30 replies to this topic

[Lboned]

Categories are fine, but I would weigh each one differently.

Percentage of how efficiency is weighed:
Damage: 40%
Damage done determines the large majority of WoT battles. The team that does more damage usually wins.

Kills: 10%
Overrated IMO. There are times when finishing shots are critical, but weighing it too heavily leads to stat nut-huggers shooting the 1% T20 instead of the full health T110. If you do damage, you will get kills. It is that simple.


Spotting: 20%
Very important. However, WoT's system is a little flawed in how they rewards scouting. Any more than 20% will lead to an increase in useless suicide scouting.

Cap: 15%
A niche stat. In tier 7+ battles, I would say 10-15% of battles end in a cap victory where the winning team WOULD NOT have won without capping.

Defense: 15%
Similar to cap, as in it fits a role on the team.


Overall, I like the mod. It is good to know who the better players are on the other team so you can give them preferential treatment.

[Skygunner]

Tweeked?

So the abstract number that have little bearing, still has little bearing?  Sorry, but there just is no magic number that can be made that tell you how "good" a player is.  Have fun chasing the higgs boson

Lboned, on Jun 09 2012 - 11:40, said:

Overall, I like the mod. It is good to know who the better players are on the other team so you can give them preferential treatment.

If you pay attention to the giants you will be eaten by the ants.   Literally the best efficiency calculator is your brain.  I'm not going to try to change your playstyle, because underestimating your opponents is a wonderful when...well you're the opponent.  In that regards, I'm not going to exert energy to make my opponents better players....same reason I stopped telling people how to effectively play the T-28 medium tank.   So I'll just leave it with this quote I often tell people in my platoon.


"A derp from a KV-2 with 1 hp left does just as much damage as one from a KV-2 with full health"

Interpret it how you wish.

[guns011]

Skygunner, on Jun 09 2012 - 11:52, said:

Tweeked?

So the abstract number that have little bearing, still has little bearing?  Sorry, but there just is no magic number that can be made that tell you how "good" a player is.  Have fun chasing the higgs boson



If you pay attention to the giants you will be eaten by the ants.   Literally the best efficiency calculator is your brain.  I'm not going to try to change your playstyle, because underestimating your opponents is a wonderful when...well you're the opponent.  In that regards, I'm not going to exert energy to make my opponents better players....same reason I stopped telling people how to effectively play the T-28 medium tank.   So I'll just leave it with this quote I often tell people in my platoon.


"A derp from a KV-2 with 1 hp left does just as much damage as one from a KV-2 with full health"

Interpret it how you wish.

well put sir o7 i agree with you 100% the "intangibles" are what make great players. attempting to quantify a player is simply ignorant imho.

[gilrad]

I still fail to see what exactly makes efficiency rating any better indicator of skill than win loss ratio over a large sample size (say 4k or 5k or more games).

Pretty much the only real improvement on win/loss ratio could be found by filtering out certain tanks/tank tiers, not because certain ones don't matter, but because certain tank tiers matter less depending on what you want to find. A clan looking for a UC recruit would be less interested in somebody's tier 1-2 tank win ratios than their tier 8-10 (mbts) or tier 6-8 (arty).

But I digress. Could somebody explain to me what efficiency rating tells you that win rate doesn't?

[Liebestod]

Sorry but kills are important. People who wait for the last shot are rare and they tend to have worse other stats.

The problem with the current efficiency rating calculator is that cap and defense are weighted too highly. They're really not very important and they are highly dependent on the type of tank you use (fast tanks will get WAY more cap and defense points since they can make it to the cap circle and they can make it back to base to defend).

Lower cap and defense and increase the importance of kills per battle.

[Liebestod]

gilrad, on Jun 09 2012 - 12:06, said:

But I digress. Could somebody explain to me what efficiency rating tells you that win rate doesn't?

Well when someone has say only 1550 efficiency but a 67%+ overall win ratio you have to conclude that they're being carried by elite platoon mates a bit.

And there are people with say only 54% win rate but 1700+ efficiency and obviously you know they're better than an average 54% tanker.

I take both of them into account to get a picture of what kind of player I'm up against.

Edited by Liebestod, Jun 09 2012 - 12:55.

[Livewyr7]

I agree the factors for calculation need tweaked (I agree with the OP's suggestion at first glance, since it is better than the current one by far) and I do think stats make a difference..  for example..  seeing an IS-4 with 640 eff rating pretty much means.. don't go with him, you'll find yourself alone.. and while that isn't ALWAYS the case, it is more often than not. Win Rate can be padded so I don't pay too much attention to that.. Eff rating not so much:  ergo why this needs a better calculation method.  (I rarely go for the cap unless I need to. I make more money and experience from damaging 4-5 tanks than I do from sharing a cap with a couple people, so my eff rating is lower than it should be)

Non-punctuated ramble complete.

