36 replies to this topic

[Kaijin]

But it won't solve the glaring disadvantages stock trial mechs have vs. custom mechs. One possible solution might be to have a player's Elo modified prior to the matchmaker operation depending on if they're using a custom mech or a trial mech. Say, add to the Elo if they're using a custom mech.

Edited by Kaijin, 09 January 2013 - 09:00 PM.

[Thirdstar]

Wouldn't someone in a trial mech be at the lower end of the ELO scale anyway, and therefore be matched against other newbies?


Genuinely curious, no prior experience with ELO mechanics.

Edited by Thirdstar, 09 January 2013 - 09:01 PM.

[Diaper Gravy]

Electric light orchestra scales well with many decades.

[Redmond Spiderhammer]

Yeah I think that mostly players using trials will be matched against other players who haven't built up a ton of skill yet so trial mechs per se are not the issue. If the stats system could ELO the various chassis as well and modify the players rating based on the variant they pilot that could add even more balance to the system, without having to have all the mechs balanced against each other.

[Kaijin]

Thirdstar, on 09 January 2013 - 09:01 PM, said:

Wouldn't someone in a trial mech be at the lower end of the ELO scale anyway, and therefore be matched against other newbies?


Genuinely curious, no prior experience with ELO mechanics.

Not necessarily. With the advent of the cadet program, newbies can purchase a mech they can customize much more rapidly, but perhaps before they've become 'elite' pilots. The mechs they build will be vastly superior to the trial mechs however, and so they will dominate the field in them, vs. other newbies in trial mechs until their Elo rating kicks them upstairs. Meanwhile, more newbies are buying their first mechs and customizing them to take the former victors' places in subduing the masses of trial mechs.

All the while, newbies coming in are having a bad experience because they're facing mechs with DHS when they only have SHS, their mechs have armor allocated by a fool, while the custom mechs have armor allocation that makes sense, trial mechs are stuck with useless weapons, while the custom mechs have optimized loadouts. Not that all the weapons on a trial mech are useless, but some of them are.

Another solution that has been put forth is to have separate queues for trial & custom. I am opposed to this solution because it might prevent players with higher Elos from being able to find a match if they're using a trial mech.

Edited by Kaijin, 09 January 2013 - 09:31 PM.

[von Pilsner]

Diaper Gravy, on 09 January 2013 - 09:12 PM, said:

Electric light orchestra scales well with many decades.

Sort of...



Will be interesting to see how matchmaking works with people who have insanely high K/D ratios from disconnecting... Wonder who they will be matched with...

[Hamzey]

It wouldn't matter, as there would need to be a seeding period of games before you receive your ELO new players get a cadet bonus making it extremely easy to get a mech quickly.

[Hamzey]

Or even solo/due/tri queue as your standard que for a rating - Full premade queue for another rating.
Works for almost every other MOBA out - and this is technically a MOBA

[Damocles]

What will ELO matchmaking do when populations dip and you cannot find anyone of your "skill level"
?

[Zylo]

Damocles, on 14 January 2013 - 12:36 AM, said:

What will ELO matchmaking do when populations dip and you cannot find anyone of your "skill level"
?

I suspect the Elo rating tolerance in the matchmaker will just be increased to handle this issue. If the tolerance starts at +/-50 and that results in too many failed matches they can probably increase it a bit until players stop getting "failed to find a match" errors.

If player population continued to drop this tolerance could probably be increased to the point that it includes the entire range of players. Of course this would mean the Elo rating becomes useless and the functions of the matchmaker return to the closed beta style where your group of 8 experienced players could be going against 8 new lone-wolf players in trial mechs.

[Exoth3rmic]

New people just got a large C-Bill buff to the new player experience.

It takes a lot longer than 25 matches to approach proficiency in any mech (both on a pilot basis and an unlock pilot tree basis). By the time Elo is of any concern to such people they will have plenty of c-bills to have invested in several non-trial mechs + fittings.

A poorly though out OP and pointlessly inflamatory as its only purpose is to beat up on the trial mechs themselves rather than having anything to do with Elo at all.

[Quxudica]

Hamzey, on 14 January 2013 - 12:27 AM, said:

Or even solo/due/tri queue as your standard que for a rating - Full premade queue for another rating.
Works for almost every other MOBA out - and this is technically a MOBA

This is in no way shape or form a MOBA in the slightest degree. No creeps, no turrets, no lanes, no abilities, no in game buying, no in game farming, no in game leveling etc. This genre is First Person Mech Sim, MOBAs are DOTA/DOTA 2/LoL/Awesomenauts/Super Monday Night Combat/Smite etc.

[Inertiaman]

Yet it's match based like mobas so enjoys several comparisons in terms of matchmaking and skill rating theories.

