#supercomputer – Under The Hammers

From Fragile to Manageable – West Ham’s Six Game Survival Path

The Under the Hammers Supercomputer projects the future.

West Ham’s survival maths has tightened again, but crucially, it’s tightened in our favour. The Under The Hammers Supercomputer, Deep Block, correctly called the Villa defeat and the Wolves win, and after that, and results elsewhere, the run-in now looks less like chaos and more like a plan. It’s in our own hands. We were right again, Villa away went as expected, Wolves at home had to be won, and was. Now at West Ham we have six games to save ourselves. And the pathway to survival is clearer than it’s been all season.

With six games now remaining, Deep Block has been rebooted once more. West Ham are still not safe, still not comfortable, but the survival picture is shifting from fragile to manageable. Since the last computer run, two fixtures have moved from projection into reality. West Ham lost 2–0 at Aston Villa and then thrashed Wolves 4–0 at the London Stadium. Both outcomes were in line with what the model expected, and that matters, because the difference between surviving and going down is often less about one surprising result, and more about reliably banking the points you must bank.

Villa Park was always low-probability for points, and so it proved. The 2–0 defeat was a blow at the time, but not a surprise, and not a result that rewrites the model. This was never the fixture West Ham’s season hinged on. You don’t build a survival plan around winning at a top-four side. You build it around what comes next.

That’s why Wolves mattered. And West Ham didn’t just win, we flattened them (at least in the second half we did!). In the original equation, Wolves at home sat in the “decisive” category, exactly the kind of game we couldn’t let slip. Those three points reduces uncertainty and reduces the number of “must-do” outcomes ahead.

The original model ran with twelve games to go. It predicted a draw at home to Bournemouth which was correct, a win at home to Wolves also correct, defeats at Liverpool, Manchester City and Villa, two correct and an unexpected free hit produced a point against City, and a draw at Fulham which turned out to be a win. We predicted that it was one of the away games where we could hope to turn a draw into a win which we duly did. So now we sit on 32 points, three points ahead of the original predictions, and two points above the bottom three with the run-in underway. We are last to play this weekend. Will we still be out of the bottom three when our game comes round on Monday?

Deep Block’s broad conclusion remains consistent with the previous run. The likely safety line sits around 38–40, which means West Ham are no longer chasing something outrageous. We are chasing a target that can be met with competent, repeatable outcomes rather than miracles. Weighting the final run-in correctly, home games like Everton and Leeds remain the true win fixtures, while Palace and Brentford away still profile as the likeliest draw opportunities, and perhaps even Newcastle too with their recent poor run. Newcastle’s recent form in the past 10 or 12 matches doesn’t match Leeds, Forest or West Ham. Statistical opportunities exist for wins in those three away games too. Deep Block’s value isn’t “perfect prediction”. It’s correctly identifying which matches are the ones that define our season, which are opportunities, and which are free hits.

1. Crystal Palace (A): Mon 20 April: Projection: Draw
Why: One of the away fixtures that still profiles as a realistic one point but with a potential to replicate what happened at Fulham. Palace have been on a run of winning every other game and drawing or losing the one in between. They won their last game, so a draw or defeat is likely if the pattern continues. They are stronger away from home, fifth in the away games league table, but just four home wins in sixteen games leaves them fifteenth in the home games table. Seven of their home games have ended in draws.

2. Everton (H): Sat 25 April: Projection: Win
Why: Previously identified as a decisive home match in the survival pathway. This will not be easy given Everton’s form with 10 points from their last six games. However, with just three home games to go, and one of those against the league leaders this game somehow has added significance for us, although Everton are one of several teams jostling for European qualification. They are strong away from home (fourth in the away games table) so we’ll have to be at our best as this is a game where we could slip up if we are not.

3. Brentford (A): Sat 2 May: Projection: Draw
Why: Still one of the likeliest away draw opportunities in the model’s framing. Despite their lofty seventh place in the league they have drawn their last four games. A realistic one point is achievable with a potential to snatch three, similar to the game at Crystal Palace. But Brentford have only lost three of their sixteen home games so no guarantees of any points here although the draw in the recent cup game gives us hope of another one.

