Some of you will have noticed that player pages now include a "WN8 Rating (alpha-test)" in place of "WN6 Rating".
This is an early testing version of the WN8 Rating being developed by Praetor77, bjshnog, and many others (this is not an exhaustive list).
Brief FAQ follows, subject to be updated later:
Q. Is it the final WN8 rating?
A. No. The rating is only in testing stage, and subject to changes by the authors.
Q. How does it work?
A. It uses the same kind of "expected stat" calculation, and comparison with the actual attained stats as NoobMeter's Performance Rating uses for damage, but it applies it not only to damage, but also to spots, frags, defense points, and win rate.
It then applies a number of caps and weights to the resulting values to attain the final rating.
The exact formulas are:
rDAMAGE = avgDmg / expDmg
rSPOT = avgSpot / expSpot
rFRAG = avgFrag / expFrag
rDEF = avgDef / expDef
rWIN = avgWinRate / expWinRate
rWINc = max(0, (rWIN - 0.71) / (1- 0.71))
rDAMAGEc= max(0, (rDAMAGE-0.22) / (1-0.22))
rFRAGc = min(rDAMAGEc+0.2 , max(0, (rFRAG-0.12) / (1-0.12)))
rSPOTc = min (rDAMAGEc+0.1 , max(0, (rSPOT-0.38) / (1-0.38)))
rDEFc = min (rDAMAGEc+0.1 , max(0, (rDEF-0.10) / (1-0.10)))
WN8 = 980*rDAMAGEc + 210*rDAMAGEc*rFRAGc + 155*rFRAGc*rSPOTc + 75*rDEFc*rFRAGc + 145*min(1.8,rWINc)
The coefficients for tanks used to calculate all the exp* ratings in case of WN8 are published here:
http://www.noobmeter.com/tankList
Edit - formula updated to version from 10th of October
Q. Is it in your opinion a good rating?
A. Yes.
While some of the parameters may still need tweaking, the ideas are extremely sound, because, same as NoobMeter Performance Rating, it considers individual tank performance.
Q. Is this rating developed by NoobMeter?
A. No. The rating is developed by the authors as listed above.
NoobMeter is likely the first (and so far only) site supporting the rating, and thus helps in the testing of the rating, but, like the WN7 and WN6 ratings before this, it is not a NoobMeter rating. Only the Performance Rating is a NoobMeter rating at the moment.
Q. Why did you remove WN6? It is still used by XVM!
A. Each rating I use takes up database space (most importantly index space, which needs to be kept in operating memory). Therefore, NoobMeter can only support a limited number of ratings without needing to upgrade to a more powerful database server (which involves additional costs).
Also, WN6 is in practice extremely similar to WN7, so I suggest everyone just uses WN7 instead.
Overall, WN8 is a significant improvement over WN6, so it is time to move on.
Q. What is the scale?
A. Expanded compared to WN7 in both directions, but with the same average. Bad players get lower ratings than they did for WN7, good players get higher ratings than they did for WN7.
Please refer to the following for how NoobMeter assigns player ranks to different WN8 values:
http://www.noobmeter.com/calculatorDescriptions
Q. I have comments / objections / etc. about the rating!
A. You can post questions about the rating's implementation in NoobMeter here, but for anything related to the actual rating, please contact the authors of the rating in the official WN8 rating thread. Thanks.
This comment has been removed by the author.
ReplyDeleteAdded a brief response above
DeleteSome people think its hard to deciede who is low tier killer or not.
Deletebecause you can see some players play a lot at tier 1~3 but you cant say they can play well at tier 10.
and problem is whan you see to someone who play only 1~3 tier playing and got a huge rating, but that rating is not same as 10tier high rating. that is what i think... :)
Performance Rating handles low-tier seal-clubbers rather well, I think.
DeleteI cannot speak for whether WN8 does or does not, I didn't see any such checks and balances there yet, but maybe they will be added later.
Okay, i have some example rating who play a lot at 1~3 tier and got a high wn8 rating.
Deletehttp://tanks.noobmeter.com/player/kr/%EC%95%84%EB%94%B0%EA%B7%B8%EB%9E%91%EA%B5%90/3000009574/
http://tanks.noobmeter.com/player/kr/%ED%8C%8C%EB%A9%B8%EC%9D%98%EC%9D%B8%ED%98%95/3000328101/
Thats why people don't believe only wn8 rating, people check average tier too. Because noob killers are not good at higher tier battle. I hope it will help your work.
Yes, you will notice they have "Great" WN8 rating but only "Good" Performance Rating.
Deletehttp://tanks.noobmeter.com/player/kr/%EA%B9%80%EB%8B%A4%EC%86%9C/3000214852/
Deletethis too.. little bit high tier but still confuse us.
Korean tankers think Wn8 is good rating system, but people want to see this rating only, without average tier.
DeleteThat's what people think. :)
This is really well work i think.
Deletehttp://tanks.noobmeter.com/player/kr/%EB%98%90%EB%B4%89%EC%9D%B4%EC%99%80%EB%8F%8C%EB%8F%8C%EC%9D%B4/3000397292/
And more example
Deletehttp://tanks.noobmeter.com/player/kr/Pushechnoye_Myaso/3000030000/
http://tanks.noobmeter.com/player/kr/%EB%A6%AC%EC%B0%A8%EB%93%9C%ED%8C%8C%EC%BB%A4/3000210013/
http://tanks.noobmeter.com/player/kr/%EA%B3%B0%EB%8C%95%EC%9D%B4/3000231786/
http://tanks.noobmeter.com/player/kr/%EC%8F%98%EB%8B%A4%EB%A7%90%EA%B3%A0%EC%BD%94%ED%8C%8C/3000328823/
http://tanks.noobmeter.com/player/kr/_Kim_Min_Ji_/3000466038/
http://tanks.noobmeter.com/player/kr/%EC%99%A0%EB%95%85%ED%81%AC/3000485232/
http://www.noobmeter.com/player/kr/%EB%8C%80%EA%B0%80%EB%A6%AC%ED%83%84/3000075021/
http://www.noobmeter.com/player/kr/%EB%88%88%EB%B6%80%EC%85%94/3000103777/
I hope it will help you :)
yea can i have a short description of efficiency pleas?
