🏁 Market Master Analysis: Phase 1 Review

It is fantastic to see the 1X2 model find its footing in Week 16. Hitting 70% in a major league like the Premier League is indeed a massive achievement for a single gameweek—that is well beyond statistical noise.

The "Man Utd 4-4 Bournemouth" result is the definition of a "Black Swan" event. No statistical model based on sane historical data predicts an 8-goal thriller draw easily. It’s frustrating for the ticket, but as a data point, it’s an outlier you generally have to forgive the model for missing.

On the flip side, the Sunderland victory is the highlight. That is a "Model's Choice" pick—where the data sees something the general public (and bookmakers) might have underestimated.

Here is the deep-dive analysis on your total hitrates for the first 4-week block.

📊 1X2 Market Analysis (Phase 1)

Total Hitrate: 52.5% (21/40 games)

Performance:

In a 3-way market (Home/Draw/Away), the baseline for random chance is roughly 33%. Sustaining a 52.5% hitrate over 40 games is statistically respectable.

The Profit Discrepancy: Despite a >50% hitrate, the Total P/L is -1.57u. This indicates that while the model is picking winners more often than not, it has likely been leaning on "heavy favorites" with short odds (e.g., 1.40 - 1.60) in previous weeks. When those fail, they wipe out multiple small wins.

The Week 16 Shift: Week 16 was the turning point. The high hitrate combined with a solid +3.53u profit suggests the model finally identified value picks (like Sunderland or Away wins for Fulham/West Ham) rather than just "safe" picks.

📉 Over/Under Market Analysis (Phase 1)

Total Hitrate: 52.5% (21/40 games)

Performance:

In a 2-way market, 50% is essentially a coin flip. To be profitable (beating the bookmaker's margin/vig), you generally need a hitrate of roughly 52.4% to 54% depending on the odds.

The Leak: While the hitrate (52.5%) is technically identical to the 1X2 model, the P/L is significantly worse (-3.53u). This suggests the model is consistently losing "value" on the O/U lines. It is winning bets with low payouts and losing bets that might have had standard odds.

Conclusion: The O/U component currently lacks an "edge." It is reacting to market averages rather than predicting game flow.

🛠️ Strategic Outlook for Phase 2

The first 4 weeks (40 games) have provided a clear baseline.

1X2 Model: It is trending in the right direction. The 70% spike in Week 16 suggests the logic is sound, but it needs to maintain the courage to pick value underdogs (like Sunderland) to fix the overall negative P/L.

O/U Model: This needs the most work. It is currently "treading water."

Tickets: The accumulator strategy is currently the biggest drain on the bankroll (-€20.98 total net). The variance in the Premier League (like a 4-4 draw) makes hitting 6-fold accumulators incredibly difficult.

🔍 Matchweek 16 Details

Match	Score	1X2 Pick	Res	O/U Pick	Res	CS Pick	Res
Chelsea v Everton	2-0	Home	✅	Over	❌	2-1	❌
Liverpool v Brighton	2-0	Home	✅	Over	❌	2-1	❌
Burnley v Fulham	2-3	Away	✅	Under	❌	0-1	❌
Arsenal v Wolves	2-1	Home	✅	Over	✅	2-0	❌
Crystal Palace v Man City	0-3	Away	✅	Over	✅	1-2	❌
Nott'm Forest v Tottenham	3-0	Draw	❌	Over	✅	2-2	❌
Sunderland v Newcastle	1-0	Home	✅	Over	❌	2-1	❌
West Ham v Aston Villa	2-3	Away	✅	Over	✅	1-2	❌
Brentford v Leeds	1-1	Home	❌	Over	❌	2-1	❌
Man United v Bournemouth	4-4	Home	❌	Over	✅	2-1	❌

📊 Model: Week 16 (1u Flat)

Model	Hitrate	Profit	ROI
1X2	70%	+3.53u	+35.3%
O/U	50%	-1.40u	-14.0%
CS	0%	N/A	N/A
TOT		+2.13u

📈 History per Week

Week	Games	1X2%	1X2 P/L	O/U%	O/U P/L	CS%	CS P/L	Total P/L
13	10	50.0%	-0.96u	50.0%	-1.06u	0.0%	0.00u	-2.02u
14	10	40.0%	-2.91u	70.0%	+2.52u	10.0%	0.00u	-0.39u
15	10	50.0%	-1.23u	40.0%	-3.59u	0.0%	0.00u	-4.82u
16	10	70.0%	+3.53u	50.0%	-1.40u	0.0%	0.00u	+2.13u
TOT	40	52.5%	-1.57u	52.5%	-3.53u	2.5%	0.00u	-5.10u

🎫 Ticket Overview (Week 16)

❌ 1X2 Ticket | Profit: €-2.20

Match	Selection	Res
Chelsea v Everton	Home	Won	✅
Liverpool v Brighton	Home	Won	✅
Man United v Bournemouth	Home	Lost	❌
Brentford v Leeds	Home	Lost	❌
Nott'm Forest v Tottenham	Draw	Lost	❌
Crystal Palace v Man City	Away	Won	✅

❌ O/U Ticket | Profit: €-2.20

Match	Selection	Res
Liverpool v Brighton	Over	Lost	❌
Man United v Bournemouth	Over	Won	✅
Crystal Palace v Man City	Over	Won	✅
Burnley v Fulham	Under	Lost	❌
Brentford v Leeds	Over	Lost	❌
West Ham v Aston Villa	Over	Won	✅

❌ CS Ticket | Profit: €-1.10

Match	Selection	Res
Arsenal v Wolves	2-0	Lost	❌
Burnley v Fulham	0-1	Lost	❌
Crystal Palace v Man City	1-2	Lost	❌
Brentford v Leeds	2-1	Lost	❌

💰 Financial Summary per Week

Week	1X2 P/L	O/U P/L	CS P/L	TOTAL
13	-2.20€	-2.20€	-1.10€	-5.50€
14	-2.20€	-1.18€	-1.10€	-4.48€
15	-2.20€	-2.20€	-1.10€	-5.50€
16	-2.20€	-2.20€	-1.10€	-5.50€
TOT	-8.80€	-7.78€	-4.40€	-20.98€

💰 Total Weekly Net: €-5.50

---

⚙️ The Engine Overhaul: Implementing xG

Backtesting is non-negotiable here. When you run the past 4 weeks through the new xG model, we have to look for these specific "deltas":

xG vs. Actual Goals (The "Luck" Factor):

If a team has scored 10 goals but only generated 5.0 xG, they are "overperforming" and due for a regression (a loss or dry spell).

Model V1 sees them as a strong winning team. Model V2 (xG) should see them as a vulnerable team. This is where we will find our edge on the Over/Under markets specifically.

xGA (Expected Goals Against):

This is arguably more important than xG for 1X2 betting. A team might win games, but if their xGA is high, they are allowing dangerous chances.

Example: Man Utd conceding 4 goals. A simple form guide might not predict that, but if their xGA over the last 3 games was rising (e.g., 1.8, 2.1, 2.4), the model would flag their defence as "leaking," even if they weren't conceding many actual goals yet.

I still have some days left.

Cheers,
Peter
Disclaimer: Educational experiment. 18+ Only.

Posted Using INLEO