.

If you care more about how markets behave once a set is decided, our first-set closeout analysis is the complementary view. Surface folklore—especially grass—has its own deep dive in the grass court specificity guide. Clay swing readers can cross-check tournament context in the clay court betting guide.

Why monthly normals instead of match-day weather

Day-level forecasts are noisy, incomplete in historical caches, and easy to overfit. A tournament’s typical May or September climate is stable: you can ask whether humid late-summer hard courts or mild Mediterranean spring clay months associate with favourite hold without scraping thousands of hourly archives. If normals show no signal, day-level weather is unlikely to be a first-order favourite-hold feature. If normals do show signal—as mid-altitude clay and mild+humid hard do here—then day-level weather becomes a candidate for a later study, not a requirement for v1.

Overall favourite baseline

Before slicing environment, sanity-check the market:

Metric	Value
Matches with odds	8,260
Joined to venues	8,254
Favourite wins	5,804
Favourite WR	70.3%
Upset rate	29.7%
Wilson 95%	69.3–71.3%

Seventy percent is the tour’s “favourites usually win” rhythm with closing-style decimal odds in our cache. Every environment cut below should be read as a deviation from that baseline, not as a standalone probability of winning a random match. When a cut moves favourite hold by three to six percentage points on a sample of hundreds or thousands of matches, that is the scale of signal we care about—large enough to notice in research, small enough that bankroll rules still matter more than any single environmental story.

---

Altitude bands: mid-elevation is the soft spot

Figure 1: Favourite win rate by altitude band, ATP 2022–2025, matches with odds. Error bars are Wilson 95% intervals.

Favourite win rate by altitude:

Altitude band	Matches	Favourite WR	Upset rate	Wilson 95%
Sea 0–200 m	6,714	71.0%	29.0%	69.9–72.0%
Mid 200–800 m	1,382	66.6%	33.4%	64.1–69.1%
High 800 m+	158	74.7%	25.3%	67.4–80.8%

Sea-level venues dominate the calendar, so the 71.0% baseline is the tour’s everyday rhythm. Mid-altitude venues sit 4.4 percentage points lower at 66.6% (n = 1,382). The Wilson intervals barely overlap the sea-level band, which is why this cut survives a first look.

The high-altitude band (74.7%, n = 158) looks more favourite-friendly, but the sample is thin and mixes indoor and outdoor events at the extremes. Treat the 800 m+ row as a pointer, not a policy.

What this does not prove:

• That every Madrid clay week is an automatic upset factory (Madrid sits in the mid band by elevation, but week-to-week field strength still dominates)
• That thin air alone explains the gap (player pools, event levels, and surface mix co-vary with altitude)

It does show that lumping all outdoor clay or all outdoor hard together hides a real altitude split in favourite hold rates.

Which venues sit where?

Elevation bands are approximate city/venue elevations, not court-by-court surveys. In plain language:

Sea (0–200 m) covers most of the calendar: Australian Open and Melbourne-week events, Miami, Indian Wells (low desert elevation in our table), Roland Garros, Rome, Monte Carlo, Wimbledon and the London grass swing, US Open, Shanghai, and the bulk of European summer clay near the coast.

Mid (200–800 m) is where the soft favourite-hold signal lives. Think Madrid-class clay elevation, Munich, parts of the South American clay swing inland, and other inland European stops. These weeks still feature strong favourites—but historically those favourites have closed less often than at sea-level clay.

High (800 m+) is a small bag: Gstaad-scale outdoor clay and a handful of indoor stops at elevation. Sample sizes collapse quickly, which is why we refuse to let the high band drive the story.

The important discipline is not memorising every city’s metres above sea level. It is noticing when a clay week sits inland and elevated versus when it sits near sea level—and checking whether the market favourite’s price already embeds that week’s usual chaos.

---

Clay × altitude: the main finding

Figure 2: Surface × altitude favourite win rates for the main cells. The clay sea vs clay mid gap is the headline.

