Braintree Table Tennis League: A 92 Per Cent Record Standing Next to a Twelve-Year-Old

Braintree Table Tennis League: A 92 Per Cent Record Standing Next to a Twelve-Year-Old

In the Braintree season dataset I reopened late one night, one line made me pause longer than all the rest: Dave Fiddeman, 92 per cent. Immediately beneath it sat another line: Ethan Collins, twelve years old, three cadets' titles and one junior boys' title.

Those two lines sit inside the same league system, the same season, the same division. The distance between them is twenty-eight years of playing experience, and a data gap no table at this level can fill.

I have spent seventeen years reading tables like that. Seven of those years have been tied to table tennis, the rest to football and match-outcome models. The earliest lesson I learned is also the one I have to remind myself of most often: a percentage without a denominator, without a reference opponent and without divisional context is just a number standing alone in a room.

That is precisely why it is worth reading.

Context: a league with no cameras

Braintree Table Tennis League is a local competition based in Essex, England, operating under Table Tennis England, the national governing body for table tennis in England. This is a community league. There are no ITTF ranking points, no prize money, no television contracts, no expected-goals tables or pressures-per-pass indices.

The operating structure is simple. Teams are sorted by division. Wins and losses decide promotion and relegation. Squad lists are flexible from match to match: a team may register many players and rotate them. This is the single most important thing to remember before reading any number below.

The way the win rate is calculated also needs spelling out. In the standard English local-league team format, each side fields three players and each player meets every opponent once. One team tie therefore produces nine individual matches. The percentage quoted in the preview is a player's wins divided by total individual matches played across the season. It does not distinguish a 3-0 win from a 3-2 win, a strong opponent from a weak one, or a match that mattered for the final standings from one that did not.

At the elite level, table tennis data is dense: service points won, receive success rate, counter-loop exchanges per five rallies, pressure indices. At Braintree, the only tool is a fraction.

A local-league win rate blends three different things: technical ability, psychological consistency and the quality of opponents faced. I call it a three-unknown equation with one line of algebra. You cannot solve it. You can only take notes and wait for more data.

When the hall is empty, the data sits and weeps alone. In the village halls and community sports centres of Braintree, the stands are usually bare. No shouting, no media pressure, no international officials. That means the data here is cleaner in terms of emotional noise, but thinner in terms of signal. A player competing in front of twenty people reacts differently from one competing in front of two thousand, and no column records that difference.

There is a personal reason I follow leagues like this. In 2026, at a major continental tournament, I noticed an eighteen-year-old midfielder because his passes into the final third outnumbered those of two far more celebrated players. He did not stand out on television. He stood out in the spreadsheet. The lesson I carried into table tennis is this: the players worth tracking usually appear first in a column, not in a commentary line.

Division two: Black Notley B and the logic of a comeback

The team rated highest in division two this season is Black Notley B. The reason lies in three names and three percentages.

Neil Freeman scored 60 per cent in division one last season. Rev Matthews scored 86 per cent in division two. Steve Kerns, a former men's singles champion, will appear in roughly half the team's matches.

Read those three lines together and a clear structure emerges. Freeman is the anchor. A player who scored 60 per cent at a higher level, dropping into division two, faces a lower average opponent. In my model, that is a variable with a positive adjustment coefficient. If form holds, the probability he exceeds 75 per cent in division two is high. I am not stating that as certainty. I am saying that if you had to bet on a single variable, that variable is downward divisional movement.

Matthews, at 86 per cent in division two last season, offers a larger and more stable sample. A player who wins almost nine of every ten matches across a local season rarely has an obvious technical weakness. They have something else: the ability to win matches they are not playing well. That is the hardest skill to measure and the most decisive one at this level, because at this level more matches are won by limiting errors than by pure technique.

Steve Kerns is the third variable, and the most structurally interesting. He plays roughly half the matches. In a league where every individual match adds to a team total, having a former men's singles champion for half a season is not a linear addition. It is a localised spike. The team is stronger in the rounds he plays and returns to baseline in the rounds he does not.

That creates a form of noise the standings never display. A team whose fixtures align with Kerns' appearances will have a completely different impression of Black Notley B from a team whose fixtures do not. I have seen this pattern many times in football data: a key player appearing in too few matches distorts an entire predictive model, and people usually blame the model rather than the input.

