Single Call vs. Double Chance
1X2 SplitSeparates outright single-outcome calls from two-way double-chance selections so true model precision is never blended.
Every published prediction settled against final score. Wins and losses, unedited.
Separates outright single-outcome calls from two-way double-chance selections so true model precision is never blended.
Verifies whether High-confidence signals genuinely outperform Moderate and Value Upset tiers in live settlement.
Both Teams To Score (BTTS) settlement is active for newly published reports. Legacy reports without BTTS fields are excluded from the denominator.
| Model Prob. | Samples | Actual Hit% |
|---|---|---|
| <45% | 10 | 40.0% |
| 45-55% | 6 | 83.3% |
| 55-65% | 2 | 100.0% |
Every settled report, including the misses.
| Date | Match | FT (HT) | 1X2 | BTTS | Goals | Score | HT/FT | All | O/U |
|---|---|---|---|---|---|---|---|---|---|
|
Sunderland
vs
UEFA Europa League
|
0-2
HT 0-2
|
MISS | — | HIT | MISS | MISS | MISS | MISS | |
|
Torreense
vs
UEFA Europa League
|
3-2
HT 1-1
|
HIT | — | MISS | MISS | MISS | MISS | HIT | |
|
Celtic
vs
UEFA Europa League
|
3-2
HT 0-0
|
MISS | — | MISS | MISS | MISS | MISS | HIT | |
|
Sturm Graz
vs
UEFA Europa League
|
2-3
HT 1-2
|
MISS | — | MISS | MISS | MISS | MISS | MISS | |
|
Benfica
vs
UEFA Europa League
|
3-1
HT 1-0
|
HIT | — | HIT | HIT | HIT | HIT | HIT | |
|
Hapoel Beer Sheva
vs
UEFA Europa League
|
0-2
HT 0-1
|
MISS | — | HIT | MISS | MISS | MISS | HIT | |
|
Celta Vigo
vs
UEFA Europa League
|
1-2
HT 1-1
|
HIT | — | HIT | HIT | HIT | HIT | HIT | |
|
AZ Alkmaar
vs
UEFA Europa League
|
1-2
HT 0-2
|
MISS | — | HIT | MISS | MISS | MISS | HIT | |
|
Bayer Leverkusen
vs
UEFA Europa League
|
0-1
HT 0-0
|
MISS | — | MISS | MISS | MISS | MISS | MISS | |
|
Sparta Praha
vs
UEFA Europa League
|
1-3
HT 1-0
|
HIT | — | MISS | MISS | MISS | MISS | MISS | |
|
Real Sociedad
vs
UEFA Europa League
|
1-1
HT 1-0
|
MISS | — | HIT | MISS | MISS | MISS | MISS | |
|
Levski Sofia
vs
UEFA Europa League
|
1-0
HT 0-0
|
HIT | — | MISS | MISS | HIT | MISS | HIT | |
|
NEC Nijmegen
vs
UEFA Europa League
|
3-1
HT 0-1
|
MISS | — | MISS | MISS | MISS | MISS | HIT | |
|
Lillestrom
vs
UEFA Europa League
|
0-1
HT 0-1
|
HIT | — | MISS | MISS | MISS | MISS | MISS | |
|
Crystal Palace
vs
UEFA Europa League
|
1-1
HT 0-1
|
MISS | — | MISS | MISS | MISS | MISS | MISS | |
|
AC Milan
vs
UEFA Europa League
|
0-2
HT 0-0
|
MISS | — | MISS | MISS | MISS | MISS | MISS | |
|
Anderlecht
vs
UEFA Europa League
|
2-0
HT 1-0
|
HIT | — | HIT | HIT | HIT | MISS | MISS | |
|
Rennes
vs
UEFA Europa League
|
0-0
HT 0-0
|
HIT | — | MISS | MISS | HIT | MISS | HIT |
Methodology: every dimension is settled against the same published report. 1X2 = full-time result matches the published direction; BTTS = Both Teams To Score outcome (tracked on reports containing an explicit BTTS projection); Goals = actual total goals is one of the predicted goal counts; Score = the final score is one of the predicted scorelines; HT/FT = the half-time/full-time combination matches; All = the core dimensions above all hit; O/U = the 2.5-goal line is settled independently. Full rules on the Methodology page.
Each report is settled on five published dimensions plus goal totals: 1X2 direction (full-time result), total goals (actual total is one of the predicted counts), correct score (final score is one of the predicted scorelines), half-time/full-time combination, and an all-correct flag when all four hit. The Over/Under 2.5 line and BTTS are settled as independent statistics.
No. The ledger is a straight count of every settled report since coverage began, and losing runs stay in the table alongside winning ones. Nothing is backfilled and no settled report is deleted.
Counting starts on the first day reports were published. Historical matches were not re-run to inflate the sample, so early totals are small by design and grow as coverage accumulates.
Yes. Every row links to its match report, which shows the original published call, the final score, and the settlement outcome on the same page.
Attribution snippet for data journalists, newsletters, and academic papers