When we cross surface with altitude, the mid-band soft spot concentrates on clay:

Cell	Matches	Favourite WR	Upset rate
Clay · sea 0–200 m	1,766	72.5%	27.5%
Clay · mid 200–800 m	755	66.6%	33.4%
Hard · sea 0–200 m	3,210	70.5%	29.5%
Hard · mid 200–800 m	459	67.3%	32.7%
Grass · sea 0–200 m	983	70.6%	29.4%

Sea-level clay favourites hold at 72.5%. Mid-altitude clay favourites hold at 66.6%—a 5.9 pp gap across hundreds of matches on each side. Hard courts move in the same direction (70.5% → 67.3%) but with a smaller sample at mid altitude (n = 459).

True high-altitude clay (Gstaad-scale events in the 800 m+ bucket) shows 73.1% favourite WR on only 93 matches—low n. Do not build a narrative on that cell alone. The robust story is mid-elevation clay versus coastal / low-elevation clay.

Key findings:

• Mid-altitude clay is where favourites bleed relative to sea-level clay
• Hard mid-altitude leans the same way but with wider uncertainty
• “High altitude clay chaos” is not supported in this sample once you demand n ≥ 100

For clay-season betting frames that start from surface rather than elevation, pair this section with the Roland Garros–oriented clay guide and remember elevation is an extra axis, not a replacement for form and matchup.

Priors we tested (and what held)

Before running the cuts, three priors showed up in tennis conversation:

1. High-altitude clay is chaos (Madrid / Gstaad / Mexico folklore).
   Result: The robust clay gap is mid vs sea (66.6% vs 72.5%), not the tiny high-altitude clay cell (n = 93). Madrid-class elevation falls in the mid band in our bins. “High altitude” as a single slogan oversells a thin Gstaad-scale sample and undersells the mid band where most elevated clay actually lives.

2. Hot + humid outdoor hard is softer for favourites (US Open / Cincinnati month normals).
   Result: On outdoor hard, the well-sampled contrast is mild+humid (68.4%, n = 1,501) vs mild+mid humidity (72.0%, n = 1,421). Explicit hot×humid cells print high favourite rates but with n < 100—we do not promote them.

3. Indoor favourites are more stable.
   Result: Rejected at this cut. Indoor 69.4% vs outdoor 70.4%.

Those three answers are the spine of the article. Everything else is context.

Surface overall (for orientation)

Environment cuts should not be confused with raw surface favourite rates in the same sample:

Surface	Matches	Favourite WR	Upset rate
Clay	2,614	70.8%	29.2%
Grass	1,070	70.6%	29.4%
Hard	3,669	70.1%	29.9%
Indoors	901	69.4%	30.6%

Surfaces alone sit within about a point of each other. Altitude inside clay moves the needle more than switching from clay to hard at the tour aggregate. That is the point of crossing dimensions instead of stopping at surface labels.

---

Indoor vs outdoor: almost no gap

Figure 3: Indoor and outdoor favourite win rates are nearly identical.

Indoor vs outdoor favourite win rate:

Setting	Matches	Favourite WR	Upset rate	Wilson 95%
Outdoor	7,353	70.4%	29.6%	69.4–71.5%
Indoor	901	69.4%	30.6%	66.3–72.3%

A one-point gap with overlapping intervals is not a story. Controlled indoor conditions do not, in this sample, produce a meaningfully more favourite-stable market than outdoor tennis overall. If you expected indoor events to lock favourites in, the data does not back a large edge.

That does not mean serve-speed or bounce variance are identical indoors and outdoors. It means that once the market has set a favourite, the hold rate is statistically similar at this cut.

Indoor tennis still changes how matches play—shorter points on fast carpets or hard courts under a roof, different return positions, less wind. Markets appear to absorb much of that into the opening price. Favourite hold after the price is set does not jump. If you are hunting indoor edges, look at serve/return profiles and schedule congestion, not at a blanket “indoor favourites cash more often” rule. This sample does not support that rule.

