9 replies to this topic

[Wyvern79]

I'm new to MWO and am starting the process of using the trial mechs to buy my first mech, but thats not here nor there. What I see over and over again are people saying Battle Value matchmaking will fix everything. I don't follow that logic and wanted to invite the proponents of BV matching to try to educate me.

From my point of view weight 'slots' would be a much better way of matching than BV. Simply because that Awesome 9M is not really twice as good as any of the other ones. I'd say that a team with 500 total tons going up against one with 550 or 450 is relatively balanced and would be good for the game.

Once I can have XL engines, and Double Heat Sinks, I want that to count as an 'advantage' for me for earning them as well. If I can afford to run them I should be allowed to have my tricked out Centurion go up against a stock one. I'm footing the bill right?

As long as the tonnage on each side is similar I think the economy will balance the rest. What I dont want to see is the game telling us we can only bring X number of tons to a match. Bring whatever you want and then have the matchmaker find you a group with something close.

[Kdogg788]

You can possibly use tonnage to more or less balance the teams within a certain range or percentage and instead of using battle value, balance the number of trials on each side.

-k

[Gaeb]

Daylighting as a designer:

If we attach BV to items (or to be lazy/efficient we just use cbill cost as a standin for BV - if we do, please up the cost of ECM to ~1.5MM) and tried matching by that for a week or two I'd be pretty excited.

Take the pimped out Raven 3L (ECM, Endo, DHS, XL engine) - it costs in the range of 8MM cbills with no pilot modules. Thats equivalent to a stock Awesome. That's equivalent to 2.5 stock jenners. If we then added an artificial reduction to trial mech costs (say, 75% standard) you get a much more interesting / dynamic match making system. Note: Pilot modules (the 'end game' planned) have CBILL costs!

2 UI changes that would help with this:
1) Mechlab view of mech value (without having to hit sell button) so you know what your mechs are worth in BV/cost
2) Matchmaker UI improvement:

"
Searching for group.........
Please note: Your likelihood of finding a match as a team is improved if you are closer to average BV
Average 4man BV this week: 32MM
Average 8man BV this week: 75MM
"

And then (admittedly I draw this directly from LoL) have outliers sit in queue for 5-10 minutes before failing to find a match or creating an unfair matchup, and match on a +/- 10% BV. Doesn't solve for ELO, but would be neat. The entire queue system probably needs an overhaul... but its beta =)

Edited by Gaeb, 12 December 2012 - 07:24 AM.

[Taryys]

from my thread-which-shall-not-be-named:

Quote

DO ... add Stat Based and Battlevalue (BV) Based Matched Making: Adding stat based match making so new players will be matched with new players. Note, I did not say trials to trials, since vets can pilot trials too. This will keep them from getting ROFLstomped by vet players, and give them a chance to learn and suck together and learn together.

With Battlevalue balancing match making via battle value will also help so people with more tricked out rides will drop against each other. This may be more effective in balancing out teams then straight tonnage or weight class, although that should be taken into account at least a little bit.

Stat based matchmaking should take a priority over BV factors. BV should modify, but not be the controlling factor since this is a skill based game. A new player that has bought was a brand new stock mech then they do not have any advantage more than any other person in a stock trial mech. BV should probably be taken into account after stats have. That way new players with their shiny new rides who have made some changes can be put together when after their stats have been compared

Edited by Taryys, 12 December 2012 - 06:26 AM.

[Enigmos]

Matchmaking IMV is considerably more complicated in a skill-centric game. BV is a tabletop mechanism intended for a mathematical model where offensive/defensive effectiveness depends on random dice rolls. Pilot skill has little bearing unless you are also factoring the tendency to become disorganized or rout depending on values asserted for morale and experience.

For mechwarrior online they would have to not only factor tonnage and loadout, but actual win/loss ratios, aggregate damage dealt and taken, and kill/death ratios.

Because the effectiveness of a lance is contingent on voice comms as opposed to keyboarded chat (to fire or change direction I cannot be texting), being grouped would also factor.

[Taryys]

and

Quote

DO NOT... put trial mechs in their own queue: putting trial mechs in a trial queue will create another set of problems, simply because vet players can also use or may need to use trial mechs. Vet players who are trying to game the system may try to jump into a trial mech to try to frag some newbs to make some easy money and destroy the game for others. We also cannot get rid of access to Trial Mechs for veteran players since they also may use them if they spend all of their cash on a new mech, or expensive upgrades, and then take significant losses while not being able to cover their Repair and Rearm costs. Plus the Trial Mechs are a way for people to test drive new mechs that they do not currently have access to.