[PTwr]

Efficiency Calculator, oh yea.
It is parody.

For example, for my account it gives score of 1251.57. Now, lets halve my damage while doubling my spots. And what we get?  Higher number: 1409.17.

Suicide scouts for the high efficiency, rush!

[gilrad]

Liebestod, on Jun 09 2012 - 12:36, said:

Well when someone has say only 1550 efficiency but a 67%+ overall win ratio you have to conclude that they're being carried by elite platoon mates a bit.

And there are people with say only 54% win rate but 1700+ efficiency and obviously you know they're better than an average 54% tanker.

I take both of them into account to get a picture of what kind of player I'm up against.

So pretty much, it's good for identifying people carried by skilled players (all those unicums must be pretty busy platooning with people who would normally be at 48%!), and deciding how far between 54% and 55% a person is.

The former doesn't seem like all that common an occurrence (at least at a level that would genuinely muddy up their win rate), and the latter doesn't seem all that helpful unless you're comparing two 54% players side by side.

[perlfreak]

A good player is generally the player who can stay engaged with the enemy the longest.

How well a person is able to do that is reflected to some degree in the stats.

The camper may live a long time in the battles, but they aren't being very effective with their time, and the stats will reflect that.

The suicider type lives a short time and doesn't contribute much, and the stats will reflect that.

The longer you live, the more shots you will take.    The more shots you take, the more likely you are to do do damage.   The more damage you do, the more kills and wins you are likely to get.   If you don't do well, then you try to engage, you get killed quick, and you do less damage.

Damage done is THE most important stat in game. Because of what it takes to do it.    How much damage is done and what kind of value the amount of damage done has is dependent on the individual tank.    The difference between the guy with good aim that hits copulas and the guy that bounces off the front turret?   Damage done.

Scouting is good, but that really needs to be it's own separate rating IMO.

Capping and Defending points always get over-valued because they are based on 0-100% each game. This causes them to go crazy high in terms of individual values, it goes unnoticed that it's got so much value because of the averages and games where you don't cap. But 1 single battle where you single cap and even if your team can win and suddenly you get multiple games worth of effeciency points.

In my first game with the M4 Jumbo I got 3 kills and I 100% capped my based(was not needed for win, but it's extra exp). For that single battle, my effeciency rating was over 16,000.   It was a good game, but 16,000?   Not even close.    I've had way better games and gotten way less eff from them.

I had a total of 1 battle in the T2 MT.   The American Tier 2 tank.     I don't even remember the battle.    But it has an eff over 12,000.    

These then carry out with a bunch of 0's from not capping, and so it gets smaller and gets a free pass.   But it's horrible that 1 game is affecting the overall so much.    And that's really what it is.     Oh you just so happen to be the first one in, and you get more % points etc, more efficiency.

It should be able to give a good efficiency rating for each individual battle, and they should be balanced in how that happens.   If you gotta include cap points, it should be somewhat equal with a good battle with kills/damage etc.

[Zebra21]

^{Im glad there is so much attetion payed to stats. As someone that has played 90%+ in pub matches and not in a platoon or tank company, Im glad you arnt targeting me. It gives me time to sneak up and shoot you.}

[Lboned]

Whether you agree with stats or not, making the mod more accurate is never a bad thing. Also, lol at the people who think kills are more important than overall damage.

[Cybergod]

How dare you make statistics that add up to 100%

[Ghoest]

The OP is definitely wrong about defense and IMO wrong about kills.

Defense is more important than capping in game. Over half the time you cap its more or less optional.
Defense on the other hand in most cases is being done as an imperative.

As for kills. Kills matter a great deal. Making sure people are dead helps you win. The killer types matter more than people realize because wounded who get away to be come snipers or capper can easily cost you a game.

[Meat_Puppet]

Without Damage, a kill is impossible.

Doing the damage puts you at far more risk than "reaping" the kill after the damage had been done.
The game rewards you for doing the work and you get very little for finishing off the opponent.

The game rightfully rewards the player who has put themselves at high risk to get the enemy down to the point where anyone can finish them off.

Finishing them off *IS* important. But first you must *WORK* to get them to that point.

Coupled with the player panel only reflecting kills and the pubbie culture (not surprisingly) looking at that as a means to "rate" their current teammates.
Until a damage total is incorporated into the player panel, this arguement will contninue to rage on.

My advice is to be situationally aware, let your lower tier tank mate or tank mate with a lesser gun finish off the kill and you concentrate on the next bigger or more threatening target. *IF POSSIBLE* and the situation will allow it.

The argument is not black and white but it is clearly the damage that is most important for the duration of the battle.
IE, before lighting the fire, you must first gather the firewood.

Everything after that is secondary and the reward system recognizes this as it should.

