Hey guys, today we discuss skill based matchmaking, a service that many, many have asked for from Wargaming.
Outro Music: Stranger Think- C418
Have a replay?
Join the Discord here!: https://discordapp.com/invite/QA7G9pr
Not ever going to happen, WG has more important things to focus on, like adding other shit to the game they said would never happen.
this would just end in high rank players taking way longer to find games
Not if it was based on exp rank within teams. That way a “1st in team” players is only ten away from a “last in team” player. And MM can simply take highest available instead of waiting for a suitably high match of players.
Maybe, yes, BUT, the new system wouldn’t need to be 12 high skill players vs 12 high skilled players. It would just need to make it so there are the EXACT or at least maybe plus minus 1 or the appropriate skill descrepancy between the two teams. So lets say Team A has 2 super unicums with around 2k-2.2k elo, the other team only has 1 BUT also have more in the “above average” elo of players.
And you could even make it so when you have to wait for X amount of time, the MMR system becomes more lenient when making teams. So lets say the first happens at 3 min queue time, then 4, then 5min etc. (just throwing out numbers) then if you’d have to wait magically for 10mins, you’d have the same kind of battles as nowadays, as far as bad matchups goes.
You just need even number of good players to be placed on both sides. It does not matter that a good player is playing a BB and another good player is playing a DD on the other side.If you have a unicum player, then place 2 good players on the other side.
I would be perfectly fine with that trade off if it means I am not held back by weekend player constantly.
I mean most game beat the crap out of veteran players because devs see they have a statistically higher retention rate.
Only stat that really matters in WoWs is damage done, until this changes you cant have real skill based matchmaking as for each class you need to take other stats into account, for BBs you gotta count potential damage, for DDs you need to take into account damage upon spotting and so on, and so on…once all of these extra stats are taken into account only then can you make a real skill based matchmaker with somewhat even teams.
Winrate is, in the end, the stat that matters. If you just stand there in a BB in a position where the enemy has to angle against you, you might be the player that wins the battle (considering the enemies havt to show broadside to your team so they can angle against you). Or you could protect a cap as a DD by simply existing near it so no enemy BB dares to get close.
at the same time there is no metric possibly measuring that – except winrate.
also damage done is a stupid metric. maybe somewhat relevant if you make it percentage based, but pure numbers are near useless. With 50k dmg I can kill 2 DDs or make a dent in a BB.
@OneHandBehind PR is the best way to evaluate a player
@OneHandBehind WR sux, specially now, where u get horrendous teams that can make lose 10+ games in a row. PR is a lot better. If 2 accounts starting the same day, one is a bot but his team wins 10 games can have better WR than one played by a real player but lost those 10 games.
@Nicolas Fidalgo and damage done is a good metric?
yes, winrate in a 12v12 game needs a lot of games to even out. But playing for a win is not equal to farming much damage (the main contributing factor to PR). except maybe if you’re in a smolensk or the like.
also in T10 games more damage is done than in a T6 game. doesn’t make me a better player though.
Argument could be made for basexp to be a metric, since WG at least tries to make every part of participation worth something.
In the end, winrate is what matters. In the best case there would be an elo-like/mmr system.
And ten games mean fuckall, yes. thats what you get when trying to measure performance in a game with 23 other people.
I think a Skill-Based MM is the *first* thing that does need to be done. If it was based on a similar system to ship selection, ie each team gets two T8 and T9 cruiser each, two T10 BBs, three T8 DDs etc etc, then each team (in theory) has the same player quality; so it puts a T8 DD with roughly equal PR rating on each team etc. Isn’t the next Ranked-Sprint being based on a skill-based system?
There are at least two other things that I think would need to change in conjunction.
Second thing is end the three tier spread. It was necessary when the game was much smaller in terms of player base, but we don’t need (For example) tier 8, 9 AND 10 ships to fill teams, there are plenty of ships that have no business fighting ships two tiers above (Emerald vs a Fiji for example!!). For me it should be either T8 and 9 OR T9 and 10 and the is applied to all tiers except T1.
Thirdly, I would COMPLETELY revise the XP-Reward system. I gather the algorithm for calculating XP is very complex, but the most heavily weighted factor is direct damage dealt. All it does is encourage stat-padding and selfish play. For me, the major factors should be the team-based play. So things like spotting damage, ships spotted, cap and defence points. I’d introduce a ‘Damage from spotting’ for BBs, so how much damage do they do to ships being spotted as well as ‘assist’ categories; things like using consumables for teammate benefit (Smoke for example) I’d also introduce kill-assists, which you’ve mentioned before and ‘saves’ so if a player saves a friendly from being destroyed by killing, covering or drawing fire. Any class, from sniper BBs to gun-boats DDs can still benefit and direct damage would still be a factor, just not the be-all-and-end-all factor.
All of this would encourage, and most importantly REWARD more team-based play, make teams more equal and not force double-up-tiering on ships which are little more than shell-magnets.
