SvR matchmaking seems weird and unbalanced

[desert surge]

I don't know if it's just me, but something about the matchmaking for this mode seriously seems to be wrong. One side frequently seems to have significantly more players than the other and close matches are rather rare because of that. I also always play the SAW side as of late for some reason and I haven't played the rebels in ages. It's kinda stale when you play on the same side all of the time and even frustrating, when you play with a character that is supposed to be a rebel (sorry Starlight).
I'm curious on what others say. I know good matchmaking is difficult when there aren't a lot of players, but I still think they are ways to potentially improve things, like by making sure both sides have the same amount of players.

[broken rose]

SvR used to be completely random but a previous update made it so matchmaking was based on your level
If 2 high level players entered a lobby they would be placed on opposite teams
One issue that may make things seem consistently unfair is when a player with a high level enters a lobby and there is a squad of high levelled players, because the squad has to be on the same team the player will be put on the opposite team as them
Its always a coin toss on what team you will be on with how many players are in the lobby
The more players the better

I think going back to assigning teams randomly would be better
and if a squad enters a lobby then the next few players joining would be placed on the opposite team to fill out the team rosters

[mystic parrot]

The weighting is also rather off.

• Too much weight on level (which itself is not a great measure)

• Too little weight on preformed party size

• Punishing 'and' disempowering of specifically higher level carries - a recipe for anger and toxicity which becomes more apparent with higher levels.

• Medic is also underpowered

.

It's play to lose, in its current format.

[errant jolt]

SvR was never completely random; players were sorted by the time they entered and had visible helmets. That's how you get a recipe for disaster, though:

• this failed to adapt to players on the smaller team lobbyhopping, leaving players who didn't want to lobbyhop stuck with a VERY imbalanced matchup.
• it made it possible for people to swap teams by leaving and rejoining. This meant that friendgroups bigger than 4 can still be on the same team, which was disastrous for the other team.

[surreal ether]

They need to balance it out by adding some really skilled bots

If there's really skilled squads on the other team

Then again this is a weird suggestion

[mystic parrot]

True. As flawed as the current system is - that which existed before it was worse.

[mystic parrot]

Also, a part of the problem is that win-rate is recorded in a mode where success and failure are at least two degrees removed from an individual's participation, while framing it as that individual's responsibility.

It's a problem because it serves to set those who try on a collision course with those who don't.

[errant jolt]

Which is why I'm perplexed when people suggest that we need to "return to old matchmaking", like, are we remembering the same matchmaking, here?

[dusty parrot]

yeah old matchmaking was a complete lottery, this has some soft matchmaking to try and balance the teams best as possible without getting too deep into a system that weighs stats, playtime, etc. (which isn't a small amount of work)

could probably use improvements of course though

[mystic parrot]

It was a lottery that could be gamed through repeating the RNG by exiting and re-entering.

Am I far off the mark in thinking that the current matchmaking also has a modifier in the background on the basis of previous games being won or lost? Asking as I've on occasion noticed puzzling matchmaking that are much less tense but then find that another challenging player in the lobby is on the same team. 🤔

[errant jolt]

Sometimes I wonder if the matchmaking is a placebo

[mystic parrot]

A placebo? How so, and to what end?

I will say, as I've mentioned in the past, that in any mode where you have faction gameplay, it is not only of importance to attempt to matchmake reasonably, but it is also important to give the game the tools that are needed to correct significant miscalculations at the matchmaking phase.

As one example of such, TF2 has the controversial rebalancing mechanic of switching colours (which also has its issues - but they also don't record win-rates, if I recall correctly).

Another mechanism is the notion of tuning the resources (or respawn times or number of units) available to a team identified as being at a disadvantage.

Frankly, only post-matchmaking tools can mitigate imbalances due to alt-account players or suboptimal class choice.

(though the class thing I do think comes down to tweaking dartguns to both self-heal 3 (on fire) and add 3 seconds of a bit (base damage (reduced for secondary darts - which could also be the self-heal value) if not 3)) of overhealth on friendlies hit at 100% - giving the class a slightly more TF2 role in SvR and also making it a little more viable as a Solo weapon)