A tweek to the eff calculator to bring it closer to something that can be read *as an indicator!* would always be welcome.
The problem with statistics is that many tend to read them as absolutes rather than an indication for further review to form an *opinion* on probability.


Just my 2 cents...

[oldkye]

Liebestod, on Jun 09 2012 - 12:36, said:

Well when someone has say only 1550 efficiency but a 67%+ overall win ratio you have to conclude that they're being carried by elite platoon mates a bit.

And there are people with say only 54% win rate but 1700+ efficiency and obviously you know they're better than an average 54% tanker.

I take both of them into account to get a picture of what kind of player I'm up against.

was just reading this cause I've never seen this calculator before(I've played since beta yeah I'm a newb i know hehe) but saw it on a clan recruitment post(never cared to join a clan cause i just randomly play this game some weeks and some weeks i don't) "whats that?" so i looked up the calculator well reading this and I'm 54% win rate and it gave me a 1805.99 lol.

Edited by oldkye, Jun 11 2012 - 02:41.

[xankludan]

I love stats when playing arty. For example, looking at number of games, if they are 1k or less they often won't be looking for tracers or moving after shots.
For eff, if they have over 2k games played and a very low eff I can also be pretty sure they won't be cbing or maybe not so good, so I only move a little bit.
However, if they have a mid to good eff with a decent number of games, I know they'll be good at spotting tracers and trees falling and predicting where arty will move.

I find XVM is right on the money about 80% of the time. The other 20% are outliers, good players having a bad game, bad players having a good game, or players using tanks outside their 'normal' style of play.

[2_minutes_hate]

does anyone have the actual ratio numbers for effeciency. it seems like  i have been getting alot more damage in. i have been playing my higer tier tanks( i normally only play tier 10s rarely) and i finally upgraded guns on ferdi and vk4502. no matter how much damage i do my rating doesnt change. so itried to be in on the cap more and that hasnt changed my rating either. i dont get it.

[2_minutes_hate]

perlfreak, on Jun 09 2012 - 14:48, said:

A good player is generally the player who can stay engaged with the enemy the longest.

How well a person is able to do that is reflected to some degree in the stats.

The camper may live a long time in the battles, but they aren't being very effective with their time, and the stats will reflect that.

The suicider type lives a short time and doesn't contribute much, and the stats will reflect that.

The longer you live, the more shots you will take. The more shots you take, the more likely you are to do do damage.   The more damage you do, the more kills and wins you are likely to get.   If you don't do well, then you try to engage, you get killed quick, and you do less damage.

Damage done is THE most important stat in game. Because of what it takes to do it. How much damage is done and what kind of value the amount of damage done has is dependent on the individual tank. The difference between the guy with good aim that hits copulas and the guy that bounces off the front turret?   Damage done.

Scouting is good, but that really needs to be it's own separate rating IMO.

Capping and Defending points always get over-valued because they are based on 0-100% each game. This causes them to go crazy high in terms of individual values, it goes unnoticed that it's got so much value because of the averages and games where you don't cap. But 1 single battle where you single cap and even if your team can win and suddenly you get multiple games worth of effeciency points.

In my first game with the M4 Jumbo I got 3 kills and I 100% capped my based(was not needed for win, but it's extra exp). For that single battle, my effeciency rating was over 16,000.   It was a good game, but 16,000?   Not even close. I've had way better games and gotten way less eff from them.

I had a total of 1 battle in the T2 MT.   The American Tier 2 tank. I don't even remember the battle. But it has an eff over 12,000.

These then carry out with a bunch of 0's from not capping, and so it gets smaller and gets a free pass.   But it's horrible that 1 game is affecting the overall so much. And that's really what it is. Oh you just so happen to be the first one in, and you get more % points etc, more efficiency.

It should be able to give a good efficiency rating for each individual battle, and they should be balanced in how that happens.   If you gotta include cap points, it should be somewhat equal with a good battle with kills/damage etc.

Dang how do check the eff rating of a single game. I could pinpoint the importence of certain items that improve eff. I know if do mad damage in my arty but I get no kills it seems like I did well. If i get a bunch low damage it seems like a waste of ammo credits. I would rather get hella damage and confederate then low damage top gun.

[boxtosser]

The point of the game is to win, whether you win by racking up damage, kills, cap, or defense point, who gives a crap exactly how you do it, as long as you win the game?

Doing 4,000 damage is great, but if the enemy caps your base while you're inflicting it on the opposite side of the map, you still lose.  Also, cap/def is an area where lower tier tanks can contribute in a high tier match since a Maus and IS-7 cap at the same rate, and even doing 1% damage with a 75mm HE shell to a tank capping will reset the timer from 99 to 0.

Artillery is a pure damage dealer and sucks for spotting, capping, or defense, so playing arty a lot is going to screw up your rating.