Picking an MVP in LoL in-houses — who actually carried each game

After an in-house, the 'who's MVP tonight' talk starts on its own. The easiest yardstick is KDA, but pick on that alone and a carry takes it every time. The tank who ate everything and opened the fight, or the support who warded the map and died buying time, both come out with low KDA.

The person who actually won the game and the person with the best KDA are often not the same. To judge MVP properly you need a different yardstick per role.

The same number means different things depending on the role. Comparing a mid's CS per minute against a support's is pointless. By role, look at things like:

• Carries (ADC, mid): damage share, CS per minute, damage dealt without dying
• Tanks and bruisers: damage taken, engages landed, time held on the front line
• Jungle: objective involvement, successful ganks, vision control
• Support: vision score, utility (heals, shields, CC), reads at dangerous moments

Seen this way, it becomes natural for a support who went 2/5/10 to be the night's MVP.

MVP is ultimately a human call, but a basis cuts down on grumbling. 'It felt like they played well' and 'a 34% damage share without dying once' don't carry the same weight.

ONNJ pulls these metrics straight from the replay, so opening a match's detail shows who did what, by role, in numbers. When you settle the MVP by reopening that game and pointing at the basis, both the pick and the runners-up accept it.

Pick only an MVP and the same strong players are always the star. Add an award for the other end and the mood lifts.

A 'reverse MVP' — longest time dead, top damage on the losing team, worst vision score — gets used as ammunition, so keep an eye on the tone. In most groups, though, awards like these spark more conversation than the MVP does.