The OP is totally off the mark anyway. As someone stated above you'll be matched up with similarly new players. Most of which by their nature will be in trials. The ones that aren't will be just bad.

[Xiang]

Damocles, on 14 January 2013 - 12:36 AM, said:

What will ELO matchmaking do when populations dip and you cannot find anyone of your "skill level"
?

ELO is supposed to be based on the team average. Im not sure how large the difference will be on a team, but the 2 teams should average out to fairly close (at least from what i read).

Example: Your team has 8 players with ELO ranging from 220-270 (so no one is more than 50 points stronger) and your team averages out to say 245-250. The other team should then also have an average of 245-250 also, but you can have some stronger and some weaker players.

In the future as you have less players the spread on a team may be higher (75-100 pts?), but the team average should still work out pretty close. Then you just have to hope that your 220 ELO player doesnt run into the other team player that is at 320.....

Again, this is a best guess...

Xiang

edit: here is the link to the devs matchmaking info

Edited by Xiang, 14 January 2013 - 02:07 AM.

[Zeke Steiner]

Battle value is a far better way to balance a match than elo as MWO is essentially a move and shoot kind of game with few other dynamics to influence tier based skill. Who wins the match isn't necessarily due to the individuals accomplishments but the whole working together and exploiting the enemy teams poor position or poor mech loadouts. Rating an individual in a team game works in league of legends because the players fight in smaller teams where an individual has a bigger impact, the game itself is also balanced better than MWO allowing for specific characters to adapt mid game to their oppenents. In MWO you're just stuck with what you drop with and if you prefer a load out that is heavily countered by the enemy team, your performance isn't going to match your elo. All elo will accomplish in MWO is to segregate players of certain builds from another.

[Inertiaman]

The trouble with BV though is that great players could pick some gimped mechs and work around it. Frankly I'd take a team of good players in 3xML ravens right now over a mixed bag of assaults. A proper personal skill rating can't be spoofed (unless you really want to go out there and purposely lose 10+ games in a row).

[Zeke Steiner]

Inertiaman, on 14 January 2013 - 02:27 AM, said:

The trouble with BV though is that great players could pick some gimped mechs and work around it. Frankly I'd take a team of good players in 3xML ravens right now over a mixed bag of assaults. A proper personal skill rating can't be spoofed (unless you really want to go out there and purposely lose 10+ games in a row).

The only problem is that "great" players only know how to move and shoot fairly accurately in the game, its not something that takes an incredible amount of play time to develop so its not applicable when it comes to determining elo. The truly great players are ones who work together and that "greatness" is easier to achieve with voice communications. By using BV, a great player could still contribute heavily with a trail mech, but they have to play smarter to do that. Going toe to toe with a custom mech is likely going to end it as most trials just can't sustain dps nor do they have high burst. BV evens the playing field because it enforces limitations, and in a heavily customizable game, thats how you balance it.

[Oderint dum Metuant]

Zeke Steiner, on 14 January 2013 - 02:56 AM, said:

The only problem is that "great" players only know how to move and shoot fairly accurately in the game, its not something that takes an incredible amount of play time to develop so its not applicable when it comes to determining elo. The truly great players are ones who work together and that "greatness" is easier to achieve with voice communications. By using BV, a great player could still contribute heavily with a trail mech, but they have to play smarter to do that. Going toe to toe with a custom mech is likely going to end it as most trials just can't sustain dps nor do they have high burst. BV evens the playing field because it enforces limitations, and in a heavily customizable game, thats how you balance it.

Errr....If you ain't got it...you ain't.
No amount of play time will make you a good player, if you just don't have what it takes.
If that was the case, we'd have more good players playing Mechwarrior, there were people that played MW4 for 5+ years and were still terrible at it.

BV is a great way to balance mechs, not pilots.

Edited by DV McKenna, 14 January 2013 - 03:02 AM.

[Inertiaman]

Zeke I have no idea why you think that good teams are not made up of good players. Or that teamplay in isolation is somehow more worthy than personal skill. Nor do I see how any of this avoids the minmax underloading that a BV system would allow.

Edited by Inertiaman, 14 January 2013 - 03:02 AM.

[Zeke Steiner]

DV McKenna, on 14 January 2013 - 03:00 AM, said:

Errr....If you ain't got it...you ain't.
No amount of play time will make you a good player, if you just don't have what it takes.

BV is a great way to balance mechs, not pilots.

For those with motion sickness or cross eyes than sure, elo would help. Best bet would be the combination of the two, but I get the feeling that in mwo, the individual elo skill cap will plateu fairly quickly while those who play in a team that communicates well will actually have a much higher elo.

Edited by Zeke Steiner, 14 January 2013 - 03:06 AM.