72 replies to this topic

[Oberost]

Anjian, on 30 December 2015 - 11:50 PM, said:

In World of Tanks, average or normalizing means you get at least a 44% ratio. If you are getting a 50% ratio, you are good. Getting 52%, you are very good.

I disagree, with a 44% WR in WoT you are a potato. Period.

You can only consider yourself to be good from 54%-55% onwards.

[Pika]

Until you have about 1000/2000 games under your belt, don't worry about stats. Right now you're not on an even playing field. You won't have many modules after so few games, I doubt you have more than 3/4 mastered 'Mechs and I seriously doubt you have a full set of modules as well as the all important Seismic.

To use a WoT reference, this would be like comparing a skill-less crew against a player who's firing APCR with a maxed out 2 skill crew. You don't stand a chance so comparing each other there is biased.

Come back in about 1540 more games, then worry about stats

Oberost, on 31 December 2015 - 04:26 AM, said:

I disagree, with a 44% WR in WoT you are a potato. Period.

You can only consider yourself to be good from 54%-55% onwards.

I believe if you're starting on the road to improving 45% is the start point if you're below that. Get to 45% before dealing with specific holes in your play style. Though to be totally fair the obsession with XVM, Stats and percentages utterly killed that game in my opinion. Stats are important but they can't account for everything. The WoT Community obsesses over the colour of a line of numbers more than the actual game itself, to the point they just "Play around" in games with lots of red because "lol we lost anyways."

And before I get called salty or anything I was "Dark Blue" rating on my overall, which is considered "Great" and in the 99 percentile. At this point I grew to detest the community's obsession over numbers. Is why I like here so much. K\D means so much less. All about that Damage Done!

Edited by Pika, 31 December 2015 - 04:42 AM.

[Bilbo]

adamts01, on 30 December 2015 - 11:44 PM, said:

If you're consistently doing well and losing then you have to bring cheesy meta builds and carry harder. Also, top damage doesn't mean you're doing what's best for the win. Not saying this is you, but it's easy to sit back, pad your damage and kdr while your teammates are mixing it up on the front lines. I'm frequently at the top of my matches in fun mechs but my W/L is close to 1. My try-hard Timby has a 1.93 W/L after over 1,000 drops.

Edit: This was almost entirely in solo drops as well.

If you are consistently doing well and losing, you are on a run of bad luck or you aren't actually doing as well as you believe.

[adamts01]

Bilbo, on 31 December 2015 - 04:45 AM, said:

If you are consistently doing well and losing, you are on a run of bad luck or you aren't actually doing as well as you believe.

As you move up the food chain, you can't just "do well" and win. There might be 6 noobs, 4 bads and 1 decent player on your team. You pretty much have to double or triple the team average to secure a win. My experience is from the Oceanic server, where the skill spread is much worse than NA or EU. Here, If you don't do as much damage as the lowest 6 players combined, you're losing.

[Anjian]

Oberost, on 31 December 2015 - 04:26 AM, said:

I disagree, with a 44% WR in WoT you are a potato. Period.

You can only consider yourself to be good from 54%-55% onwards.

Wargaming's patented MM algos looks at your recent string of games, sees your winning percentage, and if you are past 50%, will match you more with potatoes, so no matter how good you are, you are about to endure a long string of losses.

Note it does not look at your overall lifetime W/L ratio. What it looks is your recent batch of games, like your last 10 games, regardless of your overall lifetime W/L ratio.

The MM on MWO isn't trying to equalize stats for player W/L ratios. It is trying to reduce stomps instead, to achieve closer games. Its objective is not to give players 50/50 W/L, but to achieve games where you get 12 to 11 deaths, or to reduce stomps, at least what is the mathematical definition of a stomp (perhaps at least 50 to 60% deaths of the winning team if the losing team loses 100% of its mechs).

Edited by Anjian, 31 December 2015 - 06:36 AM.

[Hotthedd]

Tarogato, on 30 December 2015 - 11:20 AM, said:

Your win/loss ratio should correlate roughly to your kill/death ratio. At least, according to my research. So if you're bad at the game, you'll lose more. If you're good at the game, you'll win more. If you're good at the game and losing more, then you are a statistical anomaly and can take comfort in the knowledge that it will come to pass and you'll be out of the rut eventually.

Another statistical anomaly here.
My KDR is barely green (possibly red by the time you read this,as I am levelling a new 'mech), but my W/L is around 1.75/1

[El Bandito]

adamts01, on 31 December 2015 - 04:59 AM, said:

As you move up the food chain, you can't just "do well" and win. There might be 6 noobs, 4 bads and 1 decent player on your team. You pretty much have to double or triple the team average to secure a win. My experience is from the Oceanic server, where the skill spread is much worse than NA or EU. Here, If you don't do as much damage as the lowest 6 players combined, you're losing.