That Black Notley B were relegated last season makes the structure more notable. A relegated side has two options: rebuild and go young, or load up and bounce straight back. With Freeman, Matthews and Kerns, this team chose the second path. It is a sensible short-term probability decision and a questionable long-term one. But a season lasts only a season.

Division two: Sudbury Strollers and the depth problem

Sudbury Strollers finished second last season. This year they are treated as the main challenger to Black Notley B.

Dave Fiddeman scored 92 per cent last season. John Colvin scored 75 per cent. Taken in isolation, those two numbers would put any team in the leading group.

But one sentence in the preview stopped me: Sudbury Strollers' fate may depend on who backs them up and how often.

This is a phrase a data analyst has to read slowly. It indicates the team does not have a fixed three-player line-up. They have two high-quality players and a gap.

In my model I label this structure variance skew. A team with two players at 90 per cent and a third fluctuating around 50 per cent will have a better average score than a team of three players at 70 per cent, but a far higher variance. Over a long season, high variance is a disadvantage. In a decisive round of fixtures, high variance is a fatal risk.

That is why I place Sudbury Strollers in the challenger group rather than the favourites, despite the 92 per cent. The number does not lie; it only keeps secrets. And the secret here is the third player.

It is worth noting that Fiddeman's 92 per cent and Matthews' 86 per cent sit in two different teams in the same division. The two figures are broadly comparable, with one caveat: they say nothing about how often these two will face each other, or who would win if they did. In local leagues, direct meetings between the two strongest players of the two strongest teams usually occur once or twice a season. The sample is too small for conclusions.

I would still advise readers to note the dates of those meetings. Not to predict, but to build a small, clean dataset for future seasons.

Division three: where the data is thinnest

Division three has a more complex structure and less reliable data. Finchingfield B finished second last season. They lose Lucien Nolan-Bradford, who went through division three last season with exactly one defeat.

In compensation they have Dave Punt, moving down from division two. And they have Ray Nolan-Bradford, most likely Lucien's father, remaining in the squad. That presence preserves a club connection and a form of cultural stability that no table records.

In club-level sport, the family factor is an undervalued variable. An older player who stays with a team after a relative moves on does not merely hold a place in the line-up. They hold a training standard, a way of behaving in the hall, a memory of how the club used to operate. None of that appears in any column, but all of it influences the results of rounds in which the team is short-handed.

The most notable feature of division three is the arrival of a new team: Black Notley F. In a club-level league, a club being able to field an extra team says two things. First, that club has enough registered players to fill a new line-up. Second, it has enough organisational capacity to run an extra team all season.

In my data on local leagues, the number of teams a club registers predicts survival far better than the record of that club's strongest team. A club with three teams outlives a club with one star side. That is structure, not form.

Black Notley F are noted as having players who impressed on debut. Impressed on debut is one of the phrases I read most carefully. It rests on direct observation, not aggregated data. An impressive debut can be the start of a good season, or the high point of a player who has never competed under pressure. I keep both possibilities open.

The preview also notes that Finchingfield B could be stretched by Black Notley's new F team. That is a claim about squad depth, and I agree with it as a hypothesis. Two teams at the same club sometimes share a pool of reserve players. If Black Notley F need a body, they can draw on a source Black Notley B are not using. Finchingfield B have no equivalent source.

The pipeline: four names and a schedule

What makes this preview different from an ordinary league table is the number of junior players named in adult divisions.

Ethan Collins, twelve years old, already has three cadets' titles and one junior boys' title. I spent time reviewing Table Tennis England's age-band structure. The cadet band covers players under thirteen. The junior boys' band is wider. A twelve-year-old winning titles in both bands within the same period suggests he is competing at the top of his age group, or beyond it.

But I must state the limits of this data. Four age-group titles are a signal, not a forecast. In table tennis, the distance between twelve and eighteen runs through three biological variables: growth in height, forearm strength development, and tolerance for training volume. A player who dominates at twelve through technique and reflexes can be caught at fifteen by someone who matured physically earlier.

I have no data on those three variables. I only have results.

