65 replies to this topic

[Steven Dixon]

Ryvucz, on 26 December 2013 - 05:22 PM, said:

Let's take a look at it this way, both teams are matched by ELO, which claim they are the same, fair enough.

ELO does not care about tonnage, and one side has more firepower and armor at their disposal.

Not very fair regardless of skill, is my point.

In closed Beta (yes I went there), it was purely mech vs mech matching (albeit more difficult to do now) we had the best matched games, there wasn't steam rolling until 4 man queued up or people did 8 man sync drops.

Then ELO was introduced, and there is rarely a good, close match, it's just steam rolling, one side or the other.

I've had 100% the opposite experience, in closed beta my matches were almost always extremely 1 sided (we either slaughtered them or we were slaughtered) as soon as Elo was added by gameplay experience drastically changed and I had many extremely close matches. This is one of the reasons why I've always been a strong proponent of Elo, but this is just a personal antidote on my part. I'm all for trying to further refine and improve matchmaker and Elo, but at least in my experience it was way better than the earlier matchmaker.

[Ryvucz]

Steven Dixon, on 26 December 2013 - 06:50 PM, said:

I've had 100% the opposite experience, in closed beta my matches were almost always extremely 1 sided (we either slaughtered them or we were slaughtered) as soon as Elo was added by gameplay experience drastically changed and I had many extremely close matches. This is one of the reasons why I've always been a strong proponent of Elo, but this is just a personal antidote on my part. I'm all for trying to further refine and improve matchmaker and Elo, but at least in my experience it was way better than the earlier matchmaker.

And now we're completely opposite again. Ha ha

I'm just not feeling it with ELO, has several others are.

Elo was designed exclusively for Chess, hence the win/lose track.

Multi-player team based game with 12 players per team + one vs one stat tracker (win/lose) = bad formula in my opinion.

Sure, some people love it, but just the same hate it.

Something else needs to be used.

[Deathlike]

We already did this at @ MM v2. It wasn't strictly tonnage only but CLASS WEIGHTS.

AT THE TIME, the Awesome had the same "weight class" as the Atlas. The Commando/Spider had the same "weight class" as the Jenner/Raven. The Dragon had same "weight class" as the Cataphract.

Let's say there was whining to be had.

[BillyM]

Tonnage limits will allow ELO a much-closer starting point, and therefore greatly improve the changes of building a well matched game...

--billyM

[LoneUnknown]

I would say the bigger issue with drops is still that you can have a group of coordinated players in communication with each other go up against a complete disjointed pug. Those one sided steamrolls probably had a premade team on the winning side, and it had little to do with ELO or mech tonnage.

Either giving the game a VOIP that everyone has access to, or separating premades vs pugs are the solutions to better matchmaking.

[Ryvucz]

LoneUnknown, on 27 December 2013 - 01:23 PM, said:

I would say the bigger issue with drops is still that you can have a group of coordinated players in communication with each other go up against a complete disjointed pug. Those one sided steamrolls probably had a premade team on the winning side, and it had little to do with ELO or mech tonnage.

Either giving the game a VOIP that everyone has access to, or separating premades vs pugs are the solutions to better matchmaking.

I've actually seen quite a few pre-made 4 man groups fail horribly in PUGs. Usually the pre-mades blame the PUG players for the failure.

I just blame Elo for everything.

Deathlike, on 27 December 2013 - 01:03 PM, said:

We already did this at @ MM v2. It wasn't strictly tonnage only but CLASS WEIGHTS.

AT THE TIME, the Awesome had the same "weight class" as the Atlas. The Commando/Spider had the same "weight class" as the Jenner/Raven. The Dragon had same "weight class" as the Cataphract.

Let's say there was whining to be had.

I think that is more fair than the current system.

[cat97628]

I guess I do not understand why a new system needed to be created to balance the drops. Battletech has had a battle value system for a very long time. It could be used to balance tonnage, weapons, modules, and pilot skill. PGI keeps forgetting that this version of the game is based on something that is 30 years old and has had solutions in place to fix the problems that are seen on the field.

http://www.sarna.net/wiki/Battle_Value

[Ryvucz]

cat97628, on 27 December 2013 - 03:03 PM, said:

I guess I do not understand why a new system needed to be created to balance the drops. Battletech has had a battle value system for a very long time. It could be used to balance tonnage, weapons, modules, and pilot skill. PGI keeps forgetting that this version of the game is based on something that is 30 years old and has had solutions in place to fix the problems that are seen on the field.

http://www.sarna.net/wiki/Battle_Value

The devs have decided not to use that system, it was in one of the AtD answers.