I agree. As a mainly german battleship player I find that tanking for my team is greatly underappreciated. I support my team pushing in and taking a cap at the start of the game and survive attracting 4.5 million pot dmg. It gives us an early advantage and contributes to a win, but it is worth almost nothing as far as personal rewards go.
If skill based MM is actually implemented atleast there will be no more sealclubbers at low tier….
“Every system is perfectly design to produce exactly the results that it produces”. The MM works the way it does because that is how WG wants it to work.
As you said, skill based mm would be better than what we have now. I think that it’s too much to hope for that we get a mm that would analyze your days win/lost rating and take pity on you though. 🙂
I must be missing something. I am a 50% WR player, and would love to be pitted against roughly my level of player. Because I don’t like to be trampled, and I also don’t like to be part of a team that is so good it ROFLSTOMPs the other team. So what is the point in manipulating your own stats? I want mine to be as high as possible, in other words win as much as possible but also have fun. Tanking your stats does not make it fun when you lose, OR when you win. I would love weighted MM. Most sports have divisions where you can be promoted or demoted as the case may be, and I want to be in a division made up of my peers, not my superiors or inferiors. Neither is much fun. A little bit better is ideal, because it allows me to learn. Weighted MM can do this. And I honestly don’t understand why anyone would WANT to manipulate this, even if they could (and they probably could, I agree). Someone please explain to me why anyone would.
skill base MM does not mean putting all the good players against each other and all the bad players against each other. Skill MM means equal number of good and bad players on both side to make it fair.
@VuHien2011 If you say so. I wasn’t aware there was one set definition of skill based MM. I am fine either way btw.
If you are a bad ass, good as hell player, wouldn’t it be fun to trample all over a bunch of less experienced player and have that record hitting high damage game? Well the people tanking their stats think so.
The whole “tanking your stats” argument doesn’t really matter when the majority of matches are just one sided af, it’s basically already happening on a team wide basis instead of just one person on a team.
One simple thing they could do is balance radar MM. Yesterday I had these two games where the enemy team got 2/3 radars and my team none. It won’t solve every problem, but all radars on one side is a massive advantage.
Ya thats ONE thing that really should have been addressed by now. Obviously the MM tries to match up the same ship or same type of ship on each side…..but their really should be extra factor that takes into account what tools the ship has access to. A Radar cruiser is arguably WAY more usefull then a NON radar crusier (team wise anyway and utility) Sure the french have SPEED and the Germans have GOOD hydro but who cares when the enemy can radar you from 10 to 12kms away with little to no warning.
just split the good and bad players evenly on the teams. I start using wows monitor and most of the time the outcome is decided before the game start . 2-3 Very good players in one team and very bad on the other team. Or 60% w/r cv against 40% cv…. feels like u dont have any input in your team win or lose atm
It’d be nice not to lose 10 games in a row despite playing the objective in every single game and getting ~70k damage at tier V, just because you get weekend players and random trolls/ bots.
A lot of people have wanted this for a long time. There are so many ROFL stomps now. I am a unicum player (super uni if you look at my last few thousand games) and I believe the main issue is there are too many players at high tiers who have no idea what they are doing. If you get a couple other players on one team who do – it usually means they will roll the enemy team. I am for the change to MM as suggested, but WG has no incentive to do so at this point, so we know they will do nothing. There are SO MANY issues with this game, but WG is deaf to the communies pleas. Game has SO MUCH potential, but is held back by the development decisions of WG. Unfortunate.
Queue time is not an argument either. With the current system, two noobs in a division together get 1 fun game in a month of queueing. I’m pretty sure they wouldn’t mind waiting an additional minute for a game to pop if their odds of a fun game increase from 1% to 90%
Game is dying a slow and painful death. MM is a major part of the problem along with wildly inconsistent RNG.
Thank you for talking about this. This is a discussion that is completely ignored and an issue that needs to be addressed. I for one am tired of the blowout matches. It makes the game feel rigged and rewards poor performance. I am also tired of playing people who have 5-6 thousand battles under their belts and understand every mechanic while I am still figuring out how every class and ship works. If other major gaming companies can get it right so can Wargaming. Enough with the fiddling of ships stats and mechanics that don’t need changing. Fix the damn matchmaking first!! Thank you Sea Lord for this video, you are the man!!!
Yes, absolutely to skill based mm. don’t know why they can’t have algorithm that factors in all those things (win rate/PR/rank) to have an even distribution
“so I’m an above average player” “yeah…im not great”
What is the benefit for a good player to ruin his stat just for the purpose to be in a good team. Most of the good players are proud of their stats and ruining their stats to be in favorable MM to have more victory to go back to their original stats does not make any sense.
Yeah it’s a reason I’ve been playing mostly co-op. And stopped playing tanks altogether. When I played tanks competitively it was frustrating af to have the w/n show me a green player as the highest ranking person on a team full of tomatoes, and the other side with 5 unicoms, 4 blues, and some others. Not on 1 match, but over and over and over.
Your email address will not be published. Required fields are marked *