Sai Suresh, fourteen, and Aryaman Singh, thirteen, will make their debuts for Rayne D. The preview uses the word baptism. That is the language of direct observation, and it describes the situation accurately: the first time these two face an adult standard in an official competition.

Both are under the eye of Keith Martin, the league coach. The existence of a league-level coach is a significant fact. It means this local system does not merely organise matches; it operates a development structure. That is the difference between a league that survives and a league that grows.

JJ Calisin, eighteen, is the fourth case and arguably the one with the clearest data. He is scheduled to move up to division one at Christmas. That is a structured decision: the club or league is not waiting until the end of the season to assess him, but setting a mid-season checkpoint.

I care about this model because it creates a natural experiment. If Calisin succeeds in division one from Christmas, the internal evaluation system read the player correctly. If he fails, it read him too optimistically. Either way, we have data to compare against the prediction, which most decisions in sport never allow.

One final structural point: six junior players named in a preview of a few hundred words is a high proportion. In the professional previews I read, junior players usually appear as a footnote. Here they are part of the main story. That suggests the system is at the beginning of a generational transition, and the people writing the preview know it.

The 16-14 match: the black hole of a scoreline

Lucien Nolan-Bradford went through division three last season with exactly one defeat, to Ben Southgate, 16-14 in the fifth game.

This is the only line in the entire preview describing a specific moment of a specific match. It is extremely valuable for three reasons.

First, 16-14 in the fifth game means the match went to the limit. In table tennis a game normally ends at eleven points. For a game to reach 16-14, two players must be so evenly matched that each had at least one chance to close it out and failed. In my probability model, a game stretching to 16-14 is equivalent to two players holding near-identical win probability at that moment.

Second, the event sits inside a season in which Nolan-Bradford lost exactly once. That means Southgate was the only player who found a way to beat him, and found it in the most stressful possible circumstances.

Third, Southgate scored 87 per cent last season and is moving up a division. He is a shifting variable.

I must be clear that I lack the data to call Southgate Nolan-Bradford's nemesis. One match is one match. The denominator is one. But in a data-poor system like a local league, a match like that is an anchor point. It tells me these two players shared a similar level at a specific moment, and that any future meeting between them deserves attention.

Data cannot save a match, but it points to why the match died. Here, data could not save Nolan-Bradford from his only defeat, but it shows the defeat was not because he was weaker. It shows he met the right player on the right day.

There is a category of information a scoreline always erases: information about the points that were never written down. A 16-14 game contains at least thirty rallies in its closing phase. Each rally is a decision about serve, about position, about return direction. None of those decisions are archived. We know only the final outcome, and the final outcome cannot tell us which decision produced it.

The contrarian angle: a percentage without a denominator

Here I have to say something many readers will not want to hear.

Dave Fiddeman's 92 per cent in division two and Neil Freeman's 60 per cent in division one cannot be compared directly. More importantly: 60 per cent in division one may have a higher absolute value than 92 per cent in division two.

This is the problem of divisional calibration, and it is one of the most common errors in low-tier sports analytics. A win rate does not automatically reflect quality. It reflects the ratio between a player's ability and the average ability of the opponents that player meets.

In football, I once analysed the 2026 World Cup semi-final between France and Belgium. France took significantly fewer shots, but their expected-goals figure was clearly higher. The shot-count number told one story. The chance-quality number told another. The average reader looks only at the first.

In local table tennis the problem is worse, because there is no quality metric equivalent to expected goals. We do not know whom Fiddeman faced in his 92 per cent. We do not know how many of those wins came in a fifth game. We do not know how many came from behind.

I have followed table tennis and football long enough to know that a player winning 92 per cent mostly through 3-0 wins against weaker opponents is a different player from another winning 92 per cent. Another player might win 60 per cent with every win coming 3-2 against the strongest opponents in the division.

Those two players would have the same win rate at the same divisional level. They do not have the same value.

This is why any analysis built only on win rate at this level must come with a warning. I am writing that warning here, in plain words: the numbers below are signals, not conclusions.

A second contrarian angle concerns the junior players. The reader's instinct is to translate Ethan Collins' four age-group titles into a linear forecast. I have seen this pattern in football: a young player with an impressive youth record, and immediately every projection assigns him a first-team place. Most of those projections are wrong, not because the player lacks talent, but because human development curves are not linear.