DeleteI dont think that calculating winrate is good forr a rating since one can get a really good wr at playing TCs a lot but might otherwise be a net so good player or one can have a low winrate and be very good player
ReplyDeleteyou can have a high win rate in pub matches if you are good in your tank...TC can inflate that for players, but I have experienced great success with win rate in tanks that I am exceptional in without platoon or TC inflation
DeleteWin rate should not be used at all. All outside influences to player must be controlled or it is not statistically viable rating. A bad/mediocre/beginner player can achieve high win rates simply by platooning or TC:ing a lot with good players.
Deletesince WR is a Team factor which one cannot influence only by himself it is wrong in my opinion to count such in a PLAYER based rating
DeleteIn my Opinion using the winrate is necessary to define how good or bad a player performs.
DeleteExample. I once went YOLO-rushing into the encounter cap on Steppes rammed one tank, shot another and thereby not only prevented the enemy victory by capping but secured the victory of my team a few minutes later. For me the results were rather medicore. Instead I just could have played more cautiously, placed more shots into visible enemies, would have survived and gotten a better rating despite not being a teamplayer. IMHO this would be weird, wouldn´t it?
You can influence the outcome of battles - not 1 out of 5 maybe, but for sure 1 out of 20.
The thing is though with WR is that you can be the best team player in the world, and also be an amazing player yet the chances of you winning if you are 14-0 down? Obviously the chances increase the better you are but still i dont think it would increase enough to warrant being a good display of how good a player is.
DeleteI mean i dont know what it's like on other servers but on EU the amount of (to quote a common term) 'Siemas' which are in each match, the chances of getting a good team are so small i wouldnt be surprised if noone had above 50% by the end of the year
The argument that you are losing battle because the server has mostly poor players is extremely poor logic. There are 2 teams per battle. Therefore your odds of being on the team with the worse players is 50%. Therefore, solely based on other peoples skill levels, you have a 50% chance of winning any battle at the start. Therefore math tells us that having a win rate of over 50% over the course of thousands of battles is says you're influencing the battles to your favor. A win rate of less than 50% says your adversely influencing the battles. That is the unbiased math.
DeleteThere are a lot of elements to team play that are impossible to measure using only damage dealt , capture points, etc. There are some good factors that or not measured at all current including potential damage taken which can make a huge difference in battle. This is why its extremely important that any scientific evaluation of players include win rate. Yes, we all get those battles that are not win able, but I wouldn't have ever killed 11 tanks in one battle if I always had fair teams. Marginal players can excel when they're not outgunned. It's the exceptional players that can carry a poor team despite being outnumbered 6 to 1.
Your logic is flawed. Winrate is a poor evaluation system.you seem to forget match maker is tank weighted not player skill rated. You can have tier 4 scouts in a tier 8 battle that can do nothing but spot.or that team that has 3 arty vs 2 on the other' or that team that has 3 op kv1 against 2 t1's and a bdr. Win rate is a horrible evaluation equation
DeleteI agree with Daniel Minotto,
DeleteI think winrate is a necessary statistic because if we assume winning the matches is the main objective of World of Tanks then there are actions which are very difficult if not impossible to calculate that influences a teams chance to win. However the winrate of a player catches these actions.
Players influence every match they play, the more matches they play the more significant and accurate the winrate becomes.
I think this is an argument that most people with sub-50% win rates try to use. Personally with 15k games and a 52% overall win rate, and a 55% 60 day win rate, i will attest to the fact that over time you get better, and your win rate goes up. So personally i see win rate as a good metric to use in your statistical evaluation of overall player skill.
Delete"one can have a low winrate and be very good player"
DeleteI do not think this is true at all! Can you name even 1 such player? The goal of every player in a WoT battle should be to help their team win. If a player is getting damage and kills and still losing over a large number of games (anyone can have an unlucky streak) then there is something fundamentally bad about their playstyle. Perhaps they always camp in the back and never support their teammates.
I agree with the last comment. I have been a part of many battles where I've made excellent choices and turned the battle in our favor and was responsible for the win. I've also had many battles where my poor choices have caused a loss. I've also seen many excellent rated players hang back and do very little (hoping to scoop the kills at the end), the problem being that they hung back too long and ended up with too many enemies to effectively deal with. And we lost. Choices and skill are crucial in every game, and being an exceptional player who not only goes for high damage, but watches for opportunities to turn a bad situation around speaks volumes for win rate. I hate to bring it up because he's an arrogant ass - but go look at Garbad's winrate experiment. He played nothing but solo pub matches with a stock tank and managed (over hundreds of battles) to achieve a very high win rate (in the high 60's I believe). You can't really argue with data like that.
DeleteIn my opinion WN ratings should also use the infamous WoT bullshit RNG, it would be more realistic. No matter how good you are, if the RNG decides to screw you, you'll do less damage/kills and you will look bad. How about that?
