Upset and/or confused that you didn't get an S?

Ice Weasel X··84 votes·15,059 views

Since the Mastery and Grading system has finally been added to other permanent game modes, the confusion over how it works is on an uptick again.

Grades are not based on static prerequisites. There is no magic KDA, CS, or warding score that defines an S grade. Your end of game grade is based on a dynamic comparison to others in your region who played that same champion in that same role. Having the same end of game stats on Top Thresh and Support Thresh will give different grades. Likewise, having the same end of game stats on ADC Jhin and ADC Jinx may also give different grades.

Champion Mastery FAQ What goes into the grading system? How do you calculate it? It’s based on the champion and role that you just played in a game. So, for example, we look at Annie mid differently to Annie support. We take a series of performance metrics and compare how you do to all other players in your region in your champion / position combination. It is percentile driven - so earning a high score means you performed in the top x% of players on that champ in that position. The grade itself governs the point gains.

As an example of a dynamic percentile-based scale, consider a bell curve. I assume most of you are somewhat familiar with those from school. One of my cousin's law school professors used an extreme form of this to encourage competitive thinking. If everyone in the class scored a 98% on an assignment, everyone was given a C. It didn't matter if you performed your task very well if everyone else with the same assignment also did very well. You had to outperform them.


Let's take a look at a League example. Say we have 10 players from different games. For the sake of this discussion, we'll be the first Janna.

|Role/Champion-|----KDA----|CS-----|Wards---|Grade| |- |Support Janna|0/0/19|12|35|A |Support Janna|0/2/23|15|37|A |Support Janna|3/1/17|13|36|A |Mid Janna|0/4/6|209|15|B |Support Leona|2/5/10|35|35|B+ |Support Janna|1/3/36|20|38|S- |ADC Ashe|12/4/1|317|14|A+ |Jungle Janna|0/4/12|257|16|A+ |Support Janna|2/6/11|2|27|B |Support Janna|5/1/27|30|41|S

######(Be aware that these are likely not the only metrics used for grading)

"WTF DIDN'T I GET AN S?" Those first three Jannas have stats similar to a lot of these complaints. "Look at those assists! Look at my low deaths!" While those are definitely good stats, Janna is a disengage support. Escaping death is what she does. A low death/high assist game is the norm for her, and performance like the last Janna in the list simply outshines the others. If high stats are the average, then you need to go above and beyond. As average regional results for a champion in a specific role improve, grade requirements increase. An oversimplified way of looking at it - excellent games raise the bar, while trash performance lowers it (more realistically, an outlier result will have a lesser effect on the overall distribution of the curve).

"Why is my grade lower than that Ashe?!" That's simple. Ashe's stats mean nothing to us. Different champion, different role. If the champion/role combo is different, it's irrelevant to your grade.

"Look how much better I did than Mid and Jungle Janna!" Again, doesn't matter. Different role. Jungle Janna might even score a higher grade than the Support Jannas in these games simply because of how rarely people play Janna in the jungle. There isn't an established history of strong performance in a large number of games with her in that role, so less impressive results may provide higher grades.


So, what about ARAM? There aren't well-delineated roles. Also, everyone is supposed to die! This is true. Summoner's Rift roles are initially determined by your location on the map. In the Howling Abyss, everyone's in the same lane. Caster supports are often either played full APC mode or hybridized, so the support grading criteria from the Rift wouldn't be ideal. Deaths are almost guaranteed in ARAM, but that doesn't mean grading is fucked as a result. Remember, grades aren't based on static values. Your results are compared to everyone else in the region playing that same champion in the same role. If the average death count for your champion on ARAM is 10, then having 8 deaths isn't going to be a dealbreaker. You're actually ahead of the pack in that regard.


