[MMO] World of Tanks

Page 386 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Mar 10, 2005
14,647
2
0
The introduction of premium rounds for credits means that 5's can damage 7's, 7's can damage 9's. Can you take them on face to face... probably not.

Do you know how stale the game would be if all you had were 8's fighting 8's? Why stop there, better give the opposing teams the same amount of TD's, mediums, heavies, lights, and arty. That way it will be balanced... and utterly boring.

following that logic, chess would be better if player 1 had 2 queens to player 2's zero. such a game would have no legitimacy.
 

Ruptga

Lifer
Aug 3, 2006
10,247
207
106
you will never get me to agree with that. i'd rather wait 30 seconds to form a match of comparable opponents than instantly be dumped into a game on a massively uneven playing field. with my t7, smashing t5's is equally worthless as being smashed by t9's. this isn't competition, it's repetition of the inevitable.
Agreed.

Generally speaking those in bottom tier get smashed by those in top tier. Skill and/or (ludicrously expensive) gold rounds will take the edge off of it, but the difference between something with 1600hp that deals damage 320 at a time and something with 2200hp that deals damage 1050 at a time is simply huge, especially considering that the JP100 will bounce the KT with some reliability. The situation is actually a lot worse for a pz4 and an IS. Pz4 gets 420hp and deals 110, while the IS gets 1230hp and deals 390. With just a little bit of luck the IS will one-shot the pz4, quite possibly before he even takes aim.

WG could reduce the HUGE tier-to-tier differences in HP, armor, and alpha, but it would be a lot simpler to just reduce the MM spread to only one tier. A t9 would fight tiers 8-10, but a t8 would never see a t10. If they did that (lol) I'm sure I would take up t8s again, but as it is they just constantly get shit on by t10s, in large part because of top tier's constant power creep.

Disagree and disagree even more.
And I don't want a player skill based MM. It means I'll be placed with fucking scrubs all day, more so than I already am. No fucking thanks.
And agreed, with a caveat.

I would love to see MM account for player skill, but in the way that Starcraft 2 does it. It groups players into bronze, silver, gold, platinum, etc, and a platinum will never play with a gold, etc. Every month or so everyone qualifies again and your leagues are updated. This keeps the "play 4funners" down in their cesspit of incompetence, keeps average players fighting average players, and allows the best players to enjoy competent teammates and pitched battles, all while still allowing for nooby players to improve themselves and move up over time.

Now, there's not a lot to stop douchebags from throwing their qualifying matches to get bronze so they can noob stomp, but the system could be programmed to flag a person for movement to a higher league if they are winning by huge margins. It could also factor in their previous performance and rank them as whichever is higher (rank while qualifying or rank from pervious stats).

Another problem could be dividing the player base to the point that matches take too long to launch (more than two minutes). But, they've already split the servers... so just merge them back. And I would only have three leagues, bottom third, middle third, and top third, to keep each pool of players from getting too small.
 
Last edited:

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
as it is, i'm lucky to see more than 2 or 3 people on my team carry a 50+% win rate, including myself, so what is there to lose? all this would do is eliminate the matchups where 1 team is carrying 14 reds while the opponents are largely green/blue. that isn't a game of skill, it's a coin flip.

Again, no, since I never see instances like what you're describing on a frequent enough basis. Maybe i'll get 3/10 games like that and I'll lose those games summarily without caring much because those are the games you can't influence. I also don't care about complete stompfests (rarely happens) since those are boring with all the people stealing my damage.

I care about games where I play a big part in the influence, where if I had lost a flank the game would've been lost. Or if I let the scout through, our arty would be obliterated. Or like a replay I posted a few posts back, where if I had lost the duel with the T54E1 it would've been a complete loss for our team because our T95 is slinging fucking HE shells. I'm going to say about 40% of the games I'm playing are games I could've saved but lost because I'm not good enough, and 20% depended on good teamwork, and 40% are complete washes or curbstomps. Your suggestion will pretty much ruin this aspect of the game since it'll ensure I never have good teammates.

A better suggestion is mentioned above where you have leagues, but you have to again remember efficiency and WN7 is created by players and not endorsed by wargaming at all. Separating by win% is not the best idea since that can be padded by tank companies.