DeleteRNG isn't so significant that that it can't be overcome by good tactics or solid play. Any experienced wargamer will tell you that and be able to prove it over and over again. Everybody is subject to random numbers and have the same chances to succeed or fail based on the parameters of the game.
DeleteOh geeez. The RNG is a killer in the game some times. There are times when a Tier 10 TD can't pen a LOLTraktor. Seriously. WoT said they were going to remove it and didn't. If anything it is worse than ever. So one game you 1 shot everything in sight and another you can't pen anything. Which then goes to the win ratio. In pub matches it is pretty worthless. There are times that a single tanker can influence the overall outcome but that is rare. In team matches it is about TEAM. You can be the best player in the world and if the rest of the team is horrible you have very little chance to win. Then you see these "unicoms" come in and play like crap. Oh look. 80% win rate! They must be good. Well, they might be in CW and TC but not in pubs. And then there is arty. You depend too much on the team giving you time to do your job. If that fails you may not even get a shot off. However, the more experience you have overall, the better your stats should get. Mine have steadily gone up as I play more. That is what you need to look at is a trend. A good player's stats will be going up. And finally, if a person started out bad and then has like over 10k games, that tanker has a REAL hard time getting win rate up. Once again, the trend is what matters. And of course a lot depends on what type of tank is being driven. If you have a light tank, things like spotted should be weighted more than say base defense. Of course if a tank does something out of the ordinary it should be "rewarded". Like T92 capping.
DeleteIMO Damage should be a key factor in ratings. Win rate is useless, as many have stated it's too dependent on TEAM performance. Kills are important also, but if I can reduce my enemy from 100 to 10 or less percent - and do it to several tanks in a battle... My ISU can drop a good deal of tanks to their knees in a single shot, often followed by a team mate finishing the job. I usually top out the stats in DAMAGE. Granted the enemy still has a working gun, but they are now a rock's throw away from death. So why isn't damage considered far more when looking at a player effect on the battlefield?
DeleteIs this rating indicate LT players' skill more accurately than the others?
ReplyDeleteThe WN8 rating will indeed reflect light tanks and artillery players' skill much more accurately since it is based on the expected performance of a player, which is in turn based on the tanks which the player plays. For more info, see the WN8 thread on the wotlabs forums.
DeleteYes, it does reflect a LT drivers performance better. however it only raises your score up a little. Once the WN8 version comes out where you upload your dossier you will finally see your true rating. Currently the standard WN8 doesn't have spotting damage in the formula, but the uploaded dossier version they are coming out with will. So some sites will support this version so you can see your "real" WN8, but 3rd party programs like XVM wont be able to use it.
Deletewith the dossier version some players scores are going to raise a lot. because good players who play on the front line also avg good amounts of spot damage. most unicums always sit in back never getting any spot damage so their scores wont really change.
We won't actually be using spotting damage in the dossier version, unfortunately. The dossier files only contain data for battles since 8.8, so it's effectively useless for an overall rating.
Deletehm.. sounds good WN8.
ReplyDeleteMy recent WN8 (22 battles) is ~ 2.5k. Hm, nice nice :D
Dunno why, but I like it, as long as it shows good stats for me.
Hey ... could u explain the difference between those two players
ReplyDeletehttp://www.noobmeter.com/compareplayers/eu/all/EmbryonicJourney/gonnaFUup
both are unicums but the difference in Wn8 is huge..over 1k why is that? (wn7 is close...Performance and Eff not)
GonnaFu has more avg frags and better win rate...Embryonic has more defpoints and spots and higher avg tier...
I think the difference in Efficiency is cause of defpoints and spots
but i really dont know why the performance and wn8 of GonnaFUup are much lower than Embryonics.
Cheers
Look at average damage on GonnaFu and you know why Embryonic have better WN8.
Deletebut thats cause of the much lower tier --> 4,3 vs 6,8
Delete950 dmg on tier 4 is very high
This comment has been removed by the author.
ReplyDeleteI like ratings of any kind but it messes me up to think that i get penalized for playing with a T-46 (tier 3 light) because its one of my favorite tanks in the game. I have some experience so i end up with twice the damage as most other players in tier 3 battles and sometimes more xp than the highest on the winning team. SHOULD I RECEIVE A LOWER RATING FOR PLAYING A TANK I ENJOY PLAYING?
ReplyDeleteI'll give you simple example.
DeleteLet's say you're a professional 100m runner, but you're still way behind worlds top 1000 runners. Now you compete agains children from primary school and you want to be treated as Usain Bolt just because you enjoy it? Start playing on high tiers against people who know game mechanics and earn your Rating there not be beating children.
You can always create your own PER (Pedobear Efficiency Rating) and enjoy your high stats on a tank you like playing.
when i kill a full IS3/ FCM / IS6 (or even tier 9/10) tomato in my E75 cause they have no clue where to aim and I get around 1.5k free dmg...where is the penalty for that :D
Deletelow tier games are not easy..u need lots of map awareness, know how to use bushes, quick reactions etc cause EVERYONE who shoots at you will pen you...
DeleteThis comment has been removed by the author.
Deleteor where is the penalty for all the T54 or object 268 statpadders who drive around with there huge DPM and use prem shells only (with 450 pen for object for example)...penalties make no sense..my opinion
DeleteOn lower tiers usually 14 of 15 players are superb tomato. On higher tiers noob ratio is smaller. I also think that T54 or object 268 stat padders are in the same category as n00b tier seal clubbers. As long as there won't be some statistic that can show gold ammo usage all ratings will describe wallet rather than skill.
