MOBA esports: kills, towers, dragons, barons and gold per map
TL;DR. In MOBAs (League of Legends, Dota 2) kills are what you see and gold is what decides. Kills feed gold, towers and neutral objectives turn it into map advantage, and the gold difference is the one metric that compresses all of it into a single number. All of them are read per map, never per series.
Note. Informational and statistical content. It is not betting advice nor a promise of profit. No statistic predicts a single match: they are there to estimate probabilities over the long run. 18+ only. Play responsibly.
Per map, never per series
A best-of-three or best-of-five series aggregates maps of very different length. A 25-minute map and a 45-minute one produce incomparable kill, tower and gold counts, and adding them together blends two games that have nothing in common.
That is why we publish every one of these metrics as an average per map. It is the only way a team that plays short maps and one that stretches them out become comparable.
What each one measures
| Metric | What it measures | Watch out for |
|---|---|---|
| Kills | Aggression and game tempo | A team can win with few kills by playing for objectives |
| Towers | Real progression across the map | The metric that best tracks advance towards the win |
| Dragons / Barons (LoL) | Neutral objective control | Their weight shifts with every game patch |
| Roshans / Barracks (Dota 2) | The Dota 2 equivalents | Not comparable with the LoL ones |
| Gold | Accumulated economic advantage | In Dota 2 it is published as a difference and can be negative |
Gold comes closest to explaining the result
Everything else drains into gold: killing gives gold, taking towers gives gold, controlling objectives gives gold and tempo. Economic advantage converts into items, and items decide the fights that follow. It is a self-reinforcing loop, which is why a sustained gold lead is far harder to reverse than an unfavourable kill score.
In LoL gold is published in absolute terms (a large figure per team and map) and in Dota 2 as a difference, which can be negative. They are not the same quantity: do not compare them across titles.
The esports-specific trap: patches
No other sport rewrites its own rules every few weeks. A patch can change the value of an objective, the economy or the average game length, shifting all of these averages at once.
The practical consequence is that history ages fast. An average computed over games from several different patches blends different games. That is the underlying reason we publish descriptive statistics rather than forecasts in esports.
What you will not find
These metrics come from the provider's period data, so they only exist for MOBAs. Shooter titles (CS2, VALORANT, Call of Duty) do not carry that detail, which is why their match page does not render this card: we would rather show nothing than fill it with estimates.
You can see the available data on the esports statistics page.
Frequently asked questions
Why are the statistics measured per map instead of per series?
Because a series aggregates maps of very different length. A 25-minute map and a 45-minute one produce incomparable kill, tower and gold counts. The per-map average is the only way to compare teams with different tempo.
Which metric best explains who wins a LoL map?
The gold difference. Everything else drains into it: kills, towers and objectives give gold, gold gives items and items decide the fights that follow. It is a self-reinforcing loop.
Why are there no CS2 or VALORANT statistics?
Because the provider supplies no period data for shooter titles. We would rather not render the card than fill it with estimates.