4. Arsenal (H): Sun 10 May: Projection: Defeat
Why: A higher-difficulty fixture where the model doesn’t need points. The key is not letting it derail the plan. Despite their apparent faltering form they have still collected 13 points from their last six games. But if we have faltered ourselves in the previous 3 games then we might need something from this which will be tough.

5. Newcastle United (A): Sun 17 May: Projection: Draw
Why: Our final away game and despite their current form with four defeats in their last six games, it could still be another tough one. A point is possible and an unlikely win would ease the tension, especially if the previous results have not gone to plan.

6. Leeds United (H): Sun 24 May: Projection: Win
Why: Alongside Everton at home, Leeds was already identified as one of the defining fixtures. Who knows whether they will be safe at this point and ‘on the beach’ or still fighting for survival? It’s too far ahead but the belief is that home advantage will drive us over the line. Leeds are eighteenth in the away table so relegation form on their travels but this week’s win at Old Trafford has eased the pressure on them and the win in the cup here recently shows an upturn in away form.

Below is a projection consistent with the model’s earlier weighting that there are decisive home fixtures, that away draws provide the most plausible value and the best chance of exceeding expectation, and acceptance that a couple of games are likely free hits:

If all goes to plan the projected finish is 41 points. Two wins, three draws and one defeat: nine points from the final six games. The last six games have yielded eight points. Matching these lands us directly in the previously stated “likely safety” band, and it does it without demanding anything exotic. But it could all go wrong still. The predicted wins might not happen, and the draws could be defeats. But one away win in the last three away games could compensate for that. But a predicted win that becomes a defeat reduces the projection to 38. Will that be enough?

When we lost at home to Forest (perhaps controversially?) it didn’t look as though there was any way back and relegation seemed inevitable. The massive improvement since then has seen us climb out of the bottom three, keeping pace with Forest, and thanks to Tottenham’s decline moving two points ahead of our North London neighbours. But there are no easy games. We need our recent form to continue.

So let’s recap. The original model which ran with 12 games to go predicted a finishing position of 18^th place and likely relegation. The likely safety line at that point was predicted to be 38-42 points and that still stands. Of course, it is not just our own results that will decide our fate but also those achieved by Leeds, Forest and Tottenham.

How far do you like to go back to consider current form? Deep Block has some alternatives for you. Points gained in:

Last 4 games: West Ham 7, Leeds 6, Forest 6, Tottenham 1

Last 6 games: West Ham 8, Leeds 6, Forest 6, Tottenham 1

Last 8 games: West Ham 12, Leeds 10, Forest 7, Tottenham 1

Last 10 games: West Ham 15, Leeds 11, Forest 11, Tottenham 3

Last 12 games: West Ham 18, Leeds 14, Forest 15, Tottenham 3

However far you go back it looks good. BUT! We are not safe or comfortable. Will Tottenham will end their disastrous run of form? We hope not, but can they really continue to be so bad? Will there be a new manager bounce? Have Leeds turned the corner with their win at Manchester United? Are they already safe with a six point cushion? And what about Forest? Their fixtures aren’t straightforward but surely they will pull a further three points clear this weekend at home to Burnley?

The computer projections are based on current form, historical data, fixture difficulty and multiple other factors. But football is an unpredictable game. End of season games are perhaps even more unpredictable. Like Forrest Gump’s box of chocolates analogy, you never know what you are going to get. Football forecasting is not a perfect science. And this is nothing too serious. It‘s a bit of fun really. Anything can still happen. COYI!

The Under The Hammers Supercomputer runs again – and the margins tighten further as West Ham travel to Villa on Sunday

With eight games now remaining, Deep Block has been rebooted once more. West Ham’s survival equation continues to shift, and while nothing is settled, the picture is becoming clearer. Survival is still far from comfortable, but it is increasingly realistic.

Since the last run of the Under The Hammers Supercomputer, one more fixture has been converted from theory into fact, and crucially it was one that was previously written off as a free hit.