Is there a way to guarantee an S grade? No, but there are ways to get an advantage. Two of the most obvious ones:

  • Play a champion/role combo so far off-meta that the average performance is weak. If you're strong with that champion, you can overcome role deficiencies and outperform most people using that specific champion/role combo. As with all off-meta gaming, this can be risky and you may catch flak; don't do it in ranked without trying it out in normals first.
  • Queue with friends who are at a far lower MMR. If you've earned your MMR, you've got a good chance of playing against opponents of a lower skill. Stomp hard. Think your stats are good? Stomp harder. Make sure your results are better than good. Bear in mind this is kind of a douchebaggy approach.

TL;DR The Grading system turns a single game into a competition on two separate fronts.

  • Your end-of-game stats are your performance in that game against the other team in the same game.
  • Your grade is your performance in that game against people playing your champion and role in other games.

72 Comments

Knight Sky20 votes

Good job with this, you put alot of time into it and it all makes sense! Clarifys a few things for me (: +1 from me

Tintalor8 votes

Wait, people other than me play Jhingle?

Velzard of Koz3 votes

Jhingle Bells Curve. Oh, my head.

SkyBlade798 votes

I can confirm that using off-meta things gets easy S's. I main AD Thresh Mid, and I get an S- or higher probably 75% of the time, even when I don't do amazing.

Lazypeon1002 votes

This isn't meant to flame or anything like that, just genuinely curious, what made you want to main ad thresh mid?

A Liberal5 votes

Probably the same thing that caused my AD bard top fetish (before those pesky Koreans did it and I started getting called an LCS wannabe) insert obligatory I did it before it was cool comment

The Blue Rajah2 votes

Ok, so what inspired that?

Troll for Trump4 votes

What inspires offmeta builds for me is when I get pissed and just do a troll pick, and it ends up working out well.

For me, that would be TF support, and Brand support(way before it was cool)

The Blue Rajah3 votes

That's cool. I could totally see tf support working. That gold card, cha ching.

A Liberal1 votes

I did TF support because I was like I want to play this sexy guy but I only play support

Some other nontraditional off-meta things I have done to success are:

Cassiopeia Support (Pre-rework) Blitzcrank Jungle Chogath Support Katarina Jungle (With Magus) Lux River Lane (Jungle but gank only) Soraka Top (I had literally a 100% win rate with tank raka top build after her newest rework for a patch or two) Urgot Support (The damage reduction and undogeable poke combined with AR/MR shred... MMM....)

A Liberal1 votes

Well you see.

One game I was playing Bard and my ADC DC'd.

Than I built AD on him and played ADC myself.

I did really well, but my other attempts to replicate Bard ADC didn't do very well.

So I brought him top, thinking the slow on auto attacks would be abuseable against melee.

I was right, oh was I right.

ForeverLaxx1 votes

I know this pain, here in League and other games as well like MTG. Such is the life of no-name innovators and people willing to try different things and research.

SkyBlade796 votes

Thanks for being civil ^_^

A few things, actually. If he can his his q- e - charged e passive autoattack, he can kill basically any squishy in one combo that takes about 2-3 seconds. And, he's really good in teamfights because his ult is still strong with his soul passive, and he can randomly snipe their adc and do more than half of their health with one auto. Even more, he provides amazing utility with every part of his kit (madlife hooks and ohana lantern!), and is less squishy than other AD carries because he gets tons of extra armor with his souls. Also, it feels good to have an off-meta main. I know that I'm only at bronze 3 at the moment, but AD thresh has won me SO many games, and got me from b5 to b2 (demotion :/). At first, it was just playing an off-meta champion for giggles, but now he's almost definitely my best champion in the game.

He only has three real downsides, too. For one, his early game is really weak, unless you can land a hook to drag the enemies into your tower. Two, limited mobility (the only real escape he has is using his Q to dash to a monster or minion). Three, 60% of my games have someone flaming me, which is really his biggest weakness.

Lazypeon1004 votes

I hear you on that third complaint man. I've had many a game where someone hated my pick and decided to flame instead of focusing, always makes things harder.