Deleteunless you're on the EU server of course in which case the chance of tomatoes increases after tier VIII :)
DeleteThe formula was made by elitists for elitists. they will make the formula so that they can stay at the top while keeping the bad players down. Socialism anyone?
DeleteAnyway from the start of WN? it has catered to good players who platoon. Not the solo player who has less impact on winrate. Its why you see unicums with such high winrates.
My WN7 is over 2300 yet my winrate is 56%. why? because I do nothing but solo and I don't play lower tiers or platoon to boost winrate or kill count
2300 wn7 and 56% sucks.
DeleteYou are just a statspadder. Not a teamplayer.
An interesting point about using Gold Ammo. As there are Mods out there which inform players of what type of ammo the got hit with and from whom, can I safely assume that this data is actually available to base WNxx calculations upon? My reasoning is that if a player consistently uses Gold (Premuim) Ammo then surely his damage will go up considerably. Of course this theory is based on the fact that the player knows how to aim and shoot, but you can see where I'm going with this....
DeleteUnfortunately the mods use in game data that is transient - i.e. not available after the fact. to gather all this data would require each player to install a mod that transmitted that data as it was received - an unfeasible feat i imagine.
DeleteWhy not just lose the performance or eff rating?
ReplyDeleteBecause:
Delete* Performance Rating is IMHO more accurate than WN6/WN7.
* Efficiency Rating is more widely used than WN6.
* WN6 doesn't have any real advantages over WN7.
Thanks for doing this- these ratings do enrich my WoT experience, and help inform my play.
ReplyDeleteQuestions- what are the different rating categories and their rating thresholds (e.g., "poor", "fair", "good", etc.) for the different rating schemes, especially Win8? How are they determined, and are they dynamic? For those who are motivated to achieve the next category, this is good to know.
http://www.noobmeter.com/calculatorDescriptions
DeleteThanks for this. How often do you anticipate changing the thresholds, based on updated accumulated statistics?
DeleteCheers
Rarely.
DeleteThere are too many flaws in the ratings for them to have as much importance in the game as they have. Playing as part of a good platoon will help improve a players stats, having more matches within your own battle tier also helps, certainly in terms of damage and kills per game. Are the tanks stats based on the average across the whole WoT community, if a player does more damage and gets more kills etc than the average, then he will be better than the average player, I guess this is how it works.
ReplyDeleteActually a bad player playing in a good platoon will have bad stats. If you are outclassed by your toonies and its not close, the stats will show it.
DeleteI have an offtopic comeent:
ReplyDeletedidn't find any any "support" button to report a bud, didn't recall the profile of one of admins on KRF.
Where to report?
I'm afraid I have no idea.
Deletethe this is I found one typo in russian site version. Not that hard to fix.
DeleteWhere and what is the typo? You can just write about it here.
Deleteline
DeleteВистрели / Попадания / Процент попаданий
(Shots/Hits/Hitratio)
in replay details. The first word sould be вЫстрелЫ with "ы" letter
Thanks a lot, I have fixed it locally and it will be fixed on noobmeter.com whenever the next version deploys (usually a few days, but could be a week this time).
DeleteIs it true the rating isn't based on the average player, it only uses the players who upload their dossier cache file. How can this be considered a fair cross section of the player base on which to create another rating system. WG still don't include the amount of spotting damage in the information you use, so the suicide scout is going to get a better rating than the efficient and skillful one.
ReplyDeleteHow on earth can you use such a small cross section of the player base to develop the "best performance rating system" that could potentially affect everyone in the game. You really need to use the whole player base, otherwise you'll have players who are massively better than the real average tanker, labeled as a bad player.
WN8 excludes so much info because it isn't available, and is far too complicated as well as being based on an extremely narrow cross section of players to be a good performance rating.
If your interested to see what WN8 thinks your "expected" stats should be....
ReplyDeletehttps://www.dropbox.com/s/lfzdfv7xxi0t7nh/win8.xlsx
Copy your tank list table into the A1 (with header) A2 (without header) of the "UserStats" tab (special, unformatted). Then fill out the other blue cells (W10 - W17). It will populate the yellow cells with your expected stats.
If you did it correctly, the "UserCalc" tab's WN8 rating will match up with your noobmeter WN8 rating.
Or you can replace "/player/" with "/playerWN8/" in your NoobMeter user profile and get the coefficients printed out.
DeleteSweet! Thanks =)
DeleteMy number gets bigger therefore its a good system
ReplyDelete:P
oh and you guys really need a forum.
but the bar has been raised in every aspect. unicums are no longer 1700 but 2000. basically everyone stayed the same, the numbers are just a little different. while decent players scores went up, average players went down.
DeleteIts just a new formula making it harder for average players to get decent scores.
Big man trying to keep the little man down formula
I have 1800 battles total in WoT. 50% avg winrate, 1200 WN7. My WN8 score went up by around 100. I'm an average player. It's bad players who went down, not average.
DeleteYou all better focus on fire focus, covering allies and mini map than on all this ratings:)
ReplyDeleteWas wondering what might be the reason my Pz38Na is not among the server best by overall statistics (the third option with many games, dont know how to describe it better)? I play it from time to time, I'm an active player, the thing has 60+% winrate, and its still not on the table, while folks with 500ish fights and 58%ish wins are... any explanation? Its not that the glory is something of an issue, just would like to know why?
