Let’s explore Dynamic Queue, assuming 50% of lobbies the Matchmaker creates are 2-2-2 role-lock. How often would a DPS player theoretically be able to play the role they queued for?
Again, please keep in mind that this is completely theoretical. I don’t want to see any replies criticizing the numbers I chose. This is purely for the sake of the exercise. Feel free to plug your own numbers in.
The assumptions
Suppose 100 matches are created.
- 50 matches (50%) naturally become 2-2-2.
- Every DPS player gets to play DPS.
- 50 matches (50%) don’t naturally become 2-2-2 and need someone to flex.
- In each of these matches, one of the three queued DPS players switches to tank, creating a 2-2-2 composition.
- That means 2 out of 3 DPS players still play DPS, while 1 out of 3 flexes.
The math
For the first half:
- 50% of matches × 100% of DPS players remain DPS = 50%
For the second half:
- 50% of matches × 66.7% of DPS players remain DPS = 33.3%
Total:
50% + 33.3% = 83.3%
Under hypothetical numbers, a player who queued as DPS would get to play DPS about 83% of the time, while flexing to tank only about 17% of the time.
If matchmaking can naturally produce a large percentage of balanced lobbies, then only a minority of games require someone to flex.
Some clarification:
Under 50% 2-2-2 assumptions:
- Half of all matches require no role-to-role flexibility whatsoever.
- In the other half, only one out of three DPS players needs to flex.
- The average DPS player therefore keeps their preferred role roughly five out of every six games (83%).
That’s a much smaller sacrifice than many players might initially imagine.
If the natural rate is high enough, the system could deliver a much more consistent experience than Open Queue while avoiding the queue times associated with permanent role lock.
If Dynamic Queue can naturally build even 50% of lobbies as 2-2-2, it could provide a remarkably (in my opinion, as someone who plays mostly OQ) consistent experience without the queue times of permanent 2-2-2 role lock.
Conclusion
One interesting note is that there is no reason not to implement Dynamic Queue over keeping 1-3-2 as it is, regardless of the Matchmaker’s success rate in building a 2-2-2 lobby. Due to how the system is designed, it cannot perform worse than the baseline 1-3-2 system.