I used to play heimerdinger a lot before his rework (I should pick him up again honestly) and I'd get flamed even though I'd do well on him. I also take Vel'Koz as support more often than I do mid, because of his awesome poke and his reliable wombo combo for kills bot lane. I get people who flame me for picking him instead of a more traditional support.

The way I see it, in ranked, if you pick something off meta and you're trying your best to make it work, probably in about 95% of the cases I'm cool with it. You don't even have to do well, just so long as you try. That other 5% I'd say go to things like jungle raka (which incidentally could work before her update ages ago so long as you rushed a rylai's) who don't really intend on winning. If it's a normal, you can do whatever, I really don't care in those games xD.

I say keep working at it, if you're dedicated enough with him, I'm sure you can hit silver (or higher honestly) with him. I will say, 90% of this game is all mental, so I'm not sure how, vut you should try to find a way to calm those flamers down if at all possible before they become a problem. Typically by wrecking your foe >:). If they are just going to rage anway, just mute them. It honestly helps a lot. Good luck man!

SOVERElGN4 votes

Ice dropping common sense nukes. [zombie-brand-mindblown]

Zarvanis4 votes

Play a champion/role combo so far off-meta that the average performance is weak.

Hold on to your butts. ADC Singed Singed, coming at ya.

ManiacalCoyote1 votes

My body is ready. [slayer-jinx-wink]

DorkunedAuras2 votes

You still have to do well. If you feed your ass off, then you're still not getting a free S

Mast3rShake3 votes

one problem with the following:

Role/Champion- ----KDA---- CS----- Wards--- Grade Support Janna 0/0/19 12 35 A Support Janna 1/3/36 20 38 S-

What if the game time in example one is 22 minutes and in example two it is 60 minutes? In a 60 minutes game, you would expect to be able to accumulate more assists, kills, and wards.

Also, what if in the first example you team had a total of 25 kills and in the second example they had a total of 70? That means in game one you had a 75% kill participation rate and in game two a 53%.

The following shows two major factors that must be incorporated into your grade if they are not already.

AlmightyEyal1 votes

Its makes sense, its probably effects the grade.

the Anarchit3cht2 votes

I wonder how they determine what "role" you have. Is it based on what is chosen for you? And if thats the case, what if you just swap? LIke say you get mid and top asks to trade and you do.

DragonMastaWolfy1 votes

I'm assuming that it's determined by your build and where you were the most in the early game.

the Anarchit3cht2 votes

So then couldn't we theoretically "trick" the algorithm that decides that with lane swaps? Like say I'm mid and the top laner isnt doing too well in his match up so we swap later into the laning phase? What effect would that have?

Lord Galvatron2 votes

TL:DR

  • Farm like you have never farmed before: Farm has a disportioncate factor in your grade so more CS = better grade EVEN ON SUPPORTS (ZZ'Rot and Banner are your best friends) (also needs to be fixed it's a terrible way to measure skill on jg/supp)

  • Participate in everything dragons, towers, kills (you name it, you need your face nxt to it)

  • Low elo stomp w/ friends, shameful but if you want dat S you WANT dat S.

  • Play champs that are less played and roles on champs that are less played for MUCH easier S/S+. (rly important)

  • Longer games will yield a higher change of you getting a better grade (more things done that game then the avg. game played with the champ)

P.S: You can NOT earn lv 6 or lv 7 tokens on anything BUT normal/ranked Summoners Rift.

Zerenza2 votes

I frequently get S's as Illaoi Top does that mean i'm technically better than a vast majority of other illaoi player's in my region?

Communist Toast2 votes

I don't think your stats are compared the other team in the same game..I always thought it was just against people that play your champ in the same role.

also time length is a huge factor

aka, 0/0/19 is amazing at 20 min, not so amazing at 45 min

Carpe Victoria1 votes

I finally got an S on Ekko. An S- to be exact. It took me a game of going 24/10/15 (long crazy game) to get that damn S finally. This champ's mastery stats are gross.