In table tennis that curve breaks even more sharply. Technique at twelve is built on a body that will change completely over the following four years. My point is not that Collins will fail. My point is that we have no data to say anything about him at eighteen.

A third contrarian angle concerns the preview itself. A season preview is a document about expectations, not results. It is written by people who have observed the league for years, and it carries accumulated bias. When the preview calls Black Notley B the team to beat, that is a judgement based on a list of names, not on a probability model. I am not disputing it. I am placing it in the correct cell of my spreadsheet: this is the initial hypothesis, and every result from round one onward will test it.

The limits of data: a lesson from a season without crowds

There is a period in my career I return to often when analysing data-poor systems.

In 2026, when a European football league restarted after the pandemic, my prediction model went badly wrong. Home win rates fell from around 45 per cent to 38 per cent across a run of matches played without crowds. Five years of historical data became useless because the crowd variable had never been built into the system. I delayed publishing the report and eventually accepted releasing a revised version with an adjustment coefficient for home advantage.

The lesson I drew, and have applied to every analysis since: a variable missing from the data may still be operating.

In the Braintree League, what is that variable?

I propose three candidates.

The first is table and ball conditions at the venue. Community halls in Essex do not maintain the uniform table standards of professional arenas. A player who trains on a fast table is disadvantaged on a slow one, and vice versa. Nobody records this index.

The second is travel distance. English local leagues generally require teams to travel between towns within a region. A team with away fixtures clustered in one period is affected differently from a team with an even spread. Nobody records this index either.

The third is player availability. The preview states plainly that Steve Kerns plays roughly half the matches. But there is no data on which teams meet him in that half. With luck, a team might never face Kerns all season. With bad luck, they face him twice.

None of these three variables appear in any statistical table. All of them can decide final positions.

Every number is a recitation, every calculation a meditation. But a good analyst must accept that some things cannot be recited, because nobody has written them down.

There is one more skill worth mentioning, and it is the one data never touches: the ability to read an opponent within a specific match. A strong club-level player often wins not because the loop is harder, but because they notice after two games that the opponent struggles when pushed to the left, or that the opponent loses composure after two consecutive lost points. Those realisations leave no trace in any recording system. We only see the consequence: a higher percentage.

What I will be tracking in the early rounds

I never make unconditional predictions at this level. Instead, I build a checklist of signals to verify.

The first signal is Neil Freeman's win rate over his first five matches in division two. If he stays above 70 per cent, the divisional-drop hypothesis is confirmed. If he sits between 50 and 60 per cent, the gap between division one and division two in Braintree is narrower than I assumed.

The second signal is Steve Kerns' actual number of appearances in the first half of the season. The phrase roughly half needs quantifying. In my data, estimates of this kind tend to drift optimistic.

The third signal is Dave Fiddeman's performance against the strong teams. A 92 per cent overall figure can conceal a far lower rate against Black Notley B.

The fourth signal is Rayne D's first results with Sai Suresh and Aryaman Singh. Not the win-loss outcome, but the game scores. A junior losing 0-3 with every game reaching 9-11 is in a different position from a junior losing 0-3 with games ending 3-11.

The fifth signal is the division three standings at Christmas, when JJ Calisin moves up to division one. That move will leave a gap in his old line-up and create a test in his new one.

Do not ask the data what the future holds; ask what the past is reminding you of. Braintree's past reminds me that the teams who win these divisions are usually not the ones with the best players, but the ones whose best players appear most often.

The three teams most discussed in the preview, Black Notley B, Sudbury Strollers and Finchingfield B, share one structure: one or two leading players and an open question about the rest. That is the structure of almost every club-level league I have ever read data on.

What makes this season worth following is not who wins. It is that we can watch a generation of junior players enter the adult system and compare the initial forecast against the actual results. In a sport where every professional dataset concentrates on the world's top twelve players, having a record-keeping system good enough to track twelve- and thirteen-year-olds in Essex is rare.

I do not remember matches; I remember why they happened the way they did. For the Braintree League, the reasons will only surface after about ten rounds, when the first percentage figures finally have enough of a denominator to say something meaningful.