La Liga · Track record
How the recorded top picks for La Liga have done. Each pick is
recorded when the message goes out and settled after its match. Bets count
two ways: as % of bankroll at the stake the message suggested, and flat, as
1 unit on every bet. Predictions only count towards the hit rate, since
they came with no stake.
Overview
| Picks | Won-lost | Hit | Pending |
Bets | Profit | ROI | Flat units | Flat ROI |
| All time | 33 | 19-13 | 59.4% |
0 | 10 |
-13.4% |
-60.8% | -3.15 | -31.5% |
| Last 30 days | 18 | 13-5 | 72.2% |
0 | 2 |
+0.1% |
2.5% | +0.90 | 45% |
What went well
Best: book: DraftKings — 3-7,
-3.15 units (-31.5% flat ROI).
Worst: price: middle — 2-6,
-2.95 units (-36.9%).
By type
| Picks | Won-lost | Hit | Profit | Flat units | Flat ROI |
| bet | 11 | 3-7 | 30% |
-13.4% | -3.15 | -31.5% |
| prediction | 22 | 16-6 | 72.7% |
+0.0% | +0.00 | – |
By risk / confidence
| Picks | Won-lost | Hit | Profit | Flat units | Flat ROI |
| lower | 3 | 1-2 | 33.3% |
-8.6% | -1.20 | -40% |
| medium | 5 | 1-3 | 25% |
-3.6% | -1.85 | -46.2% |
| strong | 10 | 8-2 | 80% |
+0.0% | +0.00 | – |
| lean | 12 | 8-4 | 66.7% |
+0.0% | +0.00 | – |
| higher | 3 | 1-2 | 33.3% |
-1.2% | -0.10 | -3.3% |
By price
| Picks | Won-lost | Hit | Profit | Flat units | Flat ROI |
| middle | 9 | 2-6 | 25% |
-12.8% | -2.95 | -36.9% |
| favourite | 2 | 1-1 | 50% |
-0.6% | -0.20 | -10% |
By bookmaker
| Picks | Won-lost | Hit | Profit | Flat units | Flat ROI |
| DraftKings | 11 | 3-7 | 30% |
-13.4% | -3.15 | -31.5% |
Model vs market
Across 32 settled picks, the model gave them
54.7% on average and the bookmakers
56.6%; they won 59.4%.
Brier score (lower is better): model 0.216,
market 0.185 (on the 32 with a price).
Model vs Kalshi
On the 32 settled picks with a Kalshi market, the model gave them
54.7% on average and Kalshi 57.6%; they won 59.4%.
Brier score (lower is better): model 0.216, Kalshi 0.183
— Kalshi has been closer so far.
Closing line value
Across 6 bets with a closing line, the price taken was
-0.12%
against the close on average, and 50% of them beat it. At the closing line's
own chances those bets were worth -4.82% each.
| Day | Pick | Match | Book | Taken | Closed | CLV | Result |
| 2026-10-23 | Alavés win | Alavés v Málaga | DraftKings |
-155 | -160 |
+1.24% |
❌ lost |
| 2026-10-19 | Draw | Getafe v Rayo | DraftKings |
+190 | +190 |
+0.00% |
✅ won |
| 2026-10-18 | Draw | Celta v Alavés | DraftKings |
+215 | +210 |
+1.61% |
❌ lost |
| 2026-10-18 | Real Sociedad win | Málaga v Real Sociedad | DraftKings |
+115 | +105 |
+4.88% |
✅ won |
| 2026-10-17 | Valencia win | Valencia v Athletic | DraftKings |
+215 | +220 |
-1.56% |
❌ lost |
| 2026-10-12 | Levante win | Levante v Sevilla | DraftKings |
+170 | +190 |
-6.90% |
❌ lost |
Long shots not sent
Bets the bookmakers rated under the long-shot cutoff are left out of the WhatsApp picks and
tracked here instead, to check whether skipping them helps. So far: 1-6
(0 pending), -9.4% of bankroll at their suggested stakes,
-1.75 flat units.