[Noth]

cat97628, on 27 December 2013 - 03:03 PM, said:

I guess I do not understand why a new system needed to be created to balance the drops. Battletech has had a battle value system for a very long time. It could be used to balance tonnage, weapons, modules, and pilot skill. PGI keeps forgetting that this version of the game is based on something that is 30 years old and has had solutions in place to fix the problems that are seen on the field.

http://www.sarna.net/wiki/Battle_Value

Yet is still considered poor for balancing. It is also not accurate here. In BV an awesome is better than a Victor, but that simply is not the case in this game. THey would literally have to redo the entire formula for that to even relate to balance in this game.

[Escef]

This is a blatantly foolish idea, as it assumes that tonnage is the only valid indicator of combat effectiveness. I'm sure we've all seen people in assaults that couldn't shoot or maneuver their way out of a wet paper bag, and folk in lights that regularly challenge experienced assault jocks in damage and kills.

I'll take a flawed ELO system over tonnage balancing any day of the week.

[cat97628]

Noth, on 27 December 2013 - 03:45 PM, said:

Yet is still considered poor for balancing. It is also not accurate here. In BV an awesome is better than a Victor, but that simply is not the case in this game. THey would literally have to redo the entire formula for that to even relate to balance in this game.

not under standing why the need for a rework. An ERR PCC is and ERR PCC on any chassis and is worth X points towards BV. an XL 275 engine is an XL 275 engine across all chassis and is worth X points towards BV. the only variables not taken into account by BV would be the hit boxes which need to be addressed any ways and if they put in a known working BV system for balance they could then work on other aspects of the game.

[Noth]

cat97628, on 27 December 2013 - 08:24 PM, said:

not under standing why the need for a rework. An ERR PCC is and ERR PCC on any chassis and is worth X points towards BV. an XL 275 engine is an XL 275 engine across all chassis and is worth X points towards BV. the only variables not taken into account by BV would be the hit boxes which need to be addressed any ways and if they put in a known working BV system for balance they could then work on other aspects of the game.

You also have things like hitbox layouts, torso twist speeds and range, height and weapon positioning. THey'd have to change it quite a bit to get it to work decently in balancing. They could not just throw it in.

[cat97628]

Noth, on 27 December 2013 - 08:28 PM, said:

You also have things like hitbox layouts, torso twist speeds and range, height and weapon positioning. THey'd have to change it quite a bit to get it to work decently in balancing. They could not just throw it in.

range is part of the weapon BV cost. and i already stated that a working balance system would give the DEVs time to fix hitboxes. TT speed weapon placement and height are all chassis specific and could be worked into a total cost if one thought they needed to. I guess my point is that there is a working system in the world and no need to create or fix ELO. give us a BV system for the mechs and a base skill level for our pilots.

[Noth]

cat97628, on 27 December 2013 - 08:55 PM, said:

range is part of the weapon BV cost. and i already stated that a working balance system would give the DEVs time to fix hitboxes. TT speed weapon placement and height are all chassis specific and could be worked into a total cost if one thought they needed to. I guess my point is that there is a working system in the world and no need to create or fix ELO. give us a BV system for the mechs and a base skill level for our pilots.

And my point is that it is not working for this game. THey would literally have to rework most of the values of that system and probably even some of eth formula itself. This game is too different than battletech table top. Also I didn't mention a thing about weapon range, but torso twist range.

[Ryvucz]

Pretty interesting when you notice it's half and half with some not even saying for or against.

That should make anyone go "hmm".

And Spiders are OP, they created the Niagara Falls from the tears of their victims.

[Rebas Kradd]

I will reiterate several reasons tonnage-only should not be used and why the idea scares the {Scrap} out of me.


1. Matchmaking as a whole should not be touched at all until hit registration is fixed, hitboxes are corrected, weapon balance is finished, further class incentives are introduced, and any and all other adjustments to role warfare are made. Every one of those factors has the ability to skew the matchmaker. Really, the matchmaker needs to be one of the very last tweaks to a complete game.


2. Tonnage-based matchmaking would hurt role warfare. We'll see less assault mechs to complement lighter matches and less lighter mechs complementing clusters of assaults. That makes both Assault and Conquest hard to win. A much better way to balance things was to improve the medium class by introducing better chassis, and that just happened. Again, lots of spurious variables that need to be addressed before the matchmaker should be.


3. Premades are a huge problem (both in their favor and otherwise). As long as pugs and premades are playing in the same sandbox, the matchmaker will struggle. We need to wait on the launch module to find out what kind of shape the community takes with it. In the meantime, don't blame PGI for the exploits of Spider lance trolls.


4. The game is constantly changing and will always have effects on the matchmaker. What happens with PGI releases a new light mech? Everyone will want to play it. A tonnage-based matchmaker would then be pulling EVERYONE down in tonnage to compensate. Same with assaults. What happens with a new weapon is released that only works on certain chassis? Same thing. We need a system that's insulated from the ebb and flow of new content.


