Jump to content

[Sug] Matchmaking Phase 2.5


10 replies to this topic

#1 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 18 December 2012 - 06:58 AM

Ok, so after playing with the current matchmaking scheme for a while now, I've come up with some suggestions which (I hope) could be implemented as a stopgap measure prior to the more complex matchmaking that is due to be phased in eventually. These suggestions are meant to offer some improvements over issues that I see in the current system, while minimizing the effort required for PGI's developers, thus making it feasible to incorporate in the short term without derailing larger development plans. Of course, having not seen any of PGI's code, this assessment is purely speculative, based on an attempt to use functional elements which already exist in some form within the game's current incarnation.

My suggestion hinges upon the 8-man matchmaking, as the 4 man matchmaking is fine in its current incarnation (imho). The 8-man needs some serious work, for the following reasons:

Issues with current 8-man matchmaking
  • There is no attempt to balance mechs at all on either team. This actually has the effect of reducing flexibility in terms of workable configs, as to some extent you can (assuming competent pilots) win some using tonnage alone. There is certainly no "auto-win" config currently, but based on experience in prior incarnations of MW4, in competitive league play, I think we can do better here, fairly easily.
  • There is currently a large chasm between having 1-4 players, and then having 8. This makes it somewhat problematic, as when a unit has 1-4 players on.. that's fine. No problem. However, once you hit 5, things start getting weird. Groups need to be split, etc. Once you get more than 8, things get really bad, because then you're forced to have everyone split into really small groups and play in <4 man teams, or have a few guys just sit out by themselves. Either way, this is not optimal, and I think we can do better here as well.
Now that I've established those issues, I'm going to propose a potential solution, that I hope will be given some consideration. As I said, I hope that its implementation would be simple enough to support quick development, but it's certainly possible that it just "won't fit" with what currently exists. If that is the case, so be it.

Suggested changes to 8-man matchmaking:
  • Break the 8-man playlist up into 3 tonnage based sub-groups, at tonnages 400, 550, and 700. These numbers themselves could be adjusted, although I think those numbers actually support a decent variety of drop configs. At 700, you can basically run ANYTHING other than 8 atlases. At 400, you actually start having to decide how to fit the lance into their mechs, choosing to take a lighter mech like a commando in order to bump another mech up to a heavier chassis. Based on prior experience in league play in Mechwarriro 4, this is a good thing that makes things interesting.
  • Show the number of people currently in a given tonnage group. This is seen in most other games, allowing you to get an idea of where the population is playing. It may help avoid the potential problem of folks getting spread "too thin" across multiple drop types. (I'm not sure if this will be a problem or not, as I'm not sure what the overall player base at any one time is)
  • Allow teams to drop with less than 8 men. Note, what I'm suggesting here is that you allow a team to voluntarily drop undermanned. No lone wolves would be matched in this case. You could choose to drop with 6 pilots, for instance, but then you would be playing 6 vs 8. In many cases, this could present a disadvantage, but if you couple this with a tonnage limit, you could potentially reduce that disadvantage. That is, a team with fewer pilots could choose to drop heavier mechs, and end up matching tonnage with the other team.
That is the suggestion. The reasons I think that this would be feasible stem from the fact that the current game essentially already supports all of the pieces of it, with the exception of a tonnage calculation prior to allowing a group to drop. This seems like it could be implemented fairly easily though, when the team presses the launch button. Given that we already support multiple playlists, this just becomes a matter of adding in a few more, with specific restrictions on them.

The benefits I see are twofold. First, it gives flexibility to teams when they hit that awkward number of players around 7 or so... Because, frankly, in MW4 we would periodically choose to drop a man down to make tonnage. It's acceptable, and if a team wants to drop undermanned, I think it's fair for them to do so. I know that our unit would prefer to drop with 7 men, than have to break into two groups of 3 and 4.

Second, I think the tonnage limits would provide an important balancing aspect to the game at this point, providing some additional utility to mechs which may otherwise be limited in their utility. In past league play experiences, tonnage has always served to be a useful balancing point. Certainly, there are better potential options for such things (such as Battlevalue that is dynamically deteremined via measuring actual play statistics), but for the short term, I feel that tonnage would be better than nothing.

I think that covers it fairly well. If you bothered to read all of that, thanks.

#2 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 18 December 2012 - 07:10 AM

Would still need much programming time for the logig and expose mubers PGi does not want to show ( ppl online sort of).

Just add some lousy group leader list and let one groupleader invite the other one for match. once all players in both groups rdy... launch.

Rest will be done by community until the "real" matchmaking is finished.

Never the less, nice thoughts on your post but well... its not a fast hack to bridge the time..it would be much more a branch of the current mm.

