Match 7157953820
![]()
Average rank Legend 5 · 4 of 10 players show a rank
This match hasn't been replay-parsed by OpenDota yet, so the advantage graph and some stats (damage, healing, net worth) aren't available. Basic scoreboard data is shown.
Performance
Against recent games on Riki, from OpenDota.
Data from OpenDota, fetched . Players marked P1, P2… queued together as a party. Players who keep their profile anonymous stay anonymous here.
Strongest: gold per minute, better than 79% of Riki games.
The line marks the middle: half of Riki games do better, half worse. Per-minute stats depend on role and game length, so compare like with like.
Replay
OpenDota hasn't parsed this replay yet, so last hits at 10 minutes, item timings and wards aren't available. Ask it to: it's free and usually takes a few minutes.
| Player | K / D / A | LH / DN | GPM / XPM | Net worth | Hero dmg | Tower dmg | Healing | Items |
|---|---|---|---|---|---|---|---|---|
| 27 / 9 / 11 | 437 / 5 | 758 / 797 | 36.9k | — | — | — | ||
Anonymous Unranked · Pudge · Shard | 4 / 7 / 33 | 336 / 12 | 525 / 672 | 28.6k | — | — | — | |
| 5 / 13 / 37 | 184 / 9 | 491 / 589 | 24.4k | — | — | — | ||
| 11 / 7 / 16 | 458 / 3 | 650 / 687 | 34.9k | — | — | — | ||
Anonymous Unranked · Crystal Maiden · Shard | 6 / 10 / 22 | 72 / 1 | 334 / 608 | 14.3k | — | — | — |
| Player | K / D / A | LH / DN | GPM / XPM | Net worth | Hero dmg | Tower dmg | Healing | Items |
|---|---|---|---|---|---|---|---|---|
Anonymous Unranked · Outworld Devourer · Scepter · Shard | 15 / 6 / 13 | 306 / 5 | 576 / 772 | 33.3k | — | — | — | |
Anonymous Unranked · Riki · Shard | 20 / 9 / 12 | 321 / 18 | 603 / 859 | 29.9k | — | — | — | |
Anonymous Unranked · Techies · Shard | 7 / 13 / 17 | 169 / 6 | 429 / 603 | 21.3k | — | — | — | |
Anonymous Unranked · Ogre Magi | 4 / 12 / 22 | 259 / 21 | 574 / 701 | 28.0k | — | — | — | |
| 0 / 14 / 22 | 59 / 5 | 284 / 528 | 13.4k | — | — | — |
The draft favours Dire (52%). Radiant won despite the draft.
Radiant: C 59/100
Dire: C 59/100
Estimated from the draft alone, with the same data and report card as the draft trainer. It picked the winner 57% of the time on recent high-rank games it hadn't seen, so treat it as a read, not a prediction.