37 replies to this topic

[FiveDigits]

bob1234567890, on 20 December 2012 - 04:02 AM, said:

WOW if that's how they are doing it that is easily exploitable and i look forward to exploiting it in my rather large Merc Group.

How would you intend to go about it. Intentionally lose dozens of games, ruining your personal ratings, just to be able to drop with 8 D-DCs and get matched against a mixed weight team with 2-3 ECMs?

bob1234567890, on 20 December 2012 - 04:02 AM, said:

They dont have the populations to keep breaking people into groups like this. 8v8 almost never finds a match and that's cause of tonnage problems, now they want to make it so pilots affect how to find a match. If they don't allow rating differences of several hundred AT MINIMUM people will have trouble finding matches. Then you'll get massive differences between the teams. so dumb, matchmaking is so easy to do and balance why do we have to go through 50 phases so they just keep going in circles

There is no "tonnage problem" in 8v8 as there is no weight matching. You can drop in 8 lights and face 8 assaults, or anything in between.
Every matchmaking algorithm faces the same challenge. It starts with a rather strict matching rule. If it doesn't find two teams that match those rules then it can keep searching for a longer time or soften the rules, but those are mere implementation details.

Edited by FiveDigits, 21 December 2012 - 01:52 AM.

[Buck Cake]

As soon as Elo comes into MWO, it's gonna become both the main feature of the game and the main cause of excuses. Anyone who's at least tried LoL knows what I'm talking about.

Elo games have so called Elo Hell. Eg. when you're a beginner you're always matched with rude people and trolls until you get good enough to carry yourself out of it.

[Sarevos]

FiveDigits, on 18 December 2012 - 07:05 AM, said:

I'm not sure about that. This would come down to tracking rating for each mech component. This might become a little too complicated. It would of course be more precise than tracking variants only.
On the other hand, the mech variant decides which equipment a mech can fit - ECM or not, engine size, hard points etc. Other than that every player can effectively always choose the "optimal build" for his chassis. So, I'd consider it fair to match players based on their mech's potential instead of their actual equipment - for the sake of keeping the system simple.

such a tally of hard coded values would take microseconds your team would have a total value calculated before you all even hit launch and the match would be place against the nearest value much faster than weight limiting actually since it just matches the numbers to the nearest value incredibly quick and easy Redshift cannot be wrong iz not possible lol.

[FiveDigits]

Sarevos, on 20 December 2012 - 05:06 AM, said:

such a tally of hard coded values [...]

The whole point of the proposed system is to have the mech matchmaking values not be hard coded but based on an Elo-like rating system just like for the players.
If we went for the "per component" system a match would go like this:

• Pre-match: Get Personal Ratings for all 8 players and aggregate them. For each player's mech, get the ratings of all its components (chassis, weapons, upgrades, equipment) and aggregate them to a Mech Rating. Now somehow aggregate (e.g. average) a Team Drop Rating out of the team's Personal Ratings and Mech Ratings.
  
• Matchmaking: Look for two teams with similar Team Drop Rating and match them against each other.
  
• Post-match: Adjust each player's personal rating. Winners get +X, losers -X based on the rating difference pre game. Now comes the tricky part, adjusting the component ratings. For each player's mech on the winning team you have to add +X rating to every component. Every component on the losing team drops -X rating accordingly. That's a lot of data base operations and might be just too costly. And that's just using a blanket rating adjustment based on the Team Drop Ratings. It gets exponentially more complex if you want to adjust personal and component ratings individually.

I stand by my statement that tracking Mech Rating on a per variant base is the better approach. Rating a mech this way is very similar to rating a player. The variant dictates what you can build. The Mech Rating reflects how people build a mech on average.

[Zerbob]

I am excited to see matchmaking entering MechWarrior Online because it will add a nice competitive edge to the gameplay. It will force teams to work together more or else watch as their Matchmaking Rating keeps falling.

As for the argument of Mech Rating vs Battle Value, it's really a tough call. No matter what you pick you'll have people who claim the system is wrong. The other option is of course a drop limit but that could be tough for solo-4 man queue. A BV/MR would be much easier in terms of getting players into games on balanced teams. I personally believe that a Battle Value system would work better because some variants can be set up VERY differently (example is the Stalker. It excels both as a brawler and as a long range support 'Mech but both would have very different BV's). A Mech Rating I don't feel would be able to include that factor as well, but that's just my 0.02 C-Bills.

[FiveDigits]

Zerbob, on 21 December 2012 - 01:51 PM, said:

[...] I personally believe that a Battle Value system would work better because some variants can be set up VERY differently [...] A Mech Rating I don't feel would be able to include that factor as well, but that's just my 0.02 C-Bills.