I guess the enemy side always has more skilled players cause MM hates you? If the enemy team is just as bad as your team on average, you don't need to carry that hard.

Hotthedd, on 31 December 2015 - 06:32 AM, said:

Another statistical anomaly here.
My KDR is barely green (possibly red by the time you read this,as I am levelling a new 'mech), but my W/L is around 1.75/1

Depends on the type of mech one mains. My WLR is 1.31 and my KDR is 1.58 so I think they are pretty correlated.

Edited by El Bandito, 31 December 2015 - 06:37 AM.

[Hotthedd]

El Bandito, on 31 December 2015 - 06:33 AM, said:

Depends on the type of mech one mains. My WLR is 1.31 and my KDR is 1.58 so I think they are pretty correlated.

True. I'm not arguing with the logic, I'm just sharing that outliers always exist. (There always has to be SOMEBODY )

I'm sure my love for piloting mediums (HBK-4SP brawler build usually) has something to do with it...

[Almond Brown]

Even a 5 Kill 800 damage Match by 1 player does not mean a Win has to happen. Not sure how it doesn't happen but it happens, I know. Then the opposite happens. Some Matches you're the Man/Lass and you Lose while in other Matches, you don't factor in at all (<100 for whatever reason) and you get a Win. It is kind of funny really. ;(

[Nightmare1]

I say no to the 50/50 W/L. Frankly, I don't really care if I win or lose as long as it was a good match. Back when the 50/50 W/L was the rule of the day along with Elo, the matches were frustratingly awful. If you won a couple in a row, the MM would stick you into a couple or several where you'd get ROFLStomped in order to bring your W/L back down towards 50/50. I hated that and still do.

Currently, I'm really enjoying the current MM. I don't see stomps very often. My W/L hasn't moved much at all, but my K/D is up. Overall, I'm enjoying myself a lot more than I did during the days of Elo balancing with the 50/50 W/L rule.

So I say, let's not walk backwards.

[adamts01]

El Bandito, on 31 December 2015 - 06:33 AM, said:

I guess the enemy side always has more skilled players cause MM hates you? If the enemy team is just as bad as your team on average, you don't need to carry that hard.

From what I see, there are 2-4 good players on each team that do almost all the work, the rest are just along for the ride. If you're one of the two good players and don't bring it, you'll lose. There are great matches some times where everyone on both sides got 200-300 damage and it was down to the last few mechs, those are the good matches. I agree with nightmare1 concerning those games. I'd rather have 10 losses in a row of good fights than 100 12-0 wins.

[Thomas G Wolf]

My K/D ratio went down from 1.4 to .096 but my win ratio is between 51 and 52% and I am a pure pug solo player and I spend ALL of my time here leveling mechs. While leveling the less better mechs my K/D ratio dipped because I just could not bring enough firepower to bear or the mechs were so STOCK bad that nothing I did could help, my average damage is in the high 200s, that is from playing the mechs from the first game till they are mastered, as we all do I have games where I {LT-MOB-25} up bad and die in the first seconds of a game and then I have the Godlike games with 1k damage plus and just have to look at a mech to kill it.
I personally have stopped looking at my K/D ratio because I have come to accept the fact that it is Mech and weight class dependent , it does make a small difference if I run around in a mech with 1ER large or 1AC 5 or in a 8 med pulsa, the main thing for me is do I have fun or not, and strangly enough I have fun no matter what Mech I drive, I just keep on fiddleling with the build till I am happy with it and it is fun for ME to play.
This makes me a firm tier 3 player because I cannot for the life of me be arsed enough to play pure Meta builds, winning is fun and my DNA requires me to win, but winning with a non meta diffcult build is the greatest fun and for me equals the sting of losing when not "doing everything for the win" build, had enough of that shite in real life, time to have some fun, comes with age I guess, on the other hand you DO NOT want to meet me on a Battletech Board, that is where real life starts......

[RussianWolf]

DarthRevis, on 30 December 2015 - 05:06 PM, said:

Do you have the exact equation/formula they were using? I'd surely like to see it! Maybe would shed some light on a few things for us lol.

I don't believe anyone has the exact formula,

But from what Russ and PGI has said through the different phases of ELO and now PSR......... Their ELO system was set up where the MM used it to predict the outcome of the teams, even though that is a flawed system unless the teams are static groups. Its tried to force a 1/1 w/l and kept most people I think in a fairly close range of that (maybe 0.8 to 1.2???).

PSR is supposed to be about quick matchmaking. So i'd image that it picks the first 24 players in whatever level +/-1 and then go. My search times since they did this have been very short compared to the old days. I can image T1s and 5s may have to wait a bit longer than a T3, but all should be shorter.