_(This is not a boo hoo i didn't get an S+ post, it is a serious question and discussion)_
We all get scores that we think "yup thats an S+" and then find out that it wasn't even an S- .. and while frustrating, thats okay, we know its hard, and we know its gotten harder since the mastery 6 and 7 came out and that is fair enough, getting 7 is a big achievement to be proud of.
But sometimes the discrepancies in the stats are so wild for the same grade, that it just makes you wonder how consistant the grading system actually is.
I have heard a lot of things (but never from a RIOT member) on how the stats are graded:
1. Its a fixed score that you need to reach
2. you are compared with everyone else who plays that champ
3. you are compared with everyone else in your skill level who plays that champ
4. you are compared with everyone who plays the same role as that champ
I wont pretend to know which one it is (if it is even one of those)
and I've heard lots of things about what are the most contributing factors:
1. Its all about good KDA
2. Winning is the primary contributor
3. Deaths mean more than kills
4. Assists don't really count on anyone but supports and tanks
5. Farm and objectives are the biggest factor
6. kill contribution is everything
7. it takes into account your mmr and the mmr of your opponents
8. total gold income is how it is calculated
Again, i can't pretend to know what the answer is.
However in saying that. Let us look at Time for a minute. Time must be a consideration somewhere in the grading system, because winning a game at 15 mins with 8/2/3 with 130 CS is a good score and likely will give your grade a bump in the higher direction, however the same score at 50 mins will likely not yield the same grade as you should be expected to have more than 130 CS at 50 mins and to have since contributed to more kills than you did at 15 minutes.
**Therefore we must assume: That as time increases, the in game statistics required to achieve the same grade must rise exponentially. **
This means that if a game were to end at 20 minutes and the player achieved a B+ (scenario 1), Then if the same game was extended to 50 minutes and the player also got B+ (scenario 2) it is the logical conclusion that the difference between the two sets of stats, is exactly proportional to the amount of time elapsed between them. Thus ensuring the same grade. If the player had achieved less stats than he had in the interval between 20 minutes and 50 minutes, his grade will fall, and if he achieved more stats, it will rise. This implies an exponential curve, that if you get ahead of, your grade rises, and if you fall behind, your grade decreases. We will call this the 'Curve of Natural Progression'.
_(how this curve is calculated, and what factors contribute to where you sit on the curve is still unknown, but we do know that if time is a factor in determining a player's grade, then this curve will exist.)_
Now let us refer to the example image, this shows two of my recent ranked games playing the same champion in the same position, both games yielded the same grade, yet had extremely different in game statistics. For ease of comparison lets put the stats side by side.
Creep Score: 184
Turrets Destroyed: 1 (out of 1)
Dragon assists: 0
Creep Score: 143
Turrets destroyed: 3 (out of 9) _But i assisted in most_
Dragon assists: unknown (out of 4) _but I'm certain i helped in most of them_
Every stat was better in Game 1 except for Deaths and Creep Score. Apart from this, we can be fairly confident that Game 2 had a better score. But because they achieved the same grade, we can draw 2 conclusions:
Conclusion 1: If your statistics are the same as Game 1's, then 10 minutes later it is normally expected that you should have the stats of Game 2.
Conclusion 2: 1 Death more, 41 CS less, and 70 potential CS less (assuming that the CS continued climbing at 7 per minute) are worth more negatively, than 16 more kills, 17 more assists, and 2 more turrets, are positively.
Now neither of these conclusions sound very healthy, i think that very few people could be expected to go from 4/5/5 at 24 mins to 20/6/22 at 34 mins, Which means that either: The Curve of Natural Progression does not exist (which therefore means that Time/Game-length is NOT a factor in determining the grade) OR that the way the system works out the grade is actually Inconsistent (which is unfair).
So. If The Curve of Natural Progression does not exist, then we can only assume that the Grade is decided on stats alone regardless of game length, which means that 1 death and 111 CS is worth exactly the same towards the grade as 16 kills, 17 assists, and 2 turrets (not including various dragon/turret assists also). This means that if Game 1 had continued to 34 minutes, and the score was 4/4/5 with 295 CS, it would have resulted in an A+
One last thing that we have not yet considered, is how the stats are judged, whether it compares you against a predetermined expected score, or wether you are compared against the average scores of all the players of that champion etc etc. I do not know which one it is, but what i do know is how far apart these two games were: 1 day. Because we know this, we don't actually have to know how the grades are determined, but we know that we achieved vastly different scores, and were awarded with the same grade. This adds a 3rd possibility to what causes the discrepancies in the scores/grades, and that is that whatever is being used to compare your scores to in order to determine your grade is changing... Daily.
This means the exact same score could be worth an S one day and an A the next, which defeats the purpose of 'mastering' a champion, if some people get easy grades because they happened to play their champion on a day where everyone else was doing badly, and some people struggle just to get the same grade the next day because that day just happened to be a good 'Aurelion Sol' day for people so the average went up. I understand that with balance changes to champions, the required scores needed to obtain grades will change, but surely it is fairer for the averages (if that is how the scores are graded) to be collected over a week or more, and not just a single day.
**So what we have discovered by this analysis, is that either:**
**1. The Grade calculation per game is inconsistent
2. Kills, Assists and Objectives are virtually worthless in Grade calculation.
3. The scores required to get grades, are changing daily. **
And to me, none of those results sounds very fair in terms of mastering a champion.
So i ask you RIOT, which one of those 3 options is the truth? And if you claim none to be true, then please provide the reason why such vastly different Aurelion Sol scores yield the same grade?
_(im not mad cos i didn't get an S or anything, i already have mastery 7 on Azir (and everything after that is just a bonus to me), and i only need two 7 tokens for Sol now anyway, so I get enough S's, but i am frustrated with the lack of consistency in grading, so the point of this post is to discover why such discrepancies occur)_