This has been mentioned earlier and I agree that it's an arguable point. The option is to expand the MR to a composite Mech/Component Rating where each piece of equipment or upgrade has it's own rating. The level of detail is arbitrary and the limiting factor is complexity and performance. The rating algorithm needs to run with limited resources and in reasonable time.

[Royal Flush]

The overall system will still be broken as long as they allow Pre-Mades to drop into PUG matches.

The scores do not take into account the advantages the premades have with voice coms, ecm, and the optimizing of mech loadouts.

[FiveDigits]

I especially had 8v8 in mind when thinking about this system. It can of course not fix queue separation issues with matchmaking.

[Pachar]

I agree with this option simply because it's a simple and elegant solution that will become better the longer it's implemented.
I would like it if the value of a mech was modified by the way it was loaded out, but oh well I think this would be good enough to come close to decent matches.
In addition to personal ELO and MR I would also balance with the total tonnage on each side, no limits just equal.
The ELO systems that I've dealt with in the past always have a modifier applied to the total score of premades that increases their score to the point that they are dealing with more skilled players to offset their advantages.

Edited by Pachar, 31 January 2013 - 08:03 AM.

[FiveDigits]

Pachar, on 31 January 2013 - 08:00 AM, said:

[...] The ELO systems that I've dealt with in the past always have a modifier applied to the total score of premades that increases their score to the point that they are dealing with more skilled players to offset their advantages.

Yup, League of Legends does this successfully. Additionally I'd like the marking of premade groups in the score board like World of Tanks does it.

[MW Waldorf Statler]

Mathematics is totally unsuitable for the calculation of complex systems, because the vast number of variables and contingencies can not be measured mathematically, deshlab each of these systems will be lower than a Lobbystem, with social interaction outside of the battle field, but within the Games

[Regrets]

Please fix the match maker and end PUB stomps.

[MW Waldorf Statler]

a Match maker ,was deselect Player under 16 Years ,anticozial People,people with no Game abilitys and coordination,Idiots

[Pr8Dator]

WHAT???? Dropping 8 experienced and good player ravens against 8 inexperienced/ bad player assaults... you gotta be kidding me!!! What kind of matchmaking is that????

[FiveDigits]

Pr8Dator, on 01 February 2013 - 08:01 AM, said:

WHAT???? Dropping 8 experienced and good player ravens against 8 inexperienced/ bad player assaults... you gotta be kidding me!!! What kind of matchmaking is that????

Did you actually read the suggestion?
Mech Rating would specifically even out advantages/disadvantages of certain 'mech chassis.

• A high-rated player in a high-rated 'mech would face other high-rated players in high-rated mechs
  
• A low-rated player in a low-rated 'mech would face other low-rated players in low-rated mechs.
  
• A high-rated player in a low-rated 'mech may be matched against
  
  • low-rated players in high-rated 'mechs
    
  • average players in average 'mechs
    
  • other high-rated players in low-rated 'mechs
• ...

Edited by FiveDigits, 02 February 2013 - 08:33 AM.

[Nik Van Rhijn]

A lot of the opposition comes from people who want a lobby system where you can decide what mechs, people maps etc are involved in the drop with no "random" elements at all.

[FiveDigits]

There is no need for that.
The prime example of successful F2P games, League of Legends, offers both custom games (lobby) and match-made games.

Having a lobby and a good match maker should be our goal for MW:O, too.

[Erasus Magnus]

mech elo sounds like a good way to place FOTM builds higher up in the ranks, where they actually can be dealt with properly. however, i see a problem with the way, for example, mass srm a1 splatcats, will drive the mech elo rating up for that particular variant, preventing this chassis from being used differently. because of the fact that the srm cat is so strong, it would render almost every other take on the chassis obsolete, because you would get ranked up as high as using that said srm build, while in fact your are running it as an lrm carrier, which isnt up to par to said srm cat in most aspects.
so ist will pigeonhole variants in those said fotm specs.

so i think your mech elo shoud be expanded to a variable component battle value system, where the use of srms, lrms, ecm,etc are factored in.

perhaps merge hard value BV, Player Elo and your mech variant elo together in an (un)holy trinity of factors to search for the fitting opponents?

or make a fluid bv system, i think that would be the best of both ways, since the values of certain components get reevaluated constantly. i dont know how complicated that would be, since there are a lot of factors and equipment to be factored in.

Grüße aus dem beschaulichen Bottrop

Era

Edited by Erasus Magnus, 19 February 2013 - 08:12 AM.