---

Region: small spread, no clean hierarchy

Figure 4: Favourite win rates by region. Europe and North America dominate the sample; all regions sit roughly 68–71%.

Favourite win rate by region:

Region	Matches	Favourite WR	Upset rate
Europe	3,882	70.6%	29.4%
North America	2,293	70.5%	29.5%
Oceania	764	70.5%	29.4%
Asia	659	69.3%	30.6%
Middle East	346	69.7%	30.3%
South America	310	68.1%	31.9%

The tour’s continental labels look dramatic on a calendar graphic and quiet in a favourite-hold table. Europe, North America, and Oceania sit on top of each other near 70.5–70.6%. Asia and the Middle East are within a point. South America is the softest at 68.1% (n = 310), still inside a range you would expect from sampling noise plus a clay-heavy slate.

Takeaway: region alone is a weak favourite-hold feature. If humidity or altitude matter, they matter through venue physics and calendar month—not through the passport stamp on the tournament poster.

Asia’s humid outdoor hard weeks and South America’s clay elevation mix are better analysed through the climate and altitude cuts above than through a single “Asia upset rate” number. The region table is useful mainly as a negative result: it tells you where not to invent a story.

---

Monthly temperature and humidity normals

Climate here means typical conditions for that venue in that month, not whether Tuesday was sticky. Indoor matches stay in an “indoor” climate bin and are not assigned outdoor normals.

Temperature bands (outdoor months)

Temp band (monthly avg)	Matches	Favourite WR	Upset rate
Mild 18–26°C	4,436	70.1%	29.9%
Cool <18°C	2,682	70.7%	29.3%
Hot >26°C	235	74.0%	26.0%
Indoor (excluded)	901	69.4%	30.6%

Mild and cool months are essentially tied. The hot band (74.0%) looks favourite-friendly but rests on only 235 matches—enough to notice, not enough to crown a heat rule.

Humidity bands (outdoor months)

Humidity band (monthly avg)	Matches	Favourite WR	Upset rate
Humid >70%	4,665	70.5%	29.5%
Mid 55–70%	2,044	70.8%	29.2%
Dry <55%	644	68.6%	31.4%

Aggregate humidity alone barely moves the needle. The interesting cut appears when we restrict to outdoor hard and cross temperature with humidity.

Outdoor hard: mild + humid vs mild + mid humidity

Figure 5: Outdoor hard courts only. Mild + humid months show lower favourite hold than mild + mid-humidity months.

Outdoor hard · monthly climate cells (n ≥ 100):

Climate cell	Matches	Favourite WR	Upset rate
Mild + humid (>70%)	1,501	68.4%	31.6%
Mild + mid humidity (55–70%)	1,421	72.0%	28.0%
Mild + dry (<55%)	462	69.3%	30.7%
Cool + humid	130	63.1%	36.9%

The cleanest climate contrast: on outdoor hard in mild months, favourites hold 68.4% when the month is typically humid versus 72.0% when humidity sits in the mid band—about 3.6 pp, with both samples above 1,400 matches.

Hot × humid and hot × mid cells print high favourite rates (~76–77%) but on n < 100. We list them in the study appendix tables and leave them out of the headline.

Honest framing:

• This supports “humid mild hard-court months are slightly softer for favourites” as a calendar-normal pattern
• It does not say “bet the underdog whenever the forecast shows 80% humidity tomorrow”
• US Open / Cincinnati weeks often land in humid late-summer normals; that is context, not a betting system

How to read a hard-court week with this lens

When an outdoor hard event lands in a mild month whose long-run humidity normal is above 70%, favourites have held less often than in otherwise similar mild months with mid-range humidity. That is a 3.6 pp gap on large samples—real, modest, and easy to overtrade.

Practical use:

1. Check whether the week’s climate story is “mild and sticky” versus “mild and moderate.”
2. Ask whether the market favourite is already priced like a coin flip or like a -200 chalk.
3. Only then decide whether environment is a tie-breaker. It is rarely the whole bet.