ReplyDeleteThe ingame nick is brynky, on EU server
DeleteI am a Korean user who really enjoy noobmeter. I have been transfered from another server, and my noobmeter ID stopped updating. My ID is 체습 and I wish noobmeter developer correct that manually. Thank you. I support noobmeter. :)
ReplyDeleteHi,
ReplyDeleteFirstly, good job, guys, your site is super useful for me. I got one question though: is there any way to manually set the time periods for my stats? I have checked NM regularly lately, and my graphs are very weird looking... and I'd prefer a graph based on monthly values, like refresh every 1st of the months. Is there any option for such custom periods?
Regards,
WinG_HU
Not really. You can select custom period for the " recent battles screen" but the charts are always drawn for all available time period.
DeleteSo I should check my stats every 30 days to get a nice graph? :D
DeleteMaybe you got me wrong. I'm glad with the fact that charts are drawn for the whole period. That's good. All I want to change is the update dates. Lately I refreshed my stats daily and before that, only like every 1-2 months. It would nice if the path of the graph would represent values put in every 30 days or something, not when I visit my site.
DeleteWell, registered users get updates every night. Then they get a max of 25 snapshots and old ones get thrown out. The logic for throwing them out is a bit complex. The option to simply keep one snapshot every 30 days or so isn't something that can be easily implemented.
DeleteThis comment has been removed by the author.
DeleteAh, okay, I got it now.
DeleteOk, one more question, the last one, I promise: how about implementing an option to erase certain snapshots? Or an option to hide certain ones?
DeleteMore feasible, but still I don't believe there are enough people interested in that feature for it to be worth the time.
DeleteWell, who knows! Maybe it will be the best feature ever implemented in the history of features!
DeleteThis comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
DeleteWhere can I report bugs ?
ReplyDeleteI have a screen shot showing the "Most Recent" missing kills.
As of right now, "Yesterday"s results for me include 29 games
with an average kill per game of 1.34. However, the "Most Recent"
results, which includes 14 games of those 29 games, show an
average of 0.00 kills. I counted 7 kills in the last four games
of those 14 games alone.
It is like the kills are not being counted in "Most Recent" possibly
because they were already counted in "Yesterday".
Thanks, Hautedaug NA server
Hi, If I am reading this right, why are WN8s average dmg for tanks so much lower than your PR average dmg for tanks?
ReplyDeleteBecause the other coefficients are different and that compensates for it.
DeleteI think WN8 uses avg player dmg while NM uses "good/great" player DMG as benchmark.
surely you should use average player dmg as benchmark and not good great? that is why average is average, unless you have a higher average of good/great players?
DeleteInitially NoobMeter was mostly used by good players, so there are some historical reasons for that.
DeleteBut anyway taking a 'good player dmg' then adjusting it downwards to get 'avg player dmg' yields pretty much same results.
Late reply:
DeleteIn PR, the expected damage values are more good/great at high tiers, but even 500 WN7 players could EASILY pass those requirements at those bottom tiers (1-3). It is quite unbalanced, to say the least.
Hi!
ReplyDeleteI'd like to report a bug, if you take my yesterdays stats you will find an average tier of 21,39 which is impossible!! (date is 16.10.2013)
http://img545.imageshack.us/img545/2655/vypg.jpg
If you could investigate this case :)
Thank you, stefko
ps:
Deleteuser: stefko on eu server
Inconsistent snapshot received from WarGaming: 6,081 battles in tank results, but 6,118 battles in summary results (difference 37 battles).
DeleteDifferent amount of battles missing from each snapshot. Ratings will be imprecise
Please add to statistics mastery badges if it's possible
ReplyDeleteThanks for the suggestion, I will consider doing this.
DeleteI don't think this is good idea. On most of my tanks i don't have mastery badges simply because i was driving them before they added mastery badges. And i'm not interested in spending so much credits to re buy them.
DeleteThis comment has been removed by the author.
ReplyDeleteDoes wn8 take into consideration that lower tier games are populated by more "less-skilled" players than higher tier battles?
ReplyDeleteIt is difficult or impossible to take this factor into consideration explicitly, but it is implicit in the expected performance values for each tank. Basically, yes.
DeleteIs there any chance that we could have rating ranks based on % number of players with similiar, better and worse rating? Something like Mastery badge for exp higher than 99% players?
ReplyDeleteFor example if player has rating higher than 99% players he would be supadupaunicum, player with rating higher than 46% and lower than 55% would be average?
I realize that it would require additional tables and cpu power to calculate it lets say every 60 mins but it would also show us if we are better from other players. (color for each rank would be awesome)
Here is proposal for ranking ranks with % for them.
< 5% - disaster (black #000000)
6% - 15% - hopeless (brown #660000)
16% - 25% - very bad (red #FF0000)
26% - 35% - bad (orange #FF6600)
36% - 45% - below averge (yellow #FFFF00)
46% - 55% - average (light green #00FF00)
56% - 65% - good (dark green #009900)
66% - 75% - very good (light blue #00CCFF)
76% - 85% - excellent (dark blue #0000CC)
86% - 93% - glorious (purple #660066)
94% - 98% - unicum (pink #FF00FF)
> 99% - supaunicum (rainbow?)
I really like the new WN8 rating... but I still think PR-rating is very good. And it has a drawback.. it doesnt take spotting damage into account.
ReplyDeleteAs far i know, only XP takes spotting damage into account. Not that average XP is a very accurate measurement.. but it still say something, nobody else does.
What about a joined rating system like this Rating=(PR+WN8+avXP)/3
XP is unfair to people without a non-premium account.
DeleteWhy isn't the rating only about the last n games played (or the games in the last n months played)? In a rating I am interested most in the current performance of the player not int he performance he did years ago.
ReplyDeleteThis would also solve the problem of the former OP tanks that gift their ernstwhile owners a nice rating boost.