One thing you have to understand is as you get better at the game, you understand you can influence more and more games, and you'll feel less like you're being oppressed by the red and orange masses. And remember, sometimes that red 500 eff player is the key to winning a game. Seen it happen dozens of times where it comes down to a red vs red slapfest, and the lucky one wins.
 

mooncancook

Platinum Member
May 28, 2003
2,874
50
91
I don't expect MM to be perfect, even every player has a different opinion on what is perfect, but at least MM should try to put the same number of top tier tanks on each team. Like having 5 t9s against 7 t9s, we had no chance. And I just had a t7 game where our team had a T29, IS, T25AT, T20, while the other team had a T20, PzV/M10, AT15A. They had no chance. MM could have put the 3 mediums in one team and the 2 heavies on the other team. I've played this game for a long time, only recently I noticed that this kind of unfair match happens a lot more often. MM need to at least get the tanks match-up right.
 

stag3

Diamond Member
Feb 7, 2005
3,624
0
76
id' also rather it be more even skill wise.
i'm tired of blowouts of 15-5 or worse, shut outs
happens all the time
 

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
I don't expect MM to be perfect, even every player has a different opinion on what is perfect, but at least MM should try to put the same number of top tier tanks on each team. Like having 5 t9s against 7 t9s, we had no chance. And I just had a t7 game where our team had a T29, IS, T25AT, T20, while the other team had a T20, PzV/M10, AT15A. They had no chance. MM could have put the 3 mediums in one team and the 2 heavies on the other team. I've played this game for a long time, only recently I noticed that this kind of unfair match happens a lot more often. MM need to at least get the tanks match-up right.

This is MM fucking up, and I think WG has acknowledged that and has a fix ready in 8.8. This is because of the arty rebalancing, so I'll be cautiously optimistic before getting out my torch and pitchfork.
 

mooncancook

Platinum Member
May 28, 2003
2,874
50
91
This is MM fucking up, and I think WG has acknowledged that and has a fix ready in 8.8. This is because of the arty rebalancing, so I'll be cautiously optimistic before getting out my torch and pitchfork.

Yeah I think before 8.7 at least the arties will keep the heavies honest, now with 1 weak arty on each side we just roll right up in such an unbalanced game. I don't want to go back to old days of campfest with lots of arties, but they definitely need to rebalance matches.
 

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
I'm actually not referring to arties keeping other things in check, but more along the lines of uneven number of top tier tanks. This is, according to WG, caused by the arty being shifted around for battle weights.

AFAIK there's no plans to limit MM based on tank types, but heavies/TDs/mediums are weighted differently, and as of the arty change I believe TDs, especially tier 10 ones, need to have a very heavy MM weight to compensate so we don't get "5 foch155 vs 5 E-100s" instead of 5-arty-parties.
 

Possessed Freak

Diamond Member
Nov 4, 1999
6,045
1
0
I would love to see MM account for player skill, but in the way that Starcraft 2 does it. It groups players into bronze, silver, gold, platinum, etc, and a platinum will never play with a gold, etc.

Have you ever watched the queue for MM? I have seen the numbers of a single tier approach and cross 100 of waiting players a FEW times. Normally I have only seen 30-40 players of a single tier awaiting matches. That is across the entire NA East server.

If you want to add skill based MM this is what will happen: a full 50% of players will be in a 48% or lower category. So now skill based MM is looking at only 15-20 players in one tier for ANYONE at 49% or higher.

But let's be honest, a 49%er is no 60%er, so there probably needs to be more refinements. So they have to add other brackets of skill, somewhere in the low 50's, upper 50's, and then the upper echelon at the 60's. What percentage of players do you think are actively looking for a game at the same time that is say... at 60%? Let's just ballpark it at say 1% of the server population is at 60% or higher. As a skill based MM that would be impossible to place those players in a timely fashion with other 60%ers.

Now, this is what I have seen as objectively as I can. As a 60%er when I solo in mid tiers, I see a sea of red and yellow and myself on my team. The opposing side might have a single player or platoon of good players and the SAME sea of red/yellow.

It is VERY rare that I see a slew of purples on one side and reds on the other UNLESS it is due to platoons of horribads and platoons of unicums. If you are on the side of horribads, learn to platoon so at least it is a platoon of good+ players against the platoon of purples.

If you want skill based play, go into TCs or CWs.
 

Ruptga

Lifer
Aug 3, 2006
10,247
207
106
Have you ever watched the queue for MM? I have seen the numbers of a single tier approach and cross 100 of waiting players a FEW times. Normally I have only seen 30-40 players of a single tier awaiting matches. That is across the entire NA East server.

If you want to add skill based MM this is what will happen: a full 50% of players will be in a 48% or lower category. So now skill based MM is looking at only 15-20 players in one tier for ANYONE at 49% or higher.

But let's be honest, a 49%er is no 60%er, so there probably needs to be more refinements. So they have to add other brackets of skill, somewhere in the low 50's, upper 50's, and then the upper echelon at the 60's. What percentage of players do you think are actively looking for a game at the same time that is say... at 60%? Let's just ballpark it at say 1% of the server population is at 60% or higher. As a skill based MM that would be impossible to place those players in a timely fashion with other 60%ers.

Now, this is what I have seen as objectively as I can. As a 60%er when I solo in mid tiers, I see a sea of red and yellow and myself on my team. The opposing side might have a single player or platoon of good players and the SAME sea of red/yellow.

