Ogre Magi

Ogre Magi starts at 52.6% in Herald and ends at 48.5% in Divine, crossing fifty somewhere around Legend. The shape is familiar from Bristleback and Sniper: a hero that is genuinely good in low brackets and genuinely not in high ones.

The matchup table is unusually quiet for a hero this popular — only two favourable results against five unfavourable, from more than half a million games. Both wins are emphatic, though, and the first of them is worth dwelling on.

Ogre Magi beats Bristleback at 63.7%. Bristleback’s own page carries the same pairing at 36.3%, the worst result on it. They are the same games counted from both sides, and the pair makes a small point about what these numbers mean: the heaviest matchup win on Ogre Magi’s page is against the hero with the steepest bracket decline in the dataset.

That is not a coincidence so much as a description of what Ogre Magi does well. The two opponents it beats measurably — Bristleback and Zeus — have an average bracket skew of −4.4, well below the dataset norm. Its wins come against heroes that are themselves fading, which is a reasonable thing to be good at in Herald and a diminishing asset as the opponents who rely on those heroes stop appearing.

The losses point the other way. Phoenix at 36.7% is the heaviest, with Kez, Slardar, Dawnbreaker and Axe between 41.7% and 43.1%. Axe is the interesting one again: Axe’s page shows the reverse of this result at 56.9%, and Axe follows the same declining curve Ogre Magi does. Two low-bracket heroes, one of which reliably beats the other.

The second win is the one to actually plan around. Zeus at 60.0% rests on 200 games against Bristleback’s 113, which halves the uncertainty and leaves a result that is both large and reasonably firm. Where the Bristleback figure could plausibly sit anywhere from the high fifties to the low seventies, the Zeus number is pinned much closer to where it reads.

Hold the rest loosely. Kez rests on 115 games and Slardar on 110, both carrying seven points of uncertainty either way, which is enough to reorder them entirely. The bracket trend, built from 581,183 games, is the part of this page worth planning around.

One figure sits underneath all of this. Ogre Magi is still picked for 6.8% of its games at Divine, against Bristleback’s 4.3% — so the population has not abandoned it at the top the way it abandons the steepest decliners, even though the win rate there has gone under. Whatever is keeping it in high-bracket games is not the thing this page measures.