[MaddMaxx]

OriginalTibs, on 12 December 2012 - 06:26 AM, said:

Matchmaking IMV is considerably more complicated in a skill-centric game. BV is a tabletop mechanism intended for a mathematical model where offensive/defensive effectiveness depends on random dice rolls. Pilot skill has little bearing unless you are also factoring the tendency to become disorganized or rout depending on values asserted for morale and experience.

For mechwarrior online they would have to not only factor tonnage and loadout, but actual win/loss ratios, aggregate damage dealt and taken, and kill/death ratios.

Because the effectiveness of a lance is contingent on voice comms as opposed to keyboarded chat (to fire or change direction I cannot be texting), being grouped would also factor.

Totally agree with how complex it could become. The K.I.S.S. principle just won't cut it when the time comes.

And if they did indeed use a Group formula as part of that aggregate, then actual Lances/Companies that play together ALL the time would have to be given additional consideration (less Mechs, less total weight or less BV allotted).

BV could work if it was combined/used in conjunction with, how many Mechs a Team could field. Either value could be finite and when reached (+/- some arbitrary %) make up the Team composition. Similar to saying only 780T allowed this drop, with no limitation on Mech #'s allowed, or only 6 Mechs and or a total BV of whatever.

The permutations are literally endless. How about when any one Chassis reaches a predetermined BV, the other Team gets to field 2 Mechs that equal that same BV total?

It will be heavy duty QQ time with whatever the Dev decide, that much is assured.

Edited by MaddMaxx, 12 December 2012 - 08:10 AM.

[Rofl]

OriginalTibs, on 12 December 2012 - 06:26 AM, said:

Matchmaking IMV is considerably more complicated in a skill-centric game. BV is a tabletop mechanism intended for a mathematical model where offensive/defensive effectiveness depends on random dice rolls. Pilot skill has little bearing unless you are also factoring the tendency to become disorganized or rout depending on values asserted for morale and experience.

For mechwarrior online they would have to not only factor tonnage and loadout, but actual win/loss ratios, aggregate damage dealt and taken, and kill/death ratios.

Because the effectiveness of a lance is contingent on voice comms as opposed to keyboarded chat (to fire or change direction I cannot be texting), being grouped would also factor.

BV in table top does not take into account tactical skill or depth of knowledge of the game, either. However, I see where you are going with this and agree.

[Carnifex Requiem]

OriginalTibs, on 12 December 2012 - 06:26 AM, said:

Matchmaking IMV is considerably more complicated in a skill-centric game. BV is a tabletop mechanism intended for a mathematical model where offensive/defensive effectiveness depends on random dice rolls. Pilot skill has little bearing unless you are also factoring the tendency to become disorganized or rout depending on values asserted for morale and experience.

This is something I keep seeing and, frankly, it's making me crazy. In TT, the base BV is based purely on the mech, but the final BV includes a multiplier based on the pilot's skills (Piloting 5/Gunnery 4 is the baseline, for those who care). This means that you could potentially set up a similar multiplier scale based on mech, chassis, and total XP to get a baseline on pilot skill level. It won't be perfect (especially for a free-to-play game, where experienced pilots are that much more likely to set up other accounts to get around this), but it's something that's a straightforward adaptation.

Note: I'm well aware that straightforward != trivial.

[Jakob Knight]

The one problem with using either tonnage or BV as a guide to limiting drop configurations is that they only work if all mechs are put on the same footing (i.e. all mechs' weapons fire at the same rate, torso twist the same rate, ect.). MWO is being purposefully designed such that every mech can challenge any other mech with an even chance of success (at least, that is the overriding goal of the Devs, it seems).

To illustrate, using either BV or Tonnage, a stock Jenner should have zero chance to kill a stock Atlas, and it should require 3-4 Jenners to produce an even fight. In MWO, however, the penalties imposed on the Atlas to counterbalance it against light mech pilots means that one Jenner has an even chance to down the Atlas, made all the moreso by the 'lagshield' and light mech damage modifiers that add to this.

So I would be very, very hesitant to try to use either method of limiting force selection until either all mechs performance is equalized regardless of tonnage or engine size, or the values are overhauled to calculate in all the changes produced in MWO for the purposes of equilizing mech combat performance across tonnage levels (basically a full redesign of the rating system would be needed).

Edited by Jakob Knight, 12 December 2012 - 09:36 AM.