It is VERY rare that I see a slew of purples on one side and reds on the other UNLESS it is due to platoons of horribads and platoons of unicums. If you are on the side of horribads, learn to platoon so at least it is a platoon of good+ players against the platoon of purples.

If you want skill based play, go into TCs or CWs.

I think I already covered those objections?
Another problem could be dividing the player base to the point that matches take too long to launch (more than two minutes). But, they've already split the servers... so just merge them back. And I would only have three leagues, bottom third, middle third, and top third, to keep each pool of players from getting too small.

Merging the servers then dividing the whole server into three equal-sized groups effectively reduces the pool of players to 2/3 of what you have right now (assuming NA east and west are the same size). That will mean some longer waits, but right now a "long" wait is a bit more than 30 seconds. So maybe waits will be up to a minute, oh noes! The change will also improve the experience for everyone (except for middle and top-thirders who like to shit on the bottom third), which will result in better player retention and maybe even an increase in new players. Over time my changes would cause the player base to be larger, negating the issue of the divided player base, and making WG even moar gold. But it doesn't matter because this is WG we are talking about and inertia is king.
 

Binky

Diamond Member
Oct 9, 1999
4,046
4
81
AFAIK, the MM already definitely includes some form of skill measurement in its process. It's rare to see a bunch of green/purple names on one team, without some obvious balancing on the other team. Anybody with XVM knows that most of the teams are mostly red, but when your team is primarily green or better, the other team is almost always similar.

There's a limit on what you can do with the actual player base. I think the MM does fine, usually.
 

Chapbass

Diamond Member
May 31, 2004
3,143
88
91
I personally feel it should be one tier +/-. Even the KV-1, considered among the best if not the best T5, struggles bigtime against even average tier 7s. As boston said, I'd rather wait a little bit of time and have fun with it instead of bouncing constantly. Yeah, theres minor weakpoints, but its more or less an exercise in futility.
 

Chapbass

Diamond Member
May 31, 2004
3,143
88
91
AFAIK, the MM already definitely includes some form of skill measurement in its process. It's rare to see a bunch of green/purple names on one team, without some obvious balancing on the other team. Anybody with XVM knows that most of the teams are mostly red, but when your team is primarily green or better, the other team is almost always similar.

There's a limit on what you can do with the actual player base. I think the MM does fine, usually.

How accurate is XVM? I'm not using it, but some dude called me out for having a 42% win rate in my T1 Heavy and saying I was a noob, and yet all the stats sites as well as in game service record show my t1 heavy as 58% win rate and overall at 52%....

Could just be a typical WoT troll though...easily my least favorite part of WoT...you can't turn chat off. I swear if I wanted to play with a bunch of CoD tards, I'd go do that.
 

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
AFAIK, the MM already definitely includes some form of skill measurement in its process.

The devs have already stated they've experimented with it and it fucked up mm even more and so they scrapped the idea. The reason you find it rare to see a bunch of them on one team is because there's not a lot of good players, and even less of them solo pub. I think they have said it was about 1% of the player have winrates over 58%?

If you play late night you'll see them all platooned up, sometimes all on one side. It's always funny to see PURPL or -G- stacked all on one side while it's unnamed bads with 600 eff on the other.
 

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
Do you have any idea where you read that? I've always wanted to know what percentile I'm in.

This was a while back; some guy did a parse of the player population by crawling through the player list and just ran some simple math. Given how great (sic) the search function of the forums are (completely broken), I doubt I can find the post again. I don't remember if this was before the physics patch or not though, but I can't imagine the patch influencing win rates too much.

Does noobmeter have some sort of amalgamated stats from wargaming? If they do then that's probably the best place to look.
 

Binky

Diamond Member
Oct 9, 1999
4,046
4
81
The devs have already stated they've experimented with it and it fucked up mm even more and so they scrapped the idea.
I just don't believe this. I see too much consistency and balance between the XVM rankings on each side to think it's just random, regardless of what they say. It is FAR from even all the time, and I definitely see 5% win chances occasionally, but it's just too consistently (somewhat) balanced on a rough scale to call it random. I'm not saying that the MM cuts it as finely as XVM because I don't see that, but it at least has some sort of simplistic Good/OK/Bad rank that goes into the MM math. I don't see all red vs all green, and when there is more than one purple on one side, they are generally on both sides.

I would assume that at times of low population (late night), the possibility of extreme mismatching goes up, because the MM only has so many people to pick from.
 

Possessed Freak

Diamond Member
Nov 4, 1999
6,045
1
0
I think I already covered those objections?

And I made a nice statement regarding this interesting fact. Out of the global population the AVERAGE winrate is 48%. Which means a full 50% of the population is at or below 48%. The skill based MM in the lower third will be a complete riot. The 45%'ers and worse will just get more and more draws as there are multiple AFKs and bots per side. The mid tier will be unchanged from its current of play. The upper tier will have much more competitive games.