Hi, is there any reason for Average kills field for a certain interval to have a negative value? Cheers
ReplyDeletealso, kill count shows 0 now, although i killed lots yesterday
DeletePlease provide link to your NM profile
Deletehere's the negative value for yesterday, and you will also notice 0 value for today's interval, thx for looking into it.
Deletehttp://goo.gl/lLsN7X
That's very strange; I can only assume that WG returns some wrong data to NM, maybe due to cluster differences. No other ideas.
Deletebut the other rating calculator (http://wot-news.com/stat/calc/en/ru) shows OK numbers, and even the Bronysite webpage shows a different rating than I have here... perhaps an error/malfunction on your side? cheers
DeleteI don't see how really. I checked database and I've received 12369 kills when you had 10471 battles and 12281 kills when you had 10509 battles. So battles went up, kills went down in data received from WG.
DeleteThat should not happen.
Considering their feed is providing incorrect data in so many ways (missing tanks, missing clan IDs, etc.) recently I think odds of the root cause being a WG feed problem are 99.9% and it being a NM problem are 0.1%.
I hear you, it's just that my rating on NM keeps going down with 0 kills) but shows fine on other sites, that's all. I heard WG was implementing a new statistics api (2.0), so it's either that or some cosmic forces) thx a lot nevertheless, i will keep checking the kills number.
DeleteStats refreshed and still 0 kills, although I averaged about 2.5 per battle...
DeleteSame for me here, for yesterday stats I have 0 killed and 0 spotted, though it is obviously wrong
DeleteI don't know what changed, but all the missing kills came back, wowza, all good for now, thx.
DeleteIt's gone bad again, Average kills: -356.00...
DeleteSame thing: http://www.noobmeter.com/recent/ru/MrDarK_AngeL/13546319/13546319_00006171_20131029_160243/13546319_00006181_20131030_031608
DeleteI'd like to see this for stats post 8.8 only which could then include the spotting/tracking damage. And if anyone could figure out how to omit platoon, TC and CW stats that would truly be awesome.
ReplyDeleteKeep up the good work though like what you have done so far.
Consider only damage is stupid. You can deal 10k damage in a game but u wont be sure to win it. Kill the 15 ennemis and you will be sure to win it. Manage to spot the 15 ennemis while your teammates kill them and you will be sure to win the game.
ReplyDeleteso what i dont understand i play 13 90 and gott lots dmg on spot butt i dont get wn7 or wn8 points gor that so iff you didnt putt dmg on spot on calkulate for wn8 you didnt do nothing
ReplyDeleteand 1 more thing its nott the same iff you play public game or tc or cw so in the game on random batle iff you hawe 1 day win rate 67 % next day you will get only team with 42% overal win so first 10 batle you dont hawe chance to win .... so its nott the same iff you play standard batle or tc or cw
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteВ расчетах где-то есть ошибки. Раньше я играл на 5-9 уровнях и все считалось нормально. Сейчас стал прокачивать новую ветку с нуля, играл на 2-3 уровне. И увидел очень странные цифры в графе "уничтожено врагов -2,46". Поправьте пожалуйста.
ReplyDeletehttp://www.clantalk.net/t/noobmeter-com-discussion-thread/21/2
DeleteЗабыл указать ссылки:
ReplyDeletehttp://www.noobmeter.com/recent/ru/Memex/2842408/2842408_00009972_20131026_050804/2842408_00010007_20131027_075707;jsessionid=F88368A564F7DFB901396E9C3DFA6D64
http://www.noobmeter.com/player/ru/Memex/2842408/
http://www.clantalk.net/t/noobmeter-com-discussion-thread/21/2
DeleteI' ve encounted some problem with yesterday update of my account data:
ReplyDeletehttp://www.noobmeter.com/player/ru/Valborn/10953942/
It shows huge negative numbers, but in reality its a mistake because on another online calculators all is good and numbers another.
http://www.clantalk.net/t/noobmeter-com-discussion-thread/21/2
Deletewhat about VBE (WN9 / WN9D)?
ReplyDeleteWhat about it?
DeleteWhat about the second german TD line? If I play them I'll get low WN8. That's because currently there's no data behind them? What does the calculator do if for exampla Expected XP is blank?
ReplyDeleteUpdated damages. Hoping that fixes that.
DeleteWell this is interesting:
ReplyDeletehttp://www.noobmeter.com/recent/eu/Ludd/500103569/500103569_00005288_20131030_192523/500103569_00005301_20131031_163450
Performance: 2,239
Efficiency: 2,081
WN7: 2,431
WN8: 514
What happened here?
I have the same question. This seems a bit off. I'm not saying i'm a good player, but i've just gotten my ace tanker in my KV1 in this period, and the other matches wheren't good, but not bad either:
Deletehttp://www.noobmeter.com/recent/eu/Yominokuni/512974265/512974265_00001613_20131030_221536/512974265_00001619_20131031_204150
Performance Rating: 1,816
Efficiency Rating: 1,353
WN7 Rating: 1,881
WN8 Rating (near-final): 57
Can it be that a low amount of matches drastically lowers the rating for the interval or is there a bug in the calculation?
http://www.clantalk.net/t/noobmeter-com-discussion-thread/21/3
DeleteAlright thanks. That explains the crazy WN8 ratings I've been seeing lately on both ends of the spectrum.
