Numbers Don't Lie with Craig O'Shannessy of Brain Game Tennis

Summary

Craig O’Shannessy — Australian-born tennis analyst and founder of Brain Game Tennis, hired by all four Grand Slams for data analysis, contributor to the New York Times and major publications, and author of eight tactical products — joins Lisa Stone at the 2017 US Open to present his data-driven framework for tennis strategy and player development. O’Shannessy holds a journalism degree from Baylor and spent 20 years as a practicing coach before pivoting to analytics in 2015. His central finding: 70% of all points in professional tennis end in four shots or fewer (two shots per player), yet practice courts are “obsessed with the end of the point” — the consistency and rally work that represents only 10% of actual match play. He introduces his “Game Plan” product for junior development and discusses the tactical implications of rally length data, court positioning, gender parity in match analytics, and the correct teaching hierarchy of technique vs. tactics.

Guest Background

Craig O’Shannessy grew up in Australia playing tennis extensively and transferred through multiple US colleges before graduating from Baylor University with a journalism degree. He worked as a tennis coach for 20 years before launching braingametennis.com with eight tactical products. In 2012 he was contacted by the New York Times for a serve data article and began combining journalism and tennis analysis. He has been contracted by all four Grand Slams, writes for major tennis publications, and was at the 2017 US Open as an official analyst with full IBM data access. He describes himself as “not a stats person — a strategy person who talks stats because that’s the language that quantifies it and removes the opinion and guesswork.”

Key Findings

1. The 70% Finding — Tennis Is a Front-Loaded Sport

O’Shannessy’s foundational finding from the 2015 Australian Open: 70% of all points at the professional level end within four shots (two shots per player). 20% end in shots 5-8 (patterns of play). Only 10% of points reach nine-plus shots (extended rallies). This distribution — which O’Shannessy calls “first strike tennis” — means that practice courts are systematically inverted: “We’re spending almost 90% of our practice court practicing what happens only about 10% of the time in a match.” The average rally length across all tennis is 3.5 to 4 shots. The most common rally length by a factor of two is a rally of one — serve in, return doesn’t come back. This data applies consistently from Grand Slam to junior level, men’s and women’s equally.

2. The 90%+ Winning Correlation — Dominate 0-4 or Lose

O’Shannessy frames the winning data as a hand-raise exercise: of players who won their match, over 90% won more points in the 0-4 rally range than their opponent. “It’s almost hand in hand. It’s almost given.” In the 5-8 range, about 60% of match winners won that segment too. In the 9-plus range, the number drops to 54% — barely above coin-flip. “The longer a point goes, the more even it becomes. There is almost no advantage in the longer points. All of the winning and losing is primarily done earlier in the rally.” At the 2015 US Open Ladies, the 9-plus advantage was literally 50-50.

3. The Spanish Method Critique — Data Doesn’t Support It

O’Shannessy explicitly challenges the clay court / consistency-first development philosophy that dominated American junior development: “It sounds good in theory. I mean, what could possibly be wrong with being more consistent than my opponent or being fitter? But when you look at the raw match data, it’s not supported at all.” Evidence: even the top Spanish players (Nadal, Ferrer, Bautista Agut, Lopez) win more points in the 0-4 range than in the 9-plus range. “Especially Rafa — way more points in the 0-4 rally length than in 9-plus.” The Spanish players are also first-strike players; the narrative that they win through extended rallies is not born out by the actual match data. The label of “counterpuncher” or “defensive baseline” is a secondary label — every player on the planet is a first-strike player first.

4. The Return of Serve — The Most Underpracticed Skill in Tennis

O’Shannessy makes a provocative argument: the return of serve may matter more to winning than the serve itself. Evidence: the number one rally length in tennis is a rally of one — serve in, return does not come back. “Of all the players from 1991 to 2016 that finished number one in the world, the six most prolific return years on the ATP have been the last six years in a row.” We are in the “golden age of the return.” Yet on practice courts, the return of serve is almost never isolated and drilled systematically. “We head four hands, we hit tons of four hands, we hit tons of backhands — we almost never practice the return of serve.” The number one rally length in tennis ends with a misreturn, and that outcome is the least practiced skill at every level.

5. Gender Parity in Analytics — Coach 12-Year-Old Boys and Girls Identically

O’Shannessy presents a finding that directly challenges gender-segregated tactical coaching: men’s and women’s match analytics are nearly identical. Forehand winners vs. backhand winners at the first round of the US Open: 65% forehands for men, 65% for women. Baseline points won: 46% for men, 47% for women. Net points won: 65% for men, 65% for women. “If you show me 100 different analytics between men’s tennis and women’s tennis, they would almost always be the same.” His coaching conclusion: “If you sent me a 12-year-old boy or a 12-year-old girl, I am coaching those two kids exactly the same way. There is no such thing as the women’s game and the men’s game in the way the tactics are organized.”