Cool+humid outdoor hard (63.1%, n = 130) is directionally softer still, but the sample is thinner—note it, do not build a product around it.

---

What this means for reading matches

Environment is a background variable. It should not overpower ranking gaps, surface fit, or recent form. Used carefully, the study suggests a short checklist:

Useful checks:

• On clay, ask whether the venue is mid-elevation (roughly 200–800 m) or near sea level. Mid-altitude clay has been the softer favourite-hold bucket in 2022–2025.
• On outdoor hard, ask whether the month’s climate normal is mild + humid versus mild + mid humidity. The hold gap is modest but well sampled.
• Do not expect indoor vs outdoor alone to change favourite hold rates much.
• Do not treat continental region as a strong favourite-stability signal.

How we use this on TennisPredictor:

Live projections already fold surface, form, and market context into match-level probabilities. Environment cuts like these are research signals—candidate features for later model work, not live toggles on the site today. For today’s board, use the live dashboard and treat altitude/climate as secondary context when two players look close on clay or humid hard.

If you are building a mental model of when favourites close after winning the opener, stack this article with the favourite first-set closeout rates piece. If grass season framing is the question, start from the grass court specificity guide and remember altitude is rarely the grass story—almost all grass in this sample sits at sea level.

What we are not claiming

Environment research attracts overclaim. Here is the explicit reject list for this article:

• We are not saying markets are inefficient by X units of expected value.
• We are not publishing a staking plan keyed to metres above sea level.
• We are not substituting climate normals for injury news, travel, or matchup.
• We are not asserting that thin air causes upsets without a physical mediation study.
• We are not mixing this favourite definition with older blog cuts that used a different odds-to-winner mapping.

What we are claiming is narrower: in this 2022–2025 ATP odds sample, mid-altitude clay and mild+humid outdoor hard months show lower favourite hold rates than their natural comparison cells, with sample sizes large enough to take seriously and confidence intervals that do not shrug the gaps away.

---

Method notes readers should keep

Favourite definition. Lower decimal odds. Ties on price are rare; we treat the first listed side as favourite only when odds are equal (effectively absorbed in the sample).

Climate normals. Open-Meteo archive daily means, averaged within each calendar month over 2010–2019 for each outdoor venue coordinate. That answers “is this tournament’s typical climate associated with favourite WR?” It does not answer “did humidity spike on match day?”

Altitude. Curated approximate venue elevation. Cities with multiple courts share one elevation. Bands are chosen for interpretability, not as physics thresholds.

Coverage. 99.9% venue join; 100% outdoor climate join among joined outdoor matches; indoor climate deliberately N/A.

Low-n cells. High-altitude clay (n = 93), hot outdoor-hard climate cells (n < 100), and a few indoor × altitude splits are shown in study tables but not treated as headline evidence.

---

Reference: selected venues

The table below is a reader-facing slice of the venue file used in the study, ordered by event month (calendar order). Event month is the most common match month for that tournament in our 2022–2025 odds sample. Temp / humidity are that month’s climate normals at the venue coordinates (outdoor only). Indoor events show climate as N/A.