The latest result that was added to the model is last weekend’s unexpected 1-1 draw at home to Manchester City. That was predicted to be a defeat and a free hit. That point matters a great deal. The original projection assumed zero points from Manchester City. Instead, West Ham added one unexpected but fully deserved point, further improving the underlying survival maths and continuing the recent pattern of outperforming expectation.

Across the four games now completed since the first model was run, West Ham have taken 5 points, compared to an original projection of 2. That three-point swing can make all the difference.

Just as important as West Ham’s own result was what happened elsewhere. And nobody pulled away from us. Last week Leeds drew, Tottenham drew and Nottingham Forest drew. In fact all six teams at the bottom of the table drew. For the second week running, none of West Ham’s direct rivals managed to create separation. That collective stalling is increasingly significant.

The bottom of the table now reads:

• 15th: Leeds – 32 points
• 16th: Tottenham – 30 points
• 17th: Nottingham Forest – 29 points
• 18th: West Ham United – 29 points
• 19th: Burnley – 20 points
• 20th: Wolves – 17 points

West Ham are still just one point below the lower safety line of one point per game. With eight matches left, Deep Block updates the numbers as follows:

• Likely safety threshold: 38–40 points
• Points currently held: 29
• Points required: 9–11 from the final 8 games

That equates to roughly 1.1–1.4 points per match, a rate that now closely matches West Ham’s recent form rather than exceeding it. The current form (last 6 league games) for the teams under consideration with points per game in brackets:

• West Ham 9 (1.5)
• Leeds 6 (1.0)
• Forest 3 (0.5)
• Tottenham 1 (0.2)

Even extending ‘current form’ to eight games the results don’t differ much. Points per game (last 8):

• West Ham 1.5
• Leeds 0.88
• Forest 0.88
• Tottenham 0.38

The margins remain tight, but the task is no longer extreme. Next up we travel to Aston Villa on Sunday, a difficult task but not meaningless. The fixture remains firmly in the low-probability column based on the season as a whole. Villa sit in fourth place in the table but current form is less impressive with just five points from their last six games, four points below our total. But Villa at home are aggressive, intense and well-drilled. It won’t be easy.

The draw against Manchester City has bought us a small amount of breathing room. This is no longer a game that must deliver points to keep hope alive. Instead, it is an opportunity to add further unexpected value. A draw at Villa Park would be an excellent outcome. Even defeat does not materially damage the model, provided West Ham continue to take points in the games that follow. The pressure here is asymmetrical: Villa need to win to justify their position; West Ham can afford to be pragmatic, compact and opportunistic.

Perhaps one of the most important fixtures of the weekend from a West Ham perspective happens elsewhere when Tottenham host Nottingham Forest. This is the definition of a relegation six-pointer between two sides struggling badly for form and confidence. So what are we looking for? A draw is perhaps the best-case scenario for West Ham, freezing both rivals in place. A Forest win drags Tottenham fully into the danger zone. A Spurs win prevents Forest pulling away but keeps Tottenham within touching distance. Either way they both can’t win and that matters. Incidentally Leeds are at home to an inconsistent Brentford side who are seventh in the table and have collected nine points from their last six, losing just once, results comparable to our own.

Leeds are often thought to be strong at home but have only won six of their 15 games. Tottenham have won just twice. Forest, like West Ham have collected more points on their travels than at home.

What has changed since the original supercomputer projection? The pattern is now clear. West Ham are picking up points where none were expected. Rivals are failing to capitalise on opportunities to escape away from us. The psychological pressure has subtly shifted up the table. Deep Block originally projected an 18th-place finish on around 37 points. Based on real results, that projection now sits closer to 40 – and crucially, with a far wider range of viable pathways to get there.

The original conclusions still stand. Wolves, Everton and Leeds at home remain decisive. Palace and Brentford away remain the most likely draw opportunities. If that all happened then we’d reach 40 points. But the margin for error has widened. West Ham no longer need everything to go perfectly. In conclusion the situation is still fragile, but increasingly viable. West Ham are not safe, we are not comfortable, but we are alive, competitive, and very much in the fight. We didn’t really expect that a few weeks ago did we?

With eight games to go, Deep Block believes survival is now genuinely achievable rather than merely hopeful. It may yet go to the final day. It may yet hinge on a single moment. And if it does, well, this is the West Ham Way.

