So I finally got my friend to play MWO. I've been trying since closed beta.
I started a new account so I wouldn't totally mess up his first drops by grouping with him. Ran him through the details and a few drops on the testing grounds.
During one of the drops on testing grounds I forgot to specifically set it to testing grounds and found myself in a normal mission.
Now remember, this is a brand new account, never run a mission before, trial mechs and all. I saw splatcats, Atlas D DCs, raven 3Ls.
Is this what new players get with the Elo system? The same messed up imbalance I see on my assumedly high Elo rating?
We've all wondered why we see trials mechs at high Elo, why we're always top damage. Now from the other end all I see are cheese mechs and people I would assume shouldn't be matched with new players.
What gives?
My friend went to bed without actually dropping so I don't know how it will feel to him. Hopefully I'll find out tomorrow assuming he can stick with it.
New Player Match Making Problem?
Started by Corwin Vickers, Mar 18 2013 10:44 PM
5 replies to this topic
#1
Posted 18 March 2013 - 10:44 PM
#2
Posted 18 March 2013 - 11:31 PM
Do you have any idea what ELO means at all? Maybe you should read this first.
It does NOT mean you are paired with other trial mechs.
It does NOT mean you player that have as many matches as you.
And if you really think all Atlas and Splatcats have a good ELO you fool yourself.
One reason is as well because the ELO is not seperated by Mech. Only by player. So if someone use a "new" mech, he will maybe ruin his ELO. And have a lower ELO when going back to his pimped mech.
You are right in one point. Trial mechs should have their own matchmaking. The problem is only, if you start to add to many parameters for matchmaking, you run out of players for any.
Basicly you need sorting by: ELO (best ELO by mech), weight, XP, trial/custom, .... In the end you have to wait 15 min for a match.
It does NOT mean you are paired with other trial mechs.
It does NOT mean you player that have as many matches as you.
And if you really think all Atlas and Splatcats have a good ELO you fool yourself.
One reason is as well because the ELO is not seperated by Mech. Only by player. So if someone use a "new" mech, he will maybe ruin his ELO. And have a lower ELO when going back to his pimped mech.
You are right in one point. Trial mechs should have their own matchmaking. The problem is only, if you start to add to many parameters for matchmaking, you run out of players for any.
Basicly you need sorting by: ELO (best ELO by mech), weight, XP, trial/custom, .... In the end you have to wait 15 min for a match.
#3
Posted 18 March 2013 - 11:38 PM
New players should be fighting new players because the ELO score would be zero what the matchmaker is doing right now is its grabbing 4 vets and rookies based on ELO skill and dropping them into a match.
Its not factoring in the mech into the ELO skill.
So you could get
4 vets in Light, medium, and heavy's
4 rookies in 2 trial assault mechs and medium and heavy.
So is this a fair drop no of course not. because on the other side ELO dropped a 4 man LRM team plus 4 average players.
ELO doesn't factor in weight, type mech being dropped, or weapon dmg, just your skill level that they don't reveal to you.
That's why you see matches with 6 guys in ECM stacked up against a team with no ECM it doesn't factor these things in. That's why its fubar, and its getting so much heat. If it was working properly you would see much closer matches instead of 8-0, 8-1, 8-2, 8-3 sweeps. Its ticking people off since ELO was added my stats dropped from 75% to 68% and my win lose is down about 4% since my account is active right now, and ticking down I'm losing c-bills because your dev team is still writing in the code. ELO is Fubar.
Its not factoring in the mech into the ELO skill.
So you could get
4 vets in Light, medium, and heavy's
4 rookies in 2 trial assault mechs and medium and heavy.
So is this a fair drop no of course not. because on the other side ELO dropped a 4 man LRM team plus 4 average players.
ELO doesn't factor in weight, type mech being dropped, or weapon dmg, just your skill level that they don't reveal to you.
That's why you see matches with 6 guys in ECM stacked up against a team with no ECM it doesn't factor these things in. That's why its fubar, and its getting so much heat. If it was working properly you would see much closer matches instead of 8-0, 8-1, 8-2, 8-3 sweeps. Its ticking people off since ELO was added my stats dropped from 75% to 68% and my win lose is down about 4% since my account is active right now, and ticking down I'm losing c-bills because your dev team is still writing in the code. ELO is Fubar.
Edited by Corbon Zackery, 18 March 2013 - 11:42 PM.
#4
Posted 18 March 2013 - 11:52 PM
Or it could be working the same way it works for my normal account and throws in new players with experienced players based on an Elo target average which might be fine for me since I do well but horrible for the new player that has to fight me.
I would not expect even a horrible DC user to have an Elo as low as a brand new account.
In either case it seems like a bad situation to me. Whether it is high Elo players plus low Elo players being averaged to meet a target Elo rating for the team or everyone on my team and the enemies team having essentially a 0 Elo score (or whatever the floor is) which I don't believe is what's happening.
I would not expect even a horrible DC user to have an Elo as low as a brand new account.
In either case it seems like a bad situation to me. Whether it is high Elo players plus low Elo players being averaged to meet a target Elo rating for the team or everyone on my team and the enemies team having essentially a 0 Elo score (or whatever the floor is) which I don't believe is what's happening.
#5
Posted 18 March 2013 - 11:58 PM
I would only put players together with the same ELO, and the same for the Enemys Side. Not this, one UberSkilzplayer with 4 Noobs and some average. No one have fun with that.
Edited by Revorn, 18 March 2013 - 11:58 PM.
#6
Posted 19 March 2013 - 12:03 AM
Actually, if it's anything like how ELO works in other implementations I see...there's a range.
You don't start at 0, say the range is 700-1300, with 1000 being the average/starting point.
Assuming 1000 is the starting point, and the average, then your "average" experienced player is right at 1000, along with the brand new player. It takes a few games for them to get beaten down to the under average players.
If that's how they're doing it (which I suspect it is since that's how every other ELO system I've ever seen works), then they need to shift them to start near the bottom.
You don't start at 0, say the range is 700-1300, with 1000 being the average/starting point.
Assuming 1000 is the starting point, and the average, then your "average" experienced player is right at 1000, along with the brand new player. It takes a few games for them to get beaten down to the under average players.
If that's how they're doing it (which I suspect it is since that's how every other ELO system I've ever seen works), then they need to shift them to start near the bottom.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users
