6. Technique First, Then Tactics — But Introduce Tactics Immediately

O’Shannessy addresses the technique-vs-tactics debate directly. He agrees technique must come first: “The number one role of a junior coach is to first and foremost make sure their player has clean, simple, repeatable technique.” But he demonstrates that tactics can be embedded in technique drills from day one — his example: teaching a forehand by having a player stand slightly in the ad court, hitting to a target deep in the ad court on the other side. The player is thinking only technique, but the drill is simultaneously building depth, height over the net, run-around forehand positioning, and cross-court forehand pattern. “You can teach technique and tactics together. And you absolutely should from a young age.”

7. Court Position as a Match-Deciding Variable

O’Shannessy presents heat map data (from Hawkeye) showing that match winners consistently have superior court positioning — they play closer to the baseline, feeling what he calls “the magnetism of the baseline.” The average win percentage at the baseline is 46-47%. The average win percentage at the net is 65-66%. “We have this idea that the net is a bad thing,” but every approach to the net should be understood as moving from a coin-flip territory (baseline) to a significant-advantage territory (net). For the run-around forehand: “We give players labels — serve and a forehand — and it’s said with disdain. But what you’re saying is that player is winning more matches than anybody else.”

Actionable Advice for Families

• Evaluate your junior’s coach by asking directly: “What percentage of practice time do you spend on serve and return versus baseline consistency?” A coach whose answer is “mostly consistency” is training against the statistical reality of match play at all levels
• Watch one of your junior’s matches and count rally lengths — if most of their winning and losing is happening in four shots or fewer (it is, at every level), the skill priority in practice should be serve, return, and the third shot — not cross-court consistency drills
• Resist the framing that a player with a “big serve and a forehand” is incomplete — O’Shannessy’s data shows this player wins more points than any other, because tennis is a front-loaded sport
• Do not assume the tactical coaching for a daughter should differ from a son — the data shows men’s and women’s tennis analytics are nearly identical, and an opponent-specific game plan should be based on that individual’s patterns, not their gender

INTENNSE Relevance

• First-strike analytics as broadcast storytelling: O’Shannessy’s finding that 70% of points end in four shots is a broadcast frame that INTENNSE can use to educate fans in real time — the league’s shot-clock format and rally-scoring system already create compressed-point dynamics, and data displays showing “first-strike won percentage” by player and by bolt arc would give fans a new analytical vocabulary for what they’re watching
• Practice court alignment with match reality: INTENNSE’s coaching staff should audit practice designs against O’Shannessy’s framework — if the league’s players are spending practice time on consistency cross-court when match-winning happens in the first two shots, the practice court is misaligned; serve, return, and third ball should dominate INTENNSE practice design
• Gender parity in coaching: O’Shannessy’s finding that men’s and women’s analytics are nearly identical is a direct argument for INTENNSE’s mixed-gender coaching philosophy — coaches should develop tactical game plans for individual players based on their patterns, not by applying different frameworks to male and female players
• Net approach as positive-win territory: The 65-66% net point win rate vs. 46-47% baseline win rate argues for INTENNSE’s coaching staff to incentivize aggressive net approaches; the league’s team format with substitutions creates opportunities for doubles-style net players to enter matches specifically in net-charging scenarios
• Return of serve investment: INTENNSE’s one-serve format (if adopted) makes return-of-serve excellence even more critical — with no second serve to save players from poor first serves, the return becomes the decisive weapon; league coaching should disproportionately invest in return-of-serve work
• Data infrastructure: O’Shannessy’s 21 million data points from 700,000 points of tennis across junior-to-professional levels demonstrates that the analytics infrastructure for tennis now exists; INTENNSE should invest in a similar (smaller-scale) data collection and display system for its matches, enabling shot-by-shot rally length analysis and court positioning heat maps as broadcast features

Notable Quotes

“I’m not a stats person. I’m a strategy person. But I talk stats because that’s the language that quantifies it and removes the opinion and guesswork from coaching.”

“We’re spending almost 90% of our practice court practicing what happens only about 10% of the time in a match.”

“The number one rally length in tennis, by far — double the amount of the second highest — is a rally length of one. Serve goes in, the return does not come back.”

“If you win your match, over 90% of the time you have dominated your opponent in zero through four.”

“The longer a point goes, the more even it becomes. There is almost no advantage in the longer points.”

“How you hit the ball matters. Where you hit it matters more.”

“If you sent me a 12-year-old boy or a 12-year-old girl, I am coaching those two kids exactly the same way.”

“We say ‘all they are is a serve and a forehand’ — and it’s said with disdain, as an insult. But what you’re saying is that player is winning more matches than anybody else.”