# Chelsea vs Real Sociedad > Club Friendlies 1 · Kickoff Sat 15 Aug 2026, 13:00 UTC · [Canonical HTML](https://betsprinter.com/fixtures/38286) **Status:** Scheduled ## Model verdict - **Chelsea win:** 26% - **Draw:** 43% - **Real Sociedad win:** 31% - **Source:** model ## Pre-match deep dive ### Draw-heavy model prices a cautious, low-tempo friendly clash ## The stage This is a pre-season club friendly kicking off Sat 15 Aug 2026, 13:00 UTC, listed as Club Friendlies 1[^fact-1]. The fixture’s setting is primarily preparatory rather than competitive, and the model reflects a market that prices a deadlock as the most likely outcome[^fact-2]. The Elo framework used for context grants Chelsea a +90-point edge with home advantage applied, a substantive but not overwhelming quality gap[^fact-3]. ## Form & momentum Chelsea’s recent run reads LWLDL across the last 10 matches — three wins, one draw, six defeats — producing 1.00 points per game and averages of 1.30 goals scored and 1.50 conceded per match[^fact-4]. Real Sociedad arrive in colder shape: DLDDL in their last 10, a 1-5-4 record with 0.80 points per game and means of 1.60 goals scored and 1.90 conceded per match[^fact-5]. On pure form metrics Chelsea show a modest attacking edge in goals scored per match, while Sociedad’s conceded rate is slightly worse, but the model still prefers a draw as the single likeliest outcome[^fact-4][^fact-5][^fact-2]. The Elo differential favors the home side despite their mixed results, suggesting the quality gap underpins Chelsea’s profile here[^fact-3]. ## Personnel Chelsea’s standout in recent appearances is Enzo Fernández: two goals and one assist in his last five outings, with an average match rating of 7.26[^fact-6]. That finishing and influence has been a clear offensive highlight for Chelsea amid a tepid goals-scored average[^fact-4][^fact-6]. The heaviest absence for Chelsea is Robert Sánchez, sidelined by injury and responsible for 786 minutes in the recent run — a notable loss in continuity at a position that matters for clean-sheet stability[^fact-8]. For Real Sociedad the most productive creative spark has been Sergio Gómez, who has delivered no goals but three assists in his last five appearances and carries an average rating of 6.96[^fact-7]. Sociedad’s principal recent injury absence is Arsen Zakharyan, who logged just 65 minutes in the recent run before going out — a limited sample but nevertheless the club’s named missing piece[^fact-9]. Both sides therefore come with meaningful personnel caveats: Chelsea missing a regular shot-stopper who contributed nearly the equivalent of nine full matches, and Sociedad missing a player who featured sparingly but is explicitly listed as absent[^fact-8][^fact-9]. ## Where the model sees value The model’s probability split is Home 26% / Draw 43% / Away 31%, and that distribution was tested against two markets in the analysis[^fact-2][^fact-10]. The clearest structural call is the model’s overweighting of the draw relative to home and away outcomes, which reflects both teams’ middling recent form metrics and the nature of a friendly fixture where rotations and low intensity can compress result variance[^fact-4][^fact-5][^fact-1]. Against the markets examined, the model flags the draw as the primary edge — the single most likely result on the model’s board — while still assigning a non-trivial probability to an away victory driven by Sociedad’s capacity to nick results despite their poorer point-per-game return[^fact-2][^fact-5]. The Elo differential of +90 points for the home side tempers that view, keeping Chelsea competitive in expected outcome despite the model’s draw lean[^fact-3]. Finally, the two markets compared in the review provide the reference points that reveal these divergences between consensus pricing and the model’s probabilities[^fact-10]. ## Verdict The model leans to a draw as the likeliest single outcome in a low-intensity friendly: Chelsea hold the quality edge on Elo but carry mixed recent form and a key goalkeeper absence, while Real Sociedad arrive with fewer points per game and their own notable absences — together these inputs congeal into a high-probability stalemate on the model’s board[^fact-3][^fact-4][^fact-8][^fact-5][^fact-9][^fact-2]. ### Cited facts [^fact-1]: **Kickoff** — Sat 15 Aug 2026, 13:00 UTC — Club Friendlies 1 [^fact-2]: **Model verdict** — Home 26% / Draw 43% / Away 31% (source: model; confidence mid, 12 pp gap to runner-up). [^fact-3]: **Elo edge** — CHE vs RSO — Elo differential +90 points (with home advantage applied). [^fact-4]: **CHE recent form** — LWLDL last 10: 3-1-6 (W-D-L), 1.00 PPG, 1.30 goals scored / 1.50 conceded per match. [^fact-5]: **RSO recent form** — DLDDL last 10: 1-5-4 (W-D-L), 0.80 PPG, 1.60 goals scored / 1.90 conceded per match. [^fact-6]: **CHE in-form player** — Enzo Fernández — 2 goals, 1 assists in last 5 appearances, avg rating 7.26. [^fact-7]: **RSO in-form player** — Sergio Gómez — 0 goals, 3 assists in last 5 appearances, avg rating 6.96. [^fact-8]: **CHE key absence** — Robert Sánchez out (injury), 786 minutes in recent run. [^fact-9]: **RSO key absence** — Arsen Zakharyan out (injury), 65 minutes in recent run. [^fact-10]: **Markets analysed** — 2 market(s) compared against the model. --- Methodology: . Canonical HTML: .