I dont really like the current set up of ranked games for pick civ or random civ. For those who dont know the current rules:

Every player can vote for pick civ or random civ in ranked.

- If the majority of the votes are for random civ, then it will be random civ.
- Otherwise (majority for pick civ or both options gets 50% of all votes) it will be picked.

This system is by design already in favor of picked civ. But needed a strict majority means it is even more rare to get random civ.

I do think this system has some major flaws. As result it fells for the trap of tyranny of the majority. The majority gets much more power then they deserve. And the voice of the minority is counted less. Just let a look at the numbers:

What to me looks fair is the following: If x% of all players want to play random civ and so y=100-x% of all players want to play pick civ, then also about x% of all games is pick civ and y% of all games is random civ. With the current system, that is not the case. Just some examples for a 4v4:

If 10% of all players want to play random civ, then it will happen in 0.04% of all games.

If 20% of all players want to play random civ, then it will happen in 1.04% of all games.

If 30% of all players want to play random civ, then it will happen in 5.80% of all games.

If 40% of all players want to play random civ, then it will happen in 17,37% of all games.

If 50% of all players want to play random civ, then it will happen in 36,33% of all games.

If 60% of all players want to play random civ, then it will happen in 59,41% of all games.

If 70% of all players want to play random civ, then it will happen in 80,59% of all games.

If 80% of all players want to play random civ, then it will happen in 94,37% of all games.

If 90% of all players want to play random civ, then it will happen in 99,50% of all games.

For these percentages i assumed everyone in a game picked pick civ or random civ at random with given pick rates.

Based on this i have two oberservations:

- The numbers are in favor of pick civ. This is because in case of a 50/50 split, it will be always pick civâ€¦ This makes the numbers a bit skewed in the direction of pick civ.
- The play rates are very much skewed based on the pick rates. The majority is much more in control. In much more games it gets what it want.

Then we also have a different possible issue as well. In above case we assumed everyone makes their pick independently. This is kinda only true for solo queueing. It will be a bit different for premades. Currently, they really have a lot of power. If the premades decides they want to pick civ, then it is already guarenteed to be pick civ. That is instantly forced by the system. The premade will even have full control about this.

Even for random civ, they really have a big impact. If they all pick random civ, then there is already a 35% chance of playing random civ. Even if only 20% of the player base wants to play random civ, then already 59% of getting random civ.

Even if only 10% of the players base wants to play random civ. If a premade queued up and all will pick random civ, the probably of it being random civ will be already about 35% (assuming the other team are all solo queueing). Premades already have a huge benefit over solo queueing, but this is another benefit of premades.

So we found a few issues:

- Random vs pick civ is skewed towards pick civ by design.
- The majority has much more control over the outcomes compared to the minority.
- Pre mades has a huge impact in the outcome of random civ or pick civv.

I suggest the following change:

Change the rules about pick or random civ. Dont use majority rules to determine pick or random civ. Make it proces random, like a tail flip. If there are x players wanting pick civ, and there are n players in total then x / n it the probably of the game being pick civ and (n - x) / n is the probably of the game being random civ.

This will solve most of the above issues. The number of games with pick civ and random civ will be a better reflection of the preferences of the player base.