Deletesame here:
ReplyDeletehttp://www.noobmeter.com/recent/eu/Wice_Brullis/512348727/512348727_00003785_20131031_115931/512348727_00003798_20131031_201557
Gefechte insgesamt: 13
WN7 Rating: 1.367 1.062,2 → 1.063,7 (+1,5)
WN8 Rating: 53 1.120,2 → 1.086,8 (-33,4)
i think 53 is a bit low^^ but atleast im not the only one
http://www.clantalk.net/t/noobmeter-com-discussion-thread/21/3
DeleteThis comment has been removed by the author.
DeleteToday:
ReplyDeletePerformance Rating: 1,722 (+2.4)
Efficiency Rating: 1,433 (+2.5)
WN7 Rating: 1,505 (+2.9)
WN8 Rating (near-final): 309 (-7.0) !!!!
Is this good?
I also think i broke it :)
DeletePerformance Rating: 1,739 ;(7days) 1544
Efficiency Rating: 1015 ;1167
WN7 Rating: 1582 ;1480
WN8 Rating (near-final): 0 ; 22710
31 games ; 199 games
abused the trick or treat mission with all my tanks and all tier 1 that i could buy and sell and then i played some some normal games.
should make it so only 60 day shows up on xvm i like to see the stats of people who have played recent not overall
ReplyDeleteagree its beter that you see how som1 playt at last tome 6 monuts maibe
DeleteFor Oct. 31 there was some kind of reporting error for the WN8 rating. It gave me a 0 for the day despite doing damage, 60% WR, almost 1K average damage (VK. 3002), kills, spots, etc etc. PR and WN7 for example were over 1500. Battles for that day are still being rated a 0 as today it is still dragging down my 7 day for WN8.
ReplyDeleteI'm sick of this WN8 rating, when I play on my T34 or rest of my 6-8 tiers it's high but if I play on freaking 5 (Churchill III, T-25 or to make new 6) it falls to 300-700. And playing on ELC is painfull, 320 dmg, 1frag, 1900 ally after spot, 13 spotted (scout gained) and 320 WN8 (checked after 1 battle in day, and on most recent was only this battle)
ReplyDeleteThat is either the bug currently on Noobmeter or you're just not trying at low tiers (probably).
Deletehas there been an calculating issue with wn8 few days ago ? look here: is this true ? good wn7 but wn8 fail ?
ReplyDeletehttp://www.noobmeter.com/recent/eu/comanzatara/500761083/500761083_00011371_20131031_042420/500761083_00011381_20131031_232745
When will this site refresh? It doesnt seem to refresh all the stats. it says i have people in my clan that has not been there for long times. it does not show me in the top 50 td players for jpze100 with a 67 pct but has people listed with a 67 and less. i like the site but is there any way to keep it up to date and refresh a couple times a day?
ReplyDeleteа па рускей есть?
ReplyDeleteHow to disable refreshing everyday for registered players? I am playing less than before and refreshing twice a day (or everyday) is messy (1-2 battles in preview). I would change it to once per 3-4 days, but i do not know how.
ReplyDeleteim not a fan of wn8... i have gotten unicum in every other stat and wn8 gave me a zero? playing an average of tier 6? it seems broken.
ReplyDeleteAs with any kind of service, I'm sure that you get more complaints than compliments. With that in mind, I just want to make sure that I give a compliment about noobmeter. I'm sure that it takes a lot of time and energy to keep this site running as well as it does, and I want to give a big thanks to you. The way things are organized and presented here have done more than anything else to help me evaluate and improve my play, and thus have more fun playing, because losing sucks. So, thanks a million! I really appreciate it.
ReplyDeleteThanks for your kind words.
DeleteHey,
ReplyDeleteLooks like the beta tank is back to haunt me. Have a look here:
http://www.noobmeter.com/recent/eu/Ludd/500103569/500103569_00005361_20131106_001714/500103569_00005380_20131106_153521
Not sure whether it actually affected anything though? Number of battles don't seem to include it for example. Average tier does though. Thoughts?
Hello
ReplyDeletemy post is offtopic but i cant find link where i cant mail to noobmeter.
hi, i'm G_Maxarashvili from EU and i have Cromwell with 79.49% win in 39 game and cant find in server top. plz let me know why?
Hi! Can I have your WN8 tank data in csv format?
ReplyDeleteHi guys,
ReplyDeleteDoes anyone know how are 7vs7 matches going to be treated by wn8 rating? Are they going to be ignored, or somehow scaled? If those matches would've been considered as regular ones, it would affect wn7/wn8 negatively right?
new user here... love the performance rating.. originally I was below 999 (very bad), now I am 1148 (below avg). Most recent is 1516. what's the grading scale? ie: what is 1516 (still below avg or avg)?
ReplyDeleteThanks,
Shirva
http://www.noobmeter.com/calculatorDescriptions
Deletethanks dojchek
DeleteHi
ReplyDeleteI have something problem.
My noobmeter data has error
� (+�)<---like this
my rating was go down rapidly.
can you check my account?
http://www.noobmeter.com/player/kr/%EC%95%BC%EC%83%9D%EC%9D%98%EC%9A%B4%EC%A7%80%EB%B2%84%EC%84%AF/3000061844/
I found interesting one, and do not understand how it could be... The number of played games for some tanks is negative, so all stats got up abnormaly. http://www.noobmeter.com/recent/eu/Pestisoru_de_aur/509268446/509268446_00007425_20130710_225612/509268446_00007487_20131118_043407
Delete'Sea Server' from 'KR Server' has moved to the account.
ReplyDeleteBut Noobmeter is not updates.
I want to fix this problem.
Help me plz.
My user name is '푸르'
'KR Server'
This is a great site and all, i want to upload some epic replays but i cant because its impossible to login. I get "Login error" nomatter what option i choose. Ew_ on EU server.
