I'm curious to know why people can duo with bronze 5's in silver/gold and the B5 can have a higher pick than me.
Because in a normal match, the ranges are VERY close to each other, except if there was a duo with wide ability differences. Random in the absence of a lopsided duo is pretty much the same as non-random because players will likely be within 5-10 MMR of each other, which does not represent a material difference in skill.
Duos had very odd behavior, and that was frustrating both duos and non-duos. I believe sometimes they would be pick 4/5, other times pick 1 and I think 4? Not sure how it fell out of the elo calculations, but was awkward, and experienced ranked players actually did complain about it from time to time.
We felt that randomizing would be an improvement in terms of removing any upsides/downsides from duoing (elo already adjusts them fairly). Most of the lingering issues people are having are around edge cases of high tier rank players duoing with very weak low rank players.
If you have someone very high rank duoing with someone very low rank, the match is just going to be lopsided one way or another (either carried by the high player or fed by the low player). This problem will exist regardless of pick order.
We will most likely restrict these sorts of premades in the future, but we are prioritizing stuff like teambuilder and some additional small enhancements around duo matchmaking accuracy and general matchmaking fairness over this, because this is a relatively rare case.
So TLDR -- I think this change is strictly positive except with the edge case of a duo with very different capability levels. In that case, we should just (at some point) block the functionality to play so far out of bounds in ranked in general.
So you fixed what isn't broken because fixing what actually is broken would take more effort? Am I understanding this right?
Actually, we did this specifically because of long-term ongoing concerns from players around how duos disrupted pick order in a non-intuitive and often frustrating manner. It actually was broken.
Read what I just posted above you. You didn't fix anything, you made it worse.
You now made everyone else who got queued up with the Bronze duo partner have to be subjugated to his bans and the role he wants because he's first pick.
I agree that IN THE EXAMPLE of a high rank player duoing with their bronze 5 teammate who also doesnt take feedback on bans and insists on playing an indvidually snowballey role (top, mid espec), this is a worse system -- if said player is reasonable and works with their team, it doesnt matter anymore than a lopsided duo already matters (again, I concede, lopsided duos are not great).
I know that this scenario of lopsided duo with weak player 1st pick is not a common one and in most cases, this randomization change is therefore a positive one. I think that if we close the edge cases around say Plat 1 players duoing with Bronze players and so forth, this will be a strictly positive change. I think right now it is a net positive change despite this particular negative example.
Why was pick order changed from elo to random?
Duos had very odd behavior, and that was frustrating both duos and non-duos. I believe sometimes they would be pick 4/5, other times pick 1 and I think 4? Not sure how it fell out of the elo calculations, but was awkward, and experienced ranked players actually did complain about it from time to time.
We felt that randomizing would be an improvement in terms of removing any upsides/downsides from duoing (elo already adjusts them fairly). Most of the lingering issues people are having are around edge cases of high tier rank players duoing with very weak low rank players.
If you have someone very high rank duoing with someone very low rank, the match is just going to be lopsided one way or another (either carried by the high player or fed by the low player). This problem will exist regardless of pick order.
We will most likely restrict these sorts of premades in the future, but we are prioritizing stuff like teambuilder and some additional small enhancements around duo matchmaking accuracy and general matchmaking fairness over this, because this is a relatively rare case.
So TLDR -- I think this change is strictly positive except with the edge case of a duo with very different capability levels. In that case, we should just (at some point) block the functionality to play so far out of bounds in ranked in general.
I know that this scenario of lopsided duo with weak player 1st pick is not a common one and in most cases, this randomization change is therefore a positive one. I think that if we close the edge cases around say Plat 1 players duoing with Bronze players and so forth, this will be a strictly positive change. I think right now it is a net positive change despite this particular negative example.
- Zileas