The Under The Hammers Supercomputer has been rebooted to re-run West Ham’s survival equation

With nine games to go the supercomputer has updated the maths behind West Ham’s chances of staying up. Survival that once looked desperate is now doable. We are not safe or comfortable but it’s closer than it looked a few weeks ago.

Since the Under The Hammers Supercomputer, Deep Block, last ran its projections (prior to the Bournemouth game) three of the twelve remaining fixtures have now been completed, replacing twelve hypothetical outcomes with more hard data to assess our survival chances.

Results from those games:

West Ham 0-0 Bournemouth (H) – Draw (predicted draw)
Liverpool 5-2 West Ham (A) – Loss (predicted loss)
Fulham 0-1 West Ham (A) – Win (predicted draw)

These results delivered 4 points from a possible 9, outperforming the original projection of two draws and a loss from these fixtures by two points. In particular, the away win at Fulham is a significant upgrade on the supercomputer’s expectations and materially improves the survival equation.

After 29 matches, West Ham now sit 18th in the Premier League with 28 points, level on points with Nottingham Forest in 17th but behind on goal difference. The teams in the relegation dogfight (the bottom six) currently read:

• 15th: Leeds – 31 points
• 16th: Tottenham – 29 points
• 17th: Nottingham Forest – 28 points
• 18th: West Ham United – 28 points
• 19th: Burnley – 19 points
• 20th: Wolves – 16 points

Crucially, the safety line is currently 29 points based on the season to date (exactly a point a game), meaning West Ham are now just one point behind that current safety level, with momentum firmly shifting in our favour.

With 9 games remaining, the survival target can now be recalibrated:

• Likely safety threshold at current performance levels: 38 points
• Points currently held: 28
• Points required: 10 from the final 9 matches

That equates to 1.11 points per match, which is lower than the rate required in the original model and less than West Ham’s recent form (1.33 per match in last six games).

Of course, results elsewhere can change the safety threshold and 38 might not be enough. But based on current figures Deep Block believes that a spread of 38-40 will be the level needed.

The upcoming schedule still presents major challenges, but the pressure has eased slightly thanks to the win at Fulham:

Manchester City (H) – Free hit
Aston Villa (A) – Low probability, but a point possible
Wolves (H) – Must‑win
Crystal Palace (A) – Likely draw but winnable
Everton (H) – Crucial and winnable
Brentford (A) – Likely draw
Arsenal (H) – Low probability
Newcastle (A) – Low probability
Leeds (H) – Potential season‑definer, perhaps must win?

The original identification of Wolves, Everton and Leeds at home as decisive fixtures to win remains valid. The difference now is that the margin for error has widened slightly.

What has changed since the original projection? The Fulham win is pivotal; it reduces the required points tally and shifts psychological pressure onto our rivals. Forest and Tottenham have failed to pull away allowing West ham to close the gap. Burnley and Wolves are effectively gone, leaving only one realistic relegation place still in play.

Deep Block originally projected a 37‑point finish and 18th place. Based on real results to date, that projection now reads closer to 39 points. Our survival chances have improved materially over the past three games. While relegation remains a genuine risk, the path to safety is now clearer and more achievable than when the original model was run.

If West Ham can beat Wolves, Everton and Leeds, and add two draws elsewhere (the games at Palace and Brentford are the most likely) as per the original forecasts then Deep Block predicts that Premier League survival is more likely than not. But the margins remain tight and depend upon results elsewhere. Failure to achieve the predicted eleven points from these games will likely need points gained elsewhere in the more challenging fixtures.

Assuming that Burnley and Wolves are as good as down (albeit not yet mathematically) the current form (last six games) of the four teams likely facing the final relegation place is as follows:

West Ham: 8 points
Leeds: 5 points
Forest: 3 points
Tottenham: 1 point

If that form continues into the final nine fixtures then we would be home and dry. But the games towards the end of the season tend to be unpredictable. Three games ago Deep Block predicted that we would probably just fall short. It now predicts that we might just be safe. It may yet come down to the final day. And if it does, well this is West Ham. We wouldn’t have it any other way.