Recent picks
| Day | Pick | Match | Price | Model / Kalshi | Type | Result | Profit |
| 2026-11-08 | Celta win |
Celta v Levante (0-0) |
-125 |
47% / 52% |
prediction |
❌ lost |
– |
| 2026-11-08 | Real Sociedad win |
Real Sociedad v Rayo (1-0) |
-125 |
51% / 56% |
prediction |
✅ won |
– |
| 2026-11-08 | Villarreal win |
Villarreal v Getafe (2-0) |
-210 |
61% / 66% |
prediction |
✅ won |
– |
| 2026-11-07 | Real Madrid win |
Valencia v Real Madrid (0-3) |
-380 |
56% / 78% |
prediction |
✅ won |
– |
| 2026-11-02 | Depor win |
Depor v Osasuna (2-1) |
+110 |
46% / 46% |
prediction |
✅ won |
– |
| 2026-11-01 | Athletic win |
Athletic v Real Sociedad (2-1) |
+105 |
46% / 46% |
prediction |
✅ won |
– |
| 2026-11-01 | Elche win |
Elche v Valencia (0-1) |
+135 |
47% / 41% |
prediction |
❌ lost |
– |
| 2026-11-01 | Real Madrid win |
Racing v Real Madrid (0-3) |
-450 |
67% / 82% |
prediction |
✅ won |
– |
| 2026-10-31 | Atlético win |
Levante v Atlético (0-2) |
-200 |
48% / 64% |
prediction |
✅ won |
– |
| 2026-10-31 | Betis win |
Betis v Málaga (2-1) |
-330 |
68% / 74% |
prediction |
✅ won |
– |
| 2026-10-31 | Barça win |
Barça v Alavés (6-1) |
-1000 |
76% / 88% |
prediction |
✅ won |
– |
| 2026-10-30 | Villarreal win |
Villarreal v Espanyol (0-1) |
-245 |
62% / 70% |
prediction |
❌ lost |
– |
| 2026-10-25 | Athletic win |
Athletic v Getafe (4-0) |
-180 |
49% / 62% |
prediction |
✅ won |
– |
| 2026-10-25 | Barça win |
Barça v Real Madrid (1-2) |
-155 |
50% / 58% |
prediction |
❌ lost |
– |
| 2026-10-25 | Real Sociedad win |
Real Sociedad v Levante (2-1) |
-175 |
55% / 62% |
prediction |
✅ won |
– |
| 2026-10-24 | Atlético win |
Atlético v Depor (1-0) |
-370 |
60% / 76% |
prediction |
✅ won |
– |
| 2026-10-23 | Alavés win |
Alavés v Málaga (1-1) |
-155 |
63% / 58% |
bet · 1.57% |
❌ lost |
-1.6% |
| 2026-10-19 | Draw |
Getafe v Rayo (1-1) |
+190 |
37% / 32% |
bet · 0.86% |
✅ won |
+1.6% |
| 2026-10-18 | Draw not sent |
Real Madrid v Sevilla (4-1) |
+600 |
17% / 12% |
bet · 0.9% |
❌ lost |
-0.9% |
| 2026-10-18 | Real Madrid win |
Real Madrid v Sevilla (4-1) |
-425 |
73% / 80% |
prediction |
✅ won |
– |
Every prediction
The picks above are the few matches the model felt strongest about. This is all of them: the model's
home/draw/away chances for every match, recorded shortly before kickoff with the books' lines beside
them, and scored the same way once the match is over.
| Matches | Log-loss | Brier | RPS | Called right |
Model's log-loss on the same matches | Gap (model − it) | |
| Model | 42 | 0.9483 | 0.5627 |
0.1889 | 61.9% | | | |
| DraftKings | 42 | 0.8957 | 0.5263 |
0.1744 | 57.1% | 0.9483 |
+0.0526 (+0.0067 to +0.0944) |
no clear difference |
| Kalshi | 41 | 0.8813 | 0.5178 |
0.1689 | 58.5% | 0.9415 |
+0.0602 (+0.0130 to +0.1092) |
no clear difference |
| Polymarket | 42 | 0.8909 | 0.5245 |
0.1735 | 57.1% | 0.9483 |
+0.0574 (+0.0071 to +0.1005) |
no clear difference |
Does the injury discount help?
On the 33 matches where a missing player changed the prediction,
the model scored 0.9766 with the discount and 0.9852 without it
(gap -0.0086, -0.0294 to +0.0127):
no clear difference yet.
Do the percentages come true?
| Chance given | Times | Average given | Happened |
| 0–20% | 14 |
14.3% | 7.1% |
| 20–30% | 50 |
26% | 22% |
| 30–40% | 32 |
34.7% | 34.4% |
| 40–50% | 15 |
44.8% | 60% |
| 50–60% | 7 |
54.3% | 71.4% |
| 60–70% | 6 |
64.3% | 50% |
| 70–100% | 2 |
75.6% | 100% |