Last updated 17 May 2026. This is the final preview before kick-off at 12:30 BST, and the picture from the data has not changed significantly since the last revision, which means the analytical conclusions remain intact and, if anything, more convincing with a day's worth of market movement behind them. Heracles host FC Groningen in Almelo in what is effectively the penultimate fixture of the Eredivisie season, and while the standings tell the broad story, the numbers underneath contain something the market appears to have undervalued.

Where the Teams Stand

The Eredivisie table after 33 matches is worth spending a moment on because it tells you a great deal about the structure of this fixture. The league leaders sit on 81 points from 33 games, which is a genuinely dominant season. Below that, the positions from second through to seventh are tightly contested, with 62 points down to 50 points covering six clubs. What matters for this fixture is understanding where Heracles and Groningen fit into that picture, and the data sheet's standings block, while anonymised at team ID level, allows us to map the context clearly: this is a mid-table fixture between two sides whose seasons are effectively determined, which historically affects how teams approach the final weeks both in terms of intensity and tactical shape.

The interesting thing is what the goals data tells us about the character of the game we are likely to see. The team at position 10 in this table, for example, has scored only 38 goals against 59 conceded across 33 matches, which is a goals-against total that suggests structural defensive vulnerability. Further down the table, two clubs are sitting on goal differences of minus 19 and minus 23 respectively after 33 matches. These are not close, hard-fought campaigns. Goals have flowed across the division this season, and that broader context matters when we think about what to expect on Sunday.

What the Model Is Saying

The SportSignals model gives Heracles a 31.3% probability of winning this match. The market at Unibet has priced that at 5.0, which implies a 20% probability. That is an 11.3 percentage point gap, which is the kind of edge that makes a bet worth considering seriously rather than reflexively. The draw no bet market reinforces the story: Groningen are available at 1.25 on the draw no bet, which means the market has enormous confidence in the away side. The home draw no bet is at 3.75, which is consistent with those match result odds.

Now I want to be honest about the confidence level here, because confidence in this signal is rated at 31 out of 100. That is not a high-conviction rating, and it matters. What it tells me is that the model has identified a genuine edge in the pricing but does not have strong underlying data to back a heavy stake. The sample size for Heracles at home in the final weeks of the season, combined with the absence of granular xG and pressing data in this dataset, limits how far I am willing to push this. The edge is real. The certainty is not.

The Goals Question

This is actually the more interesting analytical territory for this match. The model puts the over 2.5 goals probability at 63%, and the BTTS Yes probability at 62%. The market at bet365 has BTTS Yes at 1.44, which implies roughly a 69% probability after the overround is stripped out. That is a significant favourite, and it means the market has already priced in a high-scoring game to a degree that limits the value on the over side.

The under 2.5 signal from the model carries an edge of just 3.1 percentage points over the market, with a 37% model probability against a 33.9% market implied probability. That is marginal. The BTTS No signal is even thinner, with a 0.6 percentage point edge at 2.7 with BetVictor. What the data actually shows is a collection of signals that are directionally consistent but not individually compelling enough to act on in isolation.

The interesting thing is the first-half goals market. Bet365 has the under 0.5 goals in the first half at 1.01 and the over at 26.0, which is an extraordinary market shape suggesting the book is heavily weighted against a first-half goal. That does not mean there will not be one, but it reflects what the market has absorbed about these teams' tendencies in the opening 45 minutes. The second-half totals are far more balanced, with over at 1.12 and under at 1.03, which points to goals expected to arrive later in the game.

The Home Win at 5.0: Is It Worth It?

Let me be direct about this, because the headline signal is the Heracles home win and I do not want to bury the conclusion. At 5.0 and with a model probability of 31.3%, the expected value calculation is positive. If you ran this bet 100 times at this edge, the long-term return would be favourable. That is the mathematical case and it is a legitimate one.

The practical case is harder to make with full conviction. We have no form data in this dataset. We have no injury information. We have no confirmed lineups. What we have is season-long standings, goals totals, and model output, and on that basis the edge is identifiable but the information set is incomplete. Any bet on the Heracles home win should be sized accordingly, which in Kelly terms means a small stake relative to your usual unit.

The draw no bet on Groningen at 1.25 is effectively the market's base case. At 80% implied probability, you are paying for near-certainty on an away side in a fixture where the model says they win just under 50% of the time and Heracles win just over 30%. That remaining 20% or so is shared between the draw and other outcomes, which means the market is not wildly irrational but is pricing Heracles below their underlying probability to a degree that has value.

Final Assessment

This is a limited-data preview, and I will not pretend otherwise. The Eredivisie dataset for this fixture has no xG splits, no PPDA, no progressive passing numbers, no form strings, and no injury list. Those absences matter because they are precisely the inputs that separate high-confidence signals from speculative ones. What remains is a clean mathematical edge on the Heracles home win at 5.0 and a thin, borderline edge on the goals markets that does not reach my threshold for a recommended bet. The home win is the only signal I would consider acting on, and only at a reduced stake relative to normal, because the edge is real but the supporting evidence base is narrow.