Tournament	Region	Altitude	Band	Setting	Event month	Month normal
Australian Open	Oceania	30 m	sea	outdoor	Jan	20.6°C / 64%
Santiago	South America	570 m	mid	outdoor	Feb	21.6°C / 54%
Indian Wells	North America	50 m	sea	outdoor	Mar	19.8°C / 35%
Miami	North America	5 m	sea	outdoor	Mar	22.1°C / 68%
Barcelona	Europe	10 m	sea	outdoor	Apr	13.9°C / 76%
Madrid	Europe	650 m	mid	outdoor	Apr	12.6°C / 61%
Monte Carlo	Europe	40 m	sea	outdoor	Apr	13.8°C / 75%
Munich	Europe	520 m	mid	outdoor	Apr	9.2°C / 71%
French Open	Europe	40 m	sea	outdoor	May	14.1°C / 72%
Rome	Europe	30 m	sea	outdoor	May	17.2°C / 73%
Queen's Club	Europe	20 m	sea	outdoor	Jun	15.7°C / 74%
Gstaad	Europe	1,050 m	high	outdoor	Jul	16.1°C / 80%
Kitzbühel	Europe	760 m	mid	outdoor	Jul	17.4°C / 80%
Wimbledon	Europe	30 m	sea	outdoor	Jul	18.0°C / 72%
Cincinnati	North America	220 m	mid	outdoor	Aug	24.2°C / 69%
US Open	North America	10 m	sea	outdoor	Aug	23.6°C / 73%
Basel	Europe	260 m	mid	indoor	Oct	N/A
Paris Masters	Europe	40 m	sea	indoor	Oct	N/A
Shanghai	Asia	5 m	sea	outdoor	Oct	18.9°C / 76%
ATP Finals	Europe	240 m	mid	indoor	Nov	N/A

Madrid’s April normal (12.6°C) is cool relative to coastal clay in May—useful context for why elevation and calendar month should be read together, not as slogans.

---

Appendix: all venues in the study

Full curated list (78 ATP venues), ordered by event month then name. Event month is the modal match month in the 2022–2025 odds sample.

Tournament	Region	Altitude	Band	Indoor	Event month
Adelaide	Oceania	50 m	sea	no	Jan
Auckland	Oceania	30 m	sea	no	Jan
Australian Open	Oceania	30 m	sea	no	Jan
Brisbane	Oceania	30 m	sea	no	Jan
Hong Kong	Asia	30 m	sea	no	Jan
Sydney	Oceania	20 m	sea	no	Jan
Acapulco	North America	30 m	sea	no	Feb
Buenos Aires	South America	25 m	sea	no	Feb
Cordoba	South America	400 m	mid	no	Feb
Dallas	North America	130 m	sea	yes	Feb
Delray Beach	North America	5 m	sea	no	Feb
Doha	Middle East	10 m	sea	no	Feb
Dubai	Middle East	10 m	sea	no	Feb
Marseille	Europe	10 m	sea	yes	Feb
Montpellier	Europe	30 m	sea	yes	Feb
Pune	Asia	560 m	mid	no	Feb
Rio de Janeiro	South America	10 m	sea	no	Feb
Rotterdam	Europe	5 m	sea	yes	Feb
Santiago	South America	570 m	mid	no	Feb
Indian Wells	North America	50 m	sea	no	Mar
Miami	North America	5 m	sea	no	Mar
Barcelona	Europe	10 m	sea	no	Apr
Belgrade	Europe	120 m	sea	no	Apr
Bucharest	Europe	90 m	sea	no	Apr
Estoril	Europe	20 m	sea	no	Apr
Houston	North America	15 m	sea	no	Apr
Madrid	Europe	650 m	mid	no	Apr
Marrakech	Middle East	450 m	mid	no	Apr
Monte Carlo	Europe	40 m	sea	no	Apr
Munich	Europe	520 m	mid	no	Apr
French Open	Europe	40 m	sea	no	May
Geneva	Europe	375 m	mid	no	May
Lyon	Europe	170 m	sea	no	May
Rome	Europe	30 m	sea	no	May
Eastbourne	Europe	10 m	sea	no	Jun
Halle	Europe	100 m	sea	no	Jun
Hertogenbosch	Europe	5 m	sea	no	Jun
Mallorca	Europe	30 m	sea	no	Jun
Queen's Club	Europe	20 m	sea	no	Jun
Stuttgart	Europe	250 m	mid	no	Jun
Atlanta	North America	320 m	mid	no	Jul
Bastad	Europe	10 m	sea	no	Jul
Gstaad	Europe	1,050 m	high	no	Jul
Hamburg	Europe	10 m	sea	no	Jul
Kitzbühel	Europe	760 m	mid	no	Jul
Newport	North America	10 m	sea	no	Jul
Umag	Europe	5 m	sea	no	Jul
Wimbledon	Europe	30 m	sea	no	Jul
Cincinnati	North America	220 m	mid	no	Aug
Los Cabos	North America	10 m	sea	no	Aug
Montreal	North America	40 m	sea	no	Aug
Toronto	North America	80 m	sea	no	Aug
US Open	North America	10 m	sea	no	Aug
Washington	North America	50 m	sea	no	Aug
Winston Salem	North America	290 m	mid	no	Aug
Beijing	Asia	50 m	sea	no	Sep
Chengdu	Asia	500 m	mid	no	Sep
Hangzhou	Asia	20 m	sea	no	Sep
Metz	Europe	180 m	sea	yes	Sep
San Diego	North America	20 m	sea	no	Sep
Seoul	Asia	40 m	sea	no	Sep
Sofia	Europe	550 m	mid	yes	Sep
Tel Aviv	Middle East	10 m	sea	yes	Sep
Almaty	Asia	800 m	high	yes	Oct
Antwerp	Europe	10 m	sea	yes	Oct
Basel	Europe	260 m	mid	yes	Oct
Brussels	Europe	50 m	sea	yes	Oct
Florence	Europe	50 m	sea	yes	Oct
Gijon	Europe	10 m	sea	yes	Oct
Napoli	Europe	20 m	sea	no	Oct
Paris Masters	Europe	40 m	sea	yes	Oct
Shanghai	Asia	5 m	sea	no	Oct
Stockholm	Europe	30 m	sea	yes	Oct
Tokyo (Japan Open)	Asia	40 m	sea	no	Oct
Vienna	Europe	170 m	sea	yes	Oct
Athens	Europe	70 m	sea	yes	Nov
ATP Finals	Europe	240 m	mid	yes	Nov
Next Gen ATP Finals	Middle East	15 m	sea	yes	Nov