5. Tonnage-only matchmaking is horribly troll-prone and newbie-hostile. Like I've said before, ELO-based matchmaking makes tonnage the swing factor right now. Not ideal. But tonnage-based matchmaking makes ELO the swing factor, and THAT is going to drive everyone batty nutso. Say what you want about the current matchmaker occasionally putting individuals out of ELO in order to balance a match. It makes no mathematical sense to say that making ELO the random factor will DECREASE the number of "out-of-ELO" matches people have to play. It will increase that number. Then we'll have newbies completely confounded by the unpredictable ELO disparities they face, while high-ELO K/D-obsessed kiddie guilds will step up their sync-dropping in lights and assaults in order to further pad their stats in the absence of having to face (fairly) equal opposition. No thanks.


6. Tonnage-only matchmaking punishes strategy. The game is well-balanced enough to where good teamwork and common sense can still overcome tonnage disparities. We see pictures of it every day.


7. If you're going to make tonnage the sole factor, you run the risk of pulling the one pillar that's really holding this game up: freedom to choose and customize one's own mech. Nobody wants to be forced out of their favorite chassis by someone else's preferences, whether directly through a dropship chat system (TROLL BAIT!) or indirectly through a matchmaking algorithm. Whether people realize it or not, their freedom to run their favorite builds is a design pillar of the game and keeps them here. Removing that and turning the community against each other is a gamebreaker.




Please, PGI, for the love of God, keep skill-based matchmaking and just let people gripe about the edge cases. It's better than the alternative.

Edited by Rebas Kradd, 28 December 2013 - 11:37 AM.

[Glenfiddich15Yr]

One thing WWII Online does to help this problem is supply limits. Choke points are made over the map (which is one global map from western Germany to northeastern France and southeastern England). The choke points presently in conflict along the front lines are limited in the type of tanks, planes, and infantry in supply. The biggest and strongest tanks are more limited, but gradually resupply. Some are forced to use smaller vehicles. Oh well. Deal with it. The same could be applied to this game.

[Steven Dixon]

Ryvucz, on 27 December 2013 - 10:17 PM, said:

Pretty interesting when you notice it's half and half with some not even saying for or against.

That should make anyone go "hmm".

To be fair polls such as this don't tend to be terribly indicative of general sentiment. I remember in college reading a study stating that feedback tends to be disproportionately negative (surveys, polls, forums, ect). The reason is that people don't generally go out of their way to post a comment or rate something if they are content, they have a high incentive to do it if they are discontented and a moderate incentive to do it if they are very satisfied. Actually the ratio tends to be somewhere along the line of a 90% satisfaction ratio tends to reflect a 50% satisfaction ratio in non-random surveys (depending on the industry and particulars of the survey). This is not to say that we can extrapolate a 90% satisfaction of MM based on this survey, just a fun fact.

If you look at the bottom of many receipts that why they do things like have contests (go to our website and rate your experience and be entered to win $1000, ect). Some people think this is to get more people to comment, but its not. The purpose is to get a more balanced view because all people (contented and discontented) have an incentive to win money in equal measure (in theory).

Although you really can't make determinations on such a poll, it would actually indicate that the 'silent majority' was actually pretty much in favor of Elo. This is also by the way one of the reasons why forums in general tend to be very negative even for very popular games and why companies don't tend to give much weight to forum opinions.

Edited by Steven Dixon, 29 December 2013 - 11:00 AM.

[Bhael Fire]

Elo should only be used when matching solo players, since it works best that way and provides more granularity in matchmaking.

Grouped players should only be able to fight against other grouped players and solo players that opt-in for matches against grouped players.

Furthermore, I'd like to see a Battle Value system put in place. One that ranks an individual mech's loadout based on it's weight, weapons, armor, equipment, modules and efficiencies.

Edited by Bhael Fire, 29 December 2013 - 11:43 AM.

[Skyfaller]

ELO is a failure. Win/Loss is the stupidest measure to match people that can possible be conceived of in a game. Leave it to PGI to embrace it.

It should be based on mech class.Aka if one side has 3 assaults the other side must also have 3 assaults.

If one side has an ECM the other side needs to have one too. Max 1 light/medium with ECM and 1 Assault/heavy with ECM per team.

Premades need to be forbidden in PUG matches. Even a 2 man premade can exploit and toss the matchmaker off. For example, two dual AC20 jagers in a premade pop into a team. Pow... that team has a massive brawling firepower advantage over the team that the matchmaker gave a Catapult C4 and a Quickdraw when matching tonnage/mech type.