Here is my feedback on the suggested changes to the group queue:
TL;DR .. in my opinion the idea will break for several reasons and not deliver on the desired goals of better balance and game play in the group queue.
-----------
1) Light mechs will always be a detriment to your team balance
If you thought light mechs were scarce before this ... I don't think you will see ANY in the group queue after this.
Not only that, but anyone in a small group that WANTS to play a light mech will be a serious disadvantage to their side in terms of tonnage balancing.
Look at the numbers.
2 players - max 150
- if a player in a 2 player group takes a light mech then the maximum for the group is 135 and their side is automatically gimped by 15 tons
3 players - max 220
- if a player takes a locust then the other two have to take 100 ton assaults ... likely ... no.
- Even a 35 ton light must be paired with two 90+ ton assaults to use the tonnage
4 players - max 285
- 1 player takes a 35 ton light and leaves 250 between 3 players or about 85 tons/player ... 1 light requires 85 ton assaults in the group to balance.
What is the bottom line? If you take a light then you force your team mates into assaults ... OR ... the entire tonnage balancing idea becomes broken.
5 player - max 365 ... a 35 ton light requires the other four to be in 80+ ton assaults on average.
2) Team optimization and maximum utilization of the least balanced mechs
12 man group - the previous standard was 3/3/3/3 ... so 3 x 35 ton lights is 105 tons ... leaving 690 tons to be split among 9 players. Average of 76 tons/player. It will be HILARIOUS (and broken) to see 12 man teams dropping in 9 x Timberwolves and 3 Arctic Cheetahs.
... and this leads into the second problem with tonnage limits instead of weight class limits. Not only will individuals min/max the meta to get the maximum performance from individual mechs. Entire TEAMS will extend this to min/max the team composition. You WILL see competitive sides with things like 9 Timberwolves and 3 Arctic Cheetahs ... simply because it will be the best team composition. Previous rules LIMITED the use of the number of most broken and unbalanced mechs ... but the tonnage limits completely break that paradigm. Unless the best mech is a 100 tonner, you will likely be able to put as MANY of them as you like on your team.
So, if you want a prediction, any tonnage limited scheme like the one described here is going to BREAK the group queue even worse than it is now. A coordinated 12 man with 12 optimaxed mechs vs. any combination of groups with anything less is going to STOMP against anything except another 12 man. On top of that, ADDING tonnage to smaller groups WON'T fix the problem since the smaller groups are more likely to be casuals, are more likely to play what they want to play (which may not be assaults) and are less likely to coordinate on the battlefield. Add to this that an organized 12 man will utilize every ton while this is much less likely for smaller groups and I think it is a recipe that won't work.
-----------------
So ... what should they do? Instead of weight class matching they could sum up total tonnage actually in each group. As the match is assembled the matchmaker will aim to give the side with more groups an advantage in total tonnage. It requires more coding, it is similar to Russ' idea, and it could possibly work ... but ONLY if actual tonnage matching is included as a factor in the matchmaker and not just throwing the groups together assuming they have whatever maximum tonnage is allocated to that group.
This still does not prevent team composition optimization by simply taking all of the best mech available. The only way I see to limit that issue is to impose both weight class and tonnage limits at group formation time. This is done in the front end and does not impact the matchmaker. If 12 mans had to be 3/3/3/3 with a tonnage limit then team min/max'ing for large groups would be reduced (though they could still field 3 Dire wolves, 3 Timberwolves, 3 Stormcrows and 3 Arctic Cheetahs with a 795 ton cap).
Edited by Mawai, 29 September 2015 - 01:32 PM.