But in the end this is ALL that will happen: The population will go back towards 50% sort of. The upper tier play will tend back to 50% which will mean that they will eventually beat the ever living snot out of the 53%ers that have also started to tend towards 50%. The players that have been somehow on the losing side of the second tier are going to slide into the first tier and will face a bunch of complete brain dead losers on both teams. As one of the few active players this guy will end up being beyond frustrated.

Sounds like a great idea. No, random play remains random. I wouldn't mind a SECOND mode of play that is skill based to allow you to progress up the chain.
 

Ruptga

Lifer
Aug 3, 2006
10,247
207
106
And I made a nice statement regarding this interesting fact. Out of the global population the AVERAGE winrate is 48%. Which means a full 50% of the population is at or below 48%. The skill based MM in the lower third will be a complete riot. The 45%'ers and worse will just get more and more draws as there are multiple AFKs and bots per side. The mid tier will be unchanged from its current of play. The upper tier will have much more competitive games.

A valid point, but AFKs and bots are also an issue that needs addressed, so why not step up enforcement and actually root those losers out as well?

But in the end this is ALL that will happen: The population will go back towards 50% sort of. The upper tier play will tend back to 50% which will mean that they will eventually beat the ever living snot out of the 53%ers that have also started to tend towards 50%. The players that have been somehow on the losing side of the second tier are going to slide into the first tier and will face a bunch of complete brain dead losers on both teams. As one of the few active players this guy will end up being beyond frustrated.

You would be right, but the MM doesn't have to be linked to win rate. Indeed, I never said a thing about win rate. If it were done right it would be based on a few important stats that, together, are hard to fake, exactly like noobmeter's Performance Rating. The program could also apply a final modifier that accounts for the fact that this person played poorly but did so against the top third of all players. Also, the leagues would be made of the bottom, middle, and top thirds of all players, not ratings. If the new system made a more competitive environment and lowered the average rating the system would simply shift its mean rating down, ensuring that each of the leagues always have an equal number of players.

Sounds like a great idea. No, random play remains random. I wouldn't mind a SECOND mode of play that is skill based to allow you to progress up the chain.

You are only saying that because you are sure it wouldn't work. It can work, SC2's continued success is proof of it, as are my walls of text. Further, matches would still be random within each third. And splitting the server into new MM and old MM pools would probably mean that neither MM would have enough users to function as they have been designed to.

If this ever happened you would probably try the beta, just to validate your bad feelings about the system, but then you would find that it's better. You would still get rolled from time to time, but you wouldn't have to put up with nearly as many deadbeat AFKs, lemming rushes that die within four minutes, or HE spammers in T95s. You would find that far more teammates have triple-digit IQs, that they respond to requests, and that they even have good ideas of their own. More matches would be battles, instead of routs, since teams' relative skills would be more closely matched. The game would be more interesting and less frustrating. Other than a potentially longer (but still short) queue time, I don't see a single insurmountable drawback to creating leagues.
 
Last edited:

Ruptga

Lifer
Aug 3, 2006
10,247
207
106
Those sites give a lot of useful information, but I haven't been able to find my percentile rank. I know I have won 55.74% of my matches, but where does that place me within the player base?
 

PowerYoga

Diamond Member
Nov 6, 2001
4,603
0
0
I just don't believe this. I see too much consistency and balance between the XVM rankings on each side to think it's just random, regardless of what they say. It is FAR from even all the time, and I definitely see 5% win chances occasionally, but it's just too consistently (somewhat) balanced on a rough scale to call it random. I'm not saying that the MM cuts it as finely as XVM because I don't see that, but it at least has some sort of simplistic Good/OK/Bad rank that goes into the MM math. I don't see all red vs all green, and when there is more than one purple on one side, they are generally on both sides.

I would assume that at times of low population (late night), the possibility of extreme mismatching goes up, because the MM only has so many people to pick from.

eh, if you say so. I really doubt the developers would put in as much effort into the MM as you think they do, seeing as how they barely give a rats ass about player opinions. Generally I take the developers' words over players unless there's some very solid statistical proof. I've seen streaks of me being on red and streaks of me being in the company of unicorns... it's just random. Most games will have 2-3 guys over 50% and that's it, the rest is a slap fest.

Those sites give a lot of useful information, but I haven't been able to find my percentile rank. I know I have won 55.74% of my matches, but where does that place me within the player base?

My guesstimate is top 10%. I don't think any sites does a full crawl on users in the WOT database though.
 

stag3

Diamond Member
Feb 7, 2005
3,624
0
76
lol just had the best game in my 268, had (2) 704's covering my side along with a t62a that lit things up for us.
complete domination. we did not even camp and snipe, we pushed forward a little behind the t62a and shit died as soon as he lit them up