#3 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 18 December 2012 - 07:17 AM

Quote

Would still need much programming time for the logig and expose mubers PGi does not want to show ( ppl online sort of).

I actually don't think so, for the reasons stated. They already support multiple playlists. They already support tracking online players (they used to show this number, actually). As I said, I think they only major part which would be new would be calculating the tonnage of the team, but this is fairly trivial... you know what mechs players have readied up, and it already is performing checks on those mechs prior to letting them drop, to avoid hacked mechs or mechs with no engines, etc. Thus, this would simply end up being a sum of the total weight of the team's mechs, and making sure that it is less than the playlist's tonnage.


Quote

Just add some lousy group leader list and let one groupleader invite the other one for match. once all players in both groups rdy... launch.


I suspect that this would actually be a far greater undertaking, as you are essentially suggesting implementation of a lobby system. Additionally, it opens the door to exploitation, as teams could farm "bot" accounts.

#4 Captain Wolfsburg

    Member

  • PipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 148 posts
  • LocationTexas

Posted 18 December 2012 - 07:33 AM

I think balance in matchmaking should be based on battle value, like the TT game. Mechs all have a BV based on tonnage, firepower, speed, armor etc. Going on tonnage alone causes balancing issues in itself, as a 60 ton mech could easily have a similar BV as a 80-100 ton mech, depending on the configuration.

#5 Tesunie

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Seeker
  • The Seeker
  • 8,586 posts
  • LocationSeraphim HQ: Asuncion

Posted 18 December 2012 - 08:27 AM

As someone who pugs all the time, I wouldn't mind occationally being dropped in a premade team of 6-7 vs another premade team. Provided that the premade I dropped with can talk to me and tell me the plans... I'd be more than happy to finish filling someone's team roster to help them out and keep them from being a man down. So I agree with that. Though, I do also feel that premades of 5-8 should be set up against other premades of same size (8) with some pugs to fill in the ranks maybe? Would certainly let us "PUGs" see how the "big boys play". Might even help us get into a group as well if the group with dropped with ends up liking us or thinking we did well.

Just my side note on the subject...

#6 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 18 December 2012 - 09:22 AM

I'll address some of the more recent statments.

Regarding Battlevalue, as I stated, I think this will be a good mechanism by which to balance teams eventually. Additionally, I suggest that such a Battlevalue be derived dynamically via a market based approach, such that mechs and equipment which are used most by the community end up having the highest battle value. These values could change on a weekly basis. This would then create an automatic balancing mechanism, as equipment and chassis which were most effective would end up being automatically balanced by the matchmaker.

However, that being said, I think that such a system is too complex for a short term implementation, and is more of a long term goal. In the short term, tonnage provides a perfectly workable, albeit imperfect, balancing system. Certainly, it does not provide the same mechanism by which to account for stronger/weaker variants and equipment that the battlevalue system described above does, but based on years of league play in MW4, it's apparent that it does provide SOME significant balancing mechanism. For instance, 200 ton drops in MW4 meant that you could take 4 Ryoken and 4 Shadowcats, but you couldn't just bring 8 novacats and large laser everything to death. You still had certain mechs of any given weight float to the top in terms of usage, but again, some balance is better than none.

In terms of mixing pugs into the pre-made, I specifically avoided this. The reason is the following:
1) If you mix in pugs to the premade, you preclude any kind of tonnage balancing, since it'd have to find appropriate numbers of pugs to fill a team, and also make sure they were running just the right amount of tonnage. This makes everything signficantly more complex.
2) It makes it a little less clear to a lone wolf what he's signing up for when he drops. He may get dropped into a random group of guys, or he might fill out a lance of premades. Either way, it's more complex, and less clear the user.

For those reasons, I decided to go with the simple approach of leaving it up to the guys dropping. When you form up your team, you can choose how many people you want to drop with, and what mechs they are in. You know ahead of time exactly what your lance is going to have in it, and can plan accordingly. This enables tonnage based balancing, by giving you the option to use that tonnage across however many mechs you know you are going to have in your team.

#7 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 19 December 2012 - 06:38 AM

Shameless bump, just to get this idea out to more folks.

Additionally, given that Paul has provided more info on the ELO matchmaking, I'd like to further urge the developers to include some kind of tonnage or Battlevalue into the matchmaking scheme.

Player skill should of course be a consideration, but the value/tonnage of the mechs being used cannot simply be ignored, or else you will severely hinder the variety of mechs you end up seeing in competitive play.

#8 MaddMaxx

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 5,911 posts
  • LocationNova Scotia, Canada

Posted 19 December 2012 - 08:08 AM

Is the tonnage selection ONE sided? Which Team set s the Tonnage for the Match? Why would any Commander take anything less than the max, one would have to assume that the opponent will always take the Max tonnage allowed?

