Data Collection - win/loss statsistics by faction

[Laplace's Demon]

Has anyone considered implementing a data-gathering element to the game where all wins and losses are compiled by race?

I have no idea how this would be done, but it might be interesting (for the sake of balance) to know which races have a higher win/loss-rate than others, if any exists at all. This exists in some form in the tournament rankings, but that represents a select subset of the players. If the game is meant to cater to a larger audience, it would be good to know the overall win-loss rates of the races.

This idea can be taken further by separating 1v1 and 3v3 numbers. If SM win more often in 3v3 than 1v1, it may be interesting to know, and would prompt discussions on whether to prioritize 3v3 balance or 1v1, and what the reasons might be for either. Could also break down by hero, if we wanted.

But most importantly, it would be a step towards objective balance. It would let us measure changes in race performance before and after updates. Having cold numbers would help clear the fog of bias, and offer insight into how players adapt and use changes as they come.

Food for thought

[Forestradio]

https://www.dawnofwar.info/esl/stats

that represents a select subset of the players

it represents the players you balance the game around

also I would not put too much stock into the data, gk have the lowest win rate, does that mean they need hardcore buffs?
no

[Dark Riku]

https://www.dawnofwar.info/esl/stats

that represents a select subset of the players

it represents the players you balance the game around

^This and 1v1 should always get priority.

[Ace of Swords]

There were statistics for the MRTs win ratio of races not sure where it went.

[Phoenix]

I would like hero winrate statistics being added to the esl stats.

[Laplace's Demon]

Those numbers are exclusively for 1v1 in a formal setting though. Might be interesting to see 3v3 and 2v2 performance.

Regardless, win ratios by hero might be even more valuable. That might be where we see more meaningful differences.

[Takadekadaka]

It would also be interesting to see the data subdivided according to the new tier system - I wonder what the differences in trends are for lower level versus higher level players