ReplyDeleteClan in NA LIMBO isn't showing in the Top 50 for 30 day WN7 due to a bug in one of the players stats.
ReplyDeletehttp://www.noobmeter.com/clan/na/LIMBO
Player here (with avg. tier of -822.42 for most recent games)
http://www.noobmeter.com/player/na/Meijin/1000167922/
My stats wasn't refresh, maybe there is some problem.
ReplyDeletehttp://www.noobmeter.com/player/eu/Sedziwoj/500095364
My stats are not refreshed since 15 Nov.
ReplyDeletemine too mate. Anyone know why??
DeleteMy stats are ALSO NOT refreshed since 15 Nov,?????????
ReplyDeleteNo stats even on planetwot since 15Nov.
ReplyDeleteIs there some problem with WG servers? My stats aren't refreshing for like 3 days or so.
ReplyDeleteAs I can see, I'm not the only one here...
My stats havent updated from 23. nov. I am a registered account. Page: http://www.noobmeter.com/player/eu/ESTBoyx/500884263
ReplyDeleteMy stats haven't refreshed for a few days either.
ReplyDeleteDear NOOBMETER manager.
ReplyDeleteI was SEA user but now I'm in KR server.
Unfortunately, NOOBMETER Rating and tank datas are not updated properly now. (no updates and it's old data.)
My KR server nickname is 추요찡
I believe you can fix this soon.
Thank you.
This is now fixed
DeleteSorry for double post and bad English. I found interesting one, and do not understand how it could be... The number of played games for some tanks is negative, so all stats got up abnormaly. http://www.noobmeter.com/recent/eu/Pestisoru_de_aur/509268446/509268446_00007425_20130710_225612/509268446_00007487_20131118_043407
ReplyDeleteI think it must be checked before each calculation for negetive numbers in data.
why not include dmg upon asist ?
ReplyDeleteHello,
ReplyDeleteIt seems that KV-220 fights are not taken into account when displaying the statistics.
can someone tell me why are u not including assist on dmg in wn 8,from my point of view it is very important parameter. pls answer
ReplyDeleteYou can't get this value from WOT API, so that's why you can't include it in the formula
Deleteok thanks
Deleteso it means all those rating system are fake. Just snipe from red line behind everybody, use ur mates as spotter and as shield, deal damage, then show up full HP to finish kills. Stupid.
DeleteIf you can do that and win the game for the team, then it is ok.
Deleteits ok ? We dont need to have rating for that, we have the win rate. Mesure who will be the more red line sucker camper is just time wasted.
DeleteI have an after reset please update statistics clan history etc nickname neros2402
ReplyDeleteThis should be fixed now
Deleteczy ktoś mi pomoże?
ReplyDeleteThis comment has been removed by the author.
DeleteDear NOOBMETER manager.
ReplyDeleteI was NA user but now I'm in KR server.
Unfortunately, NOOBMETER Rating and tank datas are not updated properly now. (no updates and it's old data.)
My KR server nickname is 곰플.
I believe you can fix this soon.
Thank you.
NA Server : http://www.noobmeter.com/player/na/Gomple/1004959144/
KR Server : http://www.noobmeter.com/player/kr/%EA%B3%B0%ED%94%8C/3000355775/
Fixed
DeleteThe tank list is at best dubious at its numbers, i tryed look up some tanks i have decent knowlegde about theyr abilities with a decent driver and som are just ranked to low or high in damage. The FV207 british arty, to do average 1750 damage is on a ok battle par level, and is a good+ battle. the foch 155 is set to 2800, if ur that low u are doing something wrong or having a bad day, an average battle is around 3K + 3x800 is 2400 so u need only to hit with one mag with ok dam. rolls to get it. Just seems like some of the numbers doesnt reflect the real average but a desired average set by the guys making the list, wich might be bad for a ranking list.... And im a below average damage dealer, more of a tacktical driver , but still trying to learn to deal some whopass damage. Beside that i like the idea behind the wn8, but the far better ranking is done by http://www.vbaddict.net/ , there u will get a picture about a players performance in all aspects also when he ends up in a shitty midday MM and only gets a win when he does 80% of hes teams damage and kills. Have more tanks then the listed iv checked and they have similar faults.
ReplyDeleteWell, for the first year I just played the game for enjoyment, to have fun, relax, get rid of stress. The second year I started watching my stats alittle. Why I don't know, but whatever. This is still just a darn game on the internet. As long as I have fun and enjoy the game is important, I don't need a Unicom status or whatever, I still kill tanks every battle. have a lot of exp, etc etc.
ReplyDeleteThanks for a great tool.
ReplyDeleteIs there a reason why my Bat Chat arty is not shown in the top tiers vehicle list?
http://www.noobmeter.com/player/eu/RawhideNOR/500410451
Thanks again!
Thanks for the report, this will be fixed in next version.
DeleteIs there a sig made for WN8 yet?
ReplyDeleteIf Praetor informs me he has officially launched WN8 I can switch signatures to use WN8.
DeleteHi,
ReplyDeleteI would like to download data from the server(login.p1.worldoftanks.eu) using FileZilla or another program to perform several experiments. I have a request and question in one. Do you betray me how to do it? Thank you in advance :)
I suggest you contact WG support.
DeleteHello Noobmeter admi,
ReplyDeleteMy stats are not showing KV-220 battles:
http://www.noobmeter.com/player/eu/d2sp/500500251/
Please take a look at it..
http://worldoftanks.eu/community/accounts/500500251-d2sp/ Not present here either
Delete