The tonnage idea works but only if the Leaders get into the queue first, agree on said Drop tonnage allowed and then bring in the other Team(s) players to fill up the slots. Sounds ok but requires full teams be in place before the negotiation ever begin.

It would appear to require a new and separate MM format...

Edited by MaddMaxx, 19 December 2012 - 08:09 AM.


#9 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 19 December 2012 - 09:22 AM

View PostMaddMaxx, on 19 December 2012 - 08:08 AM, said:

Is the tonnage selection ONE sided? Which Team set s the Tonnage for the Match? Why would any Commander take anything less than the max, one would have to assume that the opponent will always take the Max tonnage allowed?

The tonnage idea works but only if the Leaders get into the queue first, agree on said Drop tonnage allowed and then bring in the other Team(s) players to fill up the slots. Sounds ok but requires full teams be in place before the negotiation ever begin.

It would appear to require a new and separate MM format...

In my suggestion, I'm saying that you choose your playlist based on how much tonnage you want to drop.

So, for instance, when you choose to make a team, instead of simply selecting "8-Man Team", you select "8-man, 400 Ton Team".

Then, the matchmaker will attempt to match you against other teams who have decided to engage in an 8 man, 400 ton drop.

And, again, there is no "Fill up the other slots" option here.

You make a team for a fixed tonnage. If you drop with less than 8 men, then you are dropping with less than 8 men. Those empty spots will not be filled.

Edited by Roland, 19 December 2012 - 09:23 AM.


#10 Yanlowen Cage

    Member

  • PipPipPipPipPipPipPip
  • Knight Errant
  • 637 posts
  • LocationWest Virginia

Posted 19 December 2012 - 11:56 PM

wild thought! Have each mech assigned a BV (Battle value)! Derived from tonnage,weapons,ancillary systems so forth and so on. ADD the total BV for both side up and drop similiar BV units vs each other. This way groups would have similiar capabilities or weights.

#11 Captain Wolfsburg

    Member

  • PipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 148 posts
  • LocationTexas

Posted 21 December 2012 - 02:37 AM

View PostRoland, on 18 December 2012 - 09:22 AM, said:

I'll address some of the more recent statments.

Regarding Battlevalue, as I stated, I think this will be a good mechanism by which to balance teams eventually. Additionally, I suggest that such a Battlevalue be derived dynamically via a market based approach, such that mechs and equipment which are used most by the community end up having the highest battle value. These values could change on a weekly basis. This would then create an automatic balancing mechanism, as equipment and chassis which were most effective would end up being automatically balanced by the matchmaker.

However, that being said, I think that such a system is too complex for a short term implementation, and is more of a long term goal. In the short term, tonnage provides a perfectly workable, albeit imperfect, balancing system. Certainly, it does not provide the same mechanism by which to account for stronger/weaker variants and equipment that the battlevalue system described above does, but based on years of league play in MW4, it's apparent that it does provide SOME significant balancing mechanism. For instance, 200 ton drops in MW4 meant that you could take 4 Ryoken and 4 Shadowcats, but you couldn't just bring 8 novacats and large laser everything to death. You still had certain mechs of any given weight float to the top in terms of usage, but again, some balance is better than none.

In terms of mixing pugs into the pre-made, I specifically avoided this. The reason is the following:
1) If you mix in pugs to the premade, you preclude any kind of tonnage balancing, since it'd have to find appropriate numbers of pugs to fill a team, and also make sure they were running just the right amount of tonnage. This makes everything signficantly more complex.
2) It makes it a little less clear to a lone wolf what he's signing up for when he drops. He may get dropped into a random group of guys, or he might fill out a lance of premades. Either way, it's more complex, and less clear the user.

For those reasons, I decided to go with the simple approach of leaving it up to the guys dropping. When you form up your team, you can choose how many people you want to drop with, and what mechs they are in. You know ahead of time exactly what your lance is going to have in it, and can plan accordingly. This enables tonnage based balancing, by giving you the option to use that tonnage across however many mechs you know you are going to have in your team.


Why would BV have to be handled like that? It's an interesting idea, but makes little sense. BV should be based on a preset statistics of how effective your mech "should" be based on heat dissipation, weapons value, armor value, cost of mech, and tonnage. TT had this, and this game could easily have it as well. Adding a market-based BV would be way too complicated and in the end take away from the gaming experience by restricting players as to what load-outs they should take to stay within BV requirements in each match. People would either purposely nerf their builds to stay under BV, or would simply stick with what they have already. Either scenario makes that sort of BV balancing impractical.

Just stick to the tried and true, I say.





3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users