Is Spurs' New Found Form Set to Continue? What the Data Tells Us

Two wins in two games. A first league victory of the season against Coventry, followed by a remarkable 2–0 win at Stamford Bridge. Suddenly, things are looking a little brighter for Tottenham Hotspur.

And not before time. Six league games without a win had raised the prospect of another season at the wrong end of the table. After two consecutive 17th-place finishes, optimism around Roberto DeZerbi was already looking premature.

But Chelsea was huge: not just three points, but a rare victory at a ground where Spurs have won only twice in 35 Premier League visits. Pessimism is slowly turning to optimism. Will it last?

With football's weekly news cycle, only the last result ever seems to matter. It's important not to overreact, in either direction. So what does the data suggest?

Read on for the detail and subscribe here if you'd like content like this free into your email twice per week.

Are Spurs Actually in Form Yet?

Three unbeaten sounds great, but context matters. A draw at Old Trafford is historically massive. But this is that Manchester United. Nevertheless, salvaging a point with ten men deserves credit.

Coventry was a good win. Yes, they're newly promoted and struggling. But how often have Spurs failed in games like this? Last season they lost at home to West Ham and Fulham and couldn't beat Wolves. After two 17th-place finishes, Tottenham haven't earned the right to expect three points against anybody.

Then came Chelsea. An away win against a fierce rival, in a fixture that has historically brought so much disappointment, was enormous for momentum. More importantly, Spurs looked the better side: nearly 69% possession, 15 shots to Chelsea's seven, and just 0.43 expected goals (xG) conceded, according to Sky Sports.

Cole Palmer's late red card helped, of course. But was this performance a one-off?

What Did the Model See Against Chelsea?

I run Fantasy Matchup Edge, a football analytics project producing individual projections. It’s mainly for fantasy football enthusiasts, but the numbers can still tell us something about the wider game.

Rather than primarily predicting results, the model examines the actions players produce, what opponents allow, and the match environments that create those opportunities. It runs 100,000 simulations to estimate individual output, conditional on a player starting.

Before Chelsea faced Spurs, their attackers weren't projecting particularly strongly. With João Pedro unavailable, only Morgan Rogers made the model's top 50 for the Gameweek. Rogers had a 31% chance of scoring and a 22% chance of assisting; Cole Palmer's assist probability was just 5%. None happened.

That isn't a claim the model predicted a Spurs victory. It suggested Chelsea's key players lacked favourable individual matchups - and their seven shots and 0.43 xG offered some supporting evidence.

Spurs' own defensive numbers have quietly been encouraging: 9.9 xG conceded through eight matches, the league's fifth-lowest total, according to FotMob data compiled by Planet Football.

But There's an Uncomfortable Question...

Chelsea's players weren't especially fancied by the model. But neither were Spurs'. Omar

Marmoush was Tottenham's highest-ranked player before the derby, at just 87th league-wide.

Unfortunately, it's a familiar story. On 5 September, amid Spurs' dreadful start, I replied to an Opta Analyst post about their struggles to get into dangerous areas against Nottingham Forest:

“Their players aren't good enough. Only got 2 in the model's Top 100 GW3 projections (Marmoush and Senesi). For reference, Brentford have 4 in the Top 25.”

Was that slightly sensationalist, in a futile attempt to drive traffic to my fledgling project? Probably. But the thought remains: could the problem be the players?

Across many fixtures, Spurs have struggled to produce highly ranked attacking players. And their season-long numbers hardly contradict that. FotMob places Tottenham 19th for xG created after eight matches, on just 8.2. Not what you'd expect from a club with their ambitions.

There's a caveat: a fantasy projection measures expected output in a particular matchup, not

absolute player ability. New signings may still be finding their feet. And it's football - no model gets every prediction right. Mathys Tel ranked 266th that Gameweek, with a 15% scoring chance. Yet he scored against Chelsea.

Trends are easier to spot than exact individual performances.

So, Can Spurs Extend Their Run?

The encouraging news is that Tottenham's numbers are starting to change. Good performances feed gradually into the model, and the fixtures offer opportunities.

Crystal Palace come next. They've conceded 15.8 xG in eight matches, the second-worst record in the league. It's a positive matchup for Marmoush, ranked 14th across the league in FME's Gameweek 9 projections: 5.28 expected FPL points if he starts, a 40.6% scoring chance, and a 14.7% chance of a ten-point haul (usually a good barometer of an incredibly strong performance with at least two goals and assists).

The matchup evidence helps explain it. Palace allow 47% more shots than the league baseline, while Tottenham's shot supply is 12% above it, according to FME. Van Hecke, Sávio and Kudus also now make the top 100, conditional on starting.

That's a welcome change from September. But they are playing Crystal Palace. What about

afterwards?

A Better Run - And a Bigger Question

Home matches against Ipswich and Fulham feature among Tottenham's next five. The model likes their attacking prospects in those fixtures, while the defence has generally looked steadier than the attack anyway. It’s up front where Spurs need to see an improvement, and 3 of their next 5 fixtures have strong opportunities for goals.

And then there's Xavi Simons. If he returns from his knee injury and recovers his form, he could make a real difference. The model currently rates him Tottenham's best projected player over the next five Gameweeks: 24.63 expected FPL points if he starts all five, ranking 26th across the league. Clearly this won't happen; it's a hypothetical, not an availability forecast. But what I’m saying is that this is where the data has him, if he was fit and available: Spurs’ best projected player.

Yet shouldn't Spurs expect their best player to rank a little higher than 26th? It's also disheartening to see Richarlison ranked 33rd while he's been playing with the Under-21s. But we won’t get into that now.

Has Something Really Changed?

There's reason for cautious optimism: three positive results, stronger defensive numbers than the league position suggests, and fixtures that could build momentum. The individual projections are improving too.

But there's a difference between becoming competitive and truly turning a corner. The season's attacking numbers remain poor, and two wins don't erase six winless games.

If Spurs really have changed - if the new players gel and De Zerbi's ideas take hold - the model will eventually reflect it. Perhaps Chelsea was the start of something. For now, Spurs have earned something that seemed in short supply a fortnight ago: optimism.

Whether they've earned our complete confidence is another matter.

Gerry is the creator of Fantasy Matchup Edge, an independent football analytics platform producing projections and matchup analysis for Fantasy Premier League and Sorare. He enjoys uncovering the stories behind football's numbers - and occasionally discovering how wrong his model can be.