The standard coaching note for hard support is simple: ward more. Buy more observers. Buy more sentries. Keep the map lit. It’s the first piece of advice new pos 5 players receive, and it gets repeated at every level of play.
Our dataset of 403 professional matches disagrees with it.
The Near-Zero Finding
Observer and sentry ward placement counts show near-zero correlation with winning — the number of wards you place tells us almost nothing about whether your team wins.
This held across all 404 pos 5 records in the dataset. It’s not a sample size issue. Ward placement count — the metric most commonly used to evaluate support performance — simply doesn’t correlate with game outcomes at the pro level.
Ward Kills Tell the Real Story
| Observer Ward Kills | Win Rate | n |
|---|---|---|
| 2+ | 60.6% | 127 |
| 3+ | 60.7% | 61 |
| 4+ | 65.5% | 29 |
A pos 5 who kills 4+ observer wards wins 65.5% of their games — a 17-point swing over baseline.
The distinction matters: any support can place wards, but only good supports find enemy wards. Dewarding requires pattern recognition of common placements, awareness of what the enemy needs to see, and the discipline to buy sentries and use them aggressively — not just defensively around your own jungle.
Why Losing Supports Place More Wards
When behind, pos 5 often buys sentries reactively — checking Roshan after losing a fight, rewarding spots that already got caught, responding to information loss rather than creating information advantage. This inflates sentry count for losing teams.
High ward count in a losing game is often a symptom of reactive play. The map is dark because the team lost map control, and the support is spending to compensate.
Deaths: The Hard Floor
| Total Deaths | Win Rate |
|---|---|
| 1–4 | 76% |
| 5–6 | 72% |
| 7–8 | 48% |
| 9 | 39% |
| 10–11 | 17% |
| 12+ | 10% |
The cliff between 6 and 7 deaths is the sharpest threshold in the pos 5 dataset. Below 7, win rates are above 70%. At 7–8, near coin-flip territory. At 10+, almost certainly losing.
Death count is a more reliable indicator of game outcome than ward count. A support who dies 11 times — regardless of how many wards they placed — is losing their team the game.
Practical Takeaways
- Stop measuring yourself by ward count. It doesn’t predict wins. Measure by deward rate, deaths, and courier awareness.
- Deward aggressively. Buy sentries and use them offensively. Find the ward spots the enemy support uses in your bracket and contest them.
- Track the enemy courier. Especially in the first 5 minutes. Two courier kills = 71% win rate.
- Keep deaths below 7. The sharpest cliff in the dataset. Below it, you win 7+ in 10.
- Draft heals when possible. Healing pos 5 heroes are over-performing in this meta.
The Data Behind This
All findings derived from DOTApulse’s 403-match pro dataset across patch 7.41.
Read the full research →Related articles