Its getting rather old rather fast, especially since this is a great part of the game that SWTOR seems to have ignored. The AI will balance teams to be 10 to 4 then auto-abort within 30 seconds of starting the match. Why can't the AI PROPERLY BALANCE THE TEAMS SO WE CAN PLAY?

This needs to be fixed. Its a great game option that is being ignored and a lot of players see the balance by the AI as part of the issue.

This is not the AIs fault or the matchmaker. It is the fault of your fellow Swtor players actively rejecting or going AFK and missing pops. Or, someone is doing something like finishing a mission and you don’t finish the mission before the invite expires.

The matchmaker does not send out invites to an unbalanced team. It becomes unbalanced because people decline the invite disproportionately, which isn’t something the matchmaker can predict.

I have no idea why PvP in this game does not require a sufficient number of people to accept a pop before forming the match.

I'm only guessing, but I imagine its something like this:

1) Look at who is queued up.

2) Games played/Requisition spent on ships on bar = these variables are computed into some type of score, let's call it a "Rating Score."

3) Length of time waiting in queue/Grouped or Not = these variables are computed into some type of "Priority Score."

4) Make the teams dividing up the scores evenly between the highest priority 16 or 24 people in the queue

5) Send out the invites

6) Once in the match lobby, see if spots need to be backfilled.

I would guess that the system probably prioritizes starting a match, and allowing 30 seconds for backfills that could potentially be better matches to the teams, over never constructing the match in the first place.

Because, let's say 8/8 people accept on team A but only 2/8 accept on team B. So now 10 people are waiting for the match to start ... maybe they get a red system message like you do with group finder saying "A member of your group has declined the invitation." Then the matchmaker has to backfill in those remaining 6 slots ... Its already optimized the original 16 players for score, so what do you do? Take the 6 people in the queue with the highest priority score, regardless of rating score? Make the 10 people in queue wait until equally rated players enter the queue? What if there aren't 6 more people waiting for a match? All the while those ten people are basically in limbo instead of starting the match, although at least they can still do ground stuff while they wait. We've seen throughout Bioware's history that they usually prioritize fast pops over idealized matchmaking, that's certainly been the paradigm for the ground, and probably was for GSF too until they added what I call the Rating Score thing in game update 5.5. Maybe it still is, since GSF is not a popular as ground pvp.

Mind you this is all an educated guess on the steps involved in matchmaking, based on all the available evidence. Also, I'm not expressing a preference for the current system, over one that keeps the players groundside until enough people accept invites (more like activity finder does for flashpoints/uprisings/operations.) I'm simply describing how I think the process works.

I think sometimes people drop out of a match before it starts when they see the other team is far stronger and they don't want to be routed. I solo queued in an 8 vs 8 match against a well known team of 4 that was certainly using voice. They 3 capped us from the start and my magnificent total of 1 kill was better then anyone else on our side. I would not be surprised if some of the other players on our team would drop out before the start if they saw that team against them again.

Note that the matchmaker probably had no way to prevent such a mismatch. If there are 5 veteran players and 11 inexperienced players queued, and 4 are on one team, there is no way to balance the match. The only other choice would be no match at all.

On a few occasions I have seen veteran players drop out before the end if the match is going against them. I assume they just want to fluff their win/loss percentage. If you drop out of a match in the middle, does that count as a loss?

I think sometimes people drop out of a match before it starts when they see the other team is far stronger and they don't want to be routed. I solo queued in an 8 vs 8 match against a well known team of 4 that was certainly using voice. They 3 capped us from the start and my magnificent total of 1 kill was better then anyone else on our side. I would not be surprised if some of the other players on our team would drop out before the start if they saw that team against them again.

 

Note that the matchmaker probably had no way to prevent such a mismatch. If there are 5 veteran players and 11 inexperienced players queued, and 4 are on one team, there is no way to balance the match. The only other choice would be no match at all.

 

On a few occasions I have seen veteran players drop out before the end if the match is going against them. I assume they just want to fluff their win/loss percentage. If you drop out of a match in the middle, does that count as a loss?

If you drop out of a match you get a loss. But all other stats won't be affected.

Not only leaving a match gives you a loss, but all stats recorded are voided, and become a zero production for that one loss.

Which means, if you had say 4 kills 10 deaths 70k damage

And then leave the match

You get a loss

It doesnt count towards daily or weekly

Your stat line for the match is 0 kills 0 deaths 0 damage

Which, when averaged in with your stats already accumulated, is much better than if you had stayed, and got a loss anyhow.

That is correct. Your stats are out the window when you leave, or even get kicked/disconnected. AND still get a loss, even if you didnt want to leave.

So if someone wants to maintain a vey high kill/death ratio, and is in a tough match that will cause that ratio to go down, he could leave in the middle of the match and his k/d ratio would be unchanged.

I can't know if anyone is really doing anything like that, but sometimes it's hard to avoid cynicism. But I try to judge people's skill by performance in matches that I am in. A pilot who gets 6 kills and 4 deaths on the losing side of a 50-15 match could be performing just as well as a pilot who gets 15 kills on the winning side of the same match.

Yeah. A zero is better than a negative KDR, if you are looking for stat saving opportunities.

But, some do it. Most dont. Because not many know that little tidbit.

Like tonight, we were doing a DOM, it was pretty nick and tuck all thru, and my internet gave out towards the end. I dont know if we won or not, but all my stats are gone, and i get a loss, and not counted towards daily/weekly

I would have rather been there the whole way thru, lol

So, I got a loss, and fat zeros for stats. And then, it didnt pop anymore for the night.