Bands: sea 0–200 m · mid 200–800 m · high 800 m+.

Frequently asked questions

Does altitude cause more upsets?

We do not claim causation. Mid-altitude venues (200–800 m) show a lower favourite win rate (66.6%, n = 1,382) than sea-level venues (71.0%, n = 6,714). The gap is largest on clay (72.5% sea vs 66.6% mid). Field strength, event level, and surface still matter more than elevation alone.

Is this match-day weather?

No. Temperature and humidity here are monthly climate normals for the venue’s coordinates in the match’s calendar month. Indoor events are not assigned outdoor climate values.

Do indoor tournaments favour the favourite more?

Not in this sample. Indoor favourite WR is 69.4% (n = 901) versus 70.4% outdoor (n = 7,353). The intervals overlap; the gap is not material.

Which climate pattern shows the clearest hard-court signal?

Outdoor hard in mild months: humid normals (68.4%, n = 1,501) versus mid-humidity normals (72.0%, n = 1,421). Hot-month cells are too small (n < 100) for strong claims.

How many matches are in the study?

8,254 ATP matches with odds from 2022–2025 joined to the venue table. Overall favourite win rate is 70.3% (Wilson 95%: 69.3–71.3%).

Should I bet underdogs at altitude?

Not as a blanket rule. Use elevation and climate as context when prices already look soft—especially on mid-altitude clay—and always size risk responsibly. Environment does not replace matchup analysis.

---

Bottom line

Across 8,254 ATP matches with odds (2022–2025), favourites win about seven in ten. Environment does not rewrite that baseline, but it does carve meaningful niches:

The numbers that stick:

• Mid-altitude favourites: 66.6% vs sea-level 71.0%
• Clay sea vs clay mid: 72.5% vs 66.6% (−5.9 pp)
• Outdoor hard mild+humid vs mild+mid humidity: 68.4% vs 72.0%
• Indoor vs outdoor: essentially flat (69.4% vs 70.4%)
• Regions: roughly 68–71%, no dramatic hierarchy

Altitude on clay and humid mild hard-court months are the environment signals worth remembering. Everything else in this cut is either noise-adjacent or too small to trust.