Head to head · samsung
Who is named higher when two brands both appear in the same answer
Visibility and head-to-head disagree
1st on visibility, 2nd head to head. You appear more often than rivals but lose when you both show up — a competitiveness problem, not a coverage one.
Head-to-head standing
Derived from match record alone — no chosen weights
2of 4
by head-to-head record · strength 1505 (1490–1520)
Strength
1505
Win rate
53.3%
Matches
69,503
Rivals beaten
2 / 3
League table
| Brand | Strength | Confidence | Matches | Win rate |
|---|---|---|---|---|
| canon | 1586 | 1552–1619 | 2,478 | 63.0% |
| samsungyou | 1505 | 1490–1520 | 69,503 | 53.3% |
| 1483 | 1468–1498 | 67,534 | 46.8% | |
| nvidia | 1426 | 1402–1453 | 4,491 | 39.6% |
samsung head to head
| Rival | Matches | Won |
|---|---|---|
| 65,057 | 52.8% | |
| nvidia | 3,238 | 66.9% |
| canon | 1,208 | 45.7% |
Model fit
Predicted versus what actually happened
Mean error on well-populated pairs
7.1 pts
5 pairs with 300+ matches
| Pair | n | Real | Model | Off |
|---|---|---|---|---|
| samsung vs google | 65,057 | 53% | 53% | 0 |
| samsung vs nvidia | 3,238 | 67% | 61% | 6 |
| google vs canon | 1,247 | 29% | 36% | 7 |
| google vs nvidia | 1,230 | 43% | 58% | 15 |
| samsung vs canon | 1,208 | 46% | 39% | 7 |
| canon vs nvidia | 23 | 65% | 71% | 6 |
Large errors on low-match pairs are expected — that is why ratings below 300 matches are suppressed.