AlmightyEyal1 votes

And now he got nerfed so its super hard to get a good grade with him

Communist Toast1 votes

Teemo Top--22/4/4 272 CS 33:37 match S+ (victory) Teemo Top--4/7/4 265 CS 29:42 match A+ (victory) Teemo Top---4/4/12 508 CS 49:47 match S- (defeat) Teemo Jg---16/7/11 67 CS 38:19 match S+ (victory) Teemo Top---3/13/7 328 CS 45:19 match A- (victory)

DorkunedAuras1 votes

508 CS in 49 minutes, holy shit nice.

SHlTBUG1 votes

This tactic doesn't work, Riot has confirmed that playing something off-meta just compares it to the average for that role. Sorry.

DorkunedAuras1 votes

it does compare them with the average for that champ in that role. But if half the people that play janna jungle are trolls, the stats that you're being compared to are really low and thus it's easier to get really high ranks.

SHlTBUG1 votes

no not the average for that champ in that role, the average for that role.

so like, not enough people play janna jungle to get solid statistics, so it just uses the average JUNGLER of all champions to get your score.

e.g. jungle janna is calculated using jungle in general, not jungle janna

Ashe Jungle1 votes

[{quoted}](name=Ice Weasel X,realm=NA,application-id=3ErqAdtq,discussion-id=hYTXYyGh,comment-id=,timestamp=2016-06-18T01:35:59.718+0000)

  • Play a champion/role combo so far off-meta that the average performance is weak. If you're strong with that champion, you can overcome role deficiencies and outperform most people using that specific champion/role combo. As with all off-meta gaming, this can be risky and you may catch flak; don't do it in ranked without trying it out in normals first.

Can confirm.

Most of the time I get S-. Good games are S or S+. The only time I've gotten less than an A was a game where the enemy nid was all up in my jungle and pretty much killed me on respawn.

Original Blob1 votes

This definitely helped me understand the grading system better. Thank you!

Raxicor1 votes

I bow down to that support Janna with 5 kills. Lol

DorkunedAuras1 votes

Play a champion/role combo so far off-meta that the average performance is weak. If you're strong with that champion, you can overcome role deficiencies and outperform most people using that specific champion/role combo. As with all off-meta gaming, this can be risky and you may catch flak; don't do it in ranked without trying it out in normals first.

I main Rek'Sai top and I regularly stomp. This is how I have 35+ S+ Ranks and 70 S ranks on her with "only" 255k mastery.

Mast3rShake1 votes

I don't think a 57% winrate on Rek'Sai qualifies as "regularly stomping."

DorkunedAuras1 votes

Let's clarify a bit. I regularly stomp LANE. I end up with a big score, but Rek'Sai's terrible scaling means I fall off if I don't wreck everyone's shit incredibly hard early-mid. That means I've lost far too many games with fantastic scores. 4 times I've managed an S+ rank in losing games.

PK Flunr1 votes

Explains why I regularly get A or higher as Xerath support. Nice guide! ...On second thought, it might be because of my habits of ksing with R

Foxyu1 votes

I got 19/2/12, 168 cs, 19.2k gold, I played AP Ezreal Mid lane and we won, plz tell me why i got A instead of S- ?????

LoliBear641 votes

well, time to play jungle ahri Ahri

LoliBear641 votes

i went 13/1/4 ahri jungle and got a+

teritoo1 votes

Come on, i got a 38/6/9 with 189 cs with vayne and basically carried my team and still got a A+. Explain this.Vayne

The Riss1 votes

I played a game finishing 22/0/11 as Orianna Orianna and only placed an A+???? help :(

Orianna

Scary Door1 votes

Played an ARAM with Lux. 51/6/31... didn't get an S+ only an S

Adisky 1 votes

12/0/22 as a orianna and got A... Riot?

Naramiel1 votes

I still think the biggest thing the system is missing is communicating what you needed for a better score.

More from Gameplay

More by Ice Weasel X