{{
`${userDetail.user.displayName}${userDetail.user.cachedBungieGlobalDisplayNameCode ? `#${standardizeBungieGlobalCode(userDetail.user.cachedBungieGlobalDisplayNameCode)}` : ""}`}}
{{getUserClans()}}
7/4/2020 6:20:32 AM
8
[quote]I did not actually GIT GUD, I was just benefiting from how the SBMM was putting me into matches. RIP the days when I was 'good' at Crucible. Turns out I wasn't and it was an SBMM illusion.[/quote]
Literally the same thing happens when Crucible goes from CBMM to SBMM for all the good players. They suddenly get lumped together and their KDs all drop and they need to 'sweat' to compete and their skill was all just a CBMM illusion.
It goes both ways.
English
-
Huh? That doesn’t makes sense. Yeah when higher skilled players play in SBMM their KD drops because SBMM essentially keeps players at 1.00 (if opponent are truly the same skill.) On top if that with the smaller player population in the higher tiers you wait forever in a match and then deal with quite a bit of lag. Meaning now your not only playing against highly skilled opponents you’re also fighting against random BS. That doesn’t mean they aren’t still skilled. Their is no illusion created by CBMM.
-
[quote]On top if that with the smaller player population in the higher tiers you wait forever in a match[/quote] Actually, the 'smaller player population' at the top of the curve would make up more of the game's traffic. It makes sense that good players put in more time and take the game more seriously than those at the bottom of the curve. The Pareto Principle would suggest that the top 20% of players would make up about 80% of the game's throughput. It would be likely that it's actually the lower end of players who experience greater wait times. [quote]Their is no illusion created by CBMM.[/quote] There definitely is. You said right at the start, if they get put up against other people of their own skill, their KD would drop, making them think they need to try harder in order to remain competitive, that leads to sweating. It's exactly the same as what OP describes, he was getting good against the players he was being paired against, but put in with the general population he now feels that he's unable to compete. CBMM and SBMM both cater to different skill levels.That's why nobody can ever agree on which one is definitively better for the game, and why Destiny goes through these cycles like this.
-
Kd, because of matchmaking, and numerous other factors, is not a good measure of skill. The [i]median[/i] kd of the player population is almost certainly below one. What he's alluding to is that with sbmm you're generally facing players of similar skill, so kd has some meaning. In cbmm you can often face much less skilled players that don't represent much of a challenge. It's not too tough to have a high batting average if half the pitchers you face are still in Little League. It's not like players on opposing teams sort themselves out by skill and then only go after the player closest to their own level. It's only natural to go after the weakest prey again and again. Does that result in kd being a meaningful indicator of skill? You tell me.
-
[quote]Kd, because of matchmaking, and numerous other factors, is not a good measure of skill. The [i]median[/i] kd of the player population is almost certainly below one. [/quote] Kd is a great measure of skill, but not the only one. In a team deathmatch, it determines who wins and who loses. Skill like IQ is almost always a normal distribution and the mean, median and mode are very close to 1. Suicides make it slightly below. [quote] What he's alluding to is that with sbmm you're generally facing players of similar skill, so kd has some meaning. [/quote] This doesn't make sense to me. Sbmm dynamically matches you with people around your skill level so your kd should always be pushed towards 1. Sure its meaningful to have a high kd in a sbmm game because it means that you played beyond your means and did well, but that particular kd is not indicative of your overall performance. Next game, you will most likely be thrown into a higher skilled lobby and be nailed down and your kd will be brought back in line. [quote] In cbmm you can often face much less skilled players that don't represent much of a challenge. [/quote] This depends on whether you are above or below average. If below average, you have a higher chance of encountering someone more skilled and vice versa. Whether they are much of a challenge also depends on how much better they are than you. Given that most data points in a standard distribution fall in the middle, there generally is not much of a difference. A 1kd is top 40% vs a 1.28 kd being a top 12% as an actual example. [quote] It's not too tough to have a high batting average if half the pitchers you face are still in Little League. [/quote] Correct, but kd is meaningful because you have an equally opportunity to face the whole population. If you consistently do well, this shows that you are above average. [quote] It's not like players on opposing teams sort themselves out by skill and then only go after the player closest to their own level. It's only natural to go after the weakest prey again and again. Does that result in kd being a meaningful indicator of skill? You tell me.[/quote] How do you track players on the opposing team and go after the weakest link. I dont know what game you are playing, but I have not done this once in my 4000 games in D2.
-
You're missing the point. A kd of one says nothing about your skill level. It only tracks how many kills you get versus how many times you are killed. That's a flawed metric for lots of reasons. It is only valid as a tracker of skill when every player considered plays the exact same player pool as every other. That doesn't happen. It comes closest with sbmm because the comparison becomes more meaningful when every player faces opponents of about the same skill level. That's why sophisticated matchmaking systems use much more complex formulas to determine relative skill. The truth is that if you're playing a lot of sbmm pvp and your kd is constantly increasing, then you're getting better. It's much harder to say that with cbmm. The only way to establish your [i]true overall skill level[/i] is to consistently play the best or worst in the game until your success begins to change. Then you move up or down depending. But that requires skill gradients that remain consistent until you find your own level. Once you've determined that level, changes in things like kd can be said to have some kind of meaning. But the skill level of opponents with cbmm is too inconsistent to allow for those kind of comparisons. And once you reach the top of the ladder in Destiny pvp you can start at the bottom rung in a [i]real[/i] competitive shooter.
-
[quote]The [i]median [/i]kd of the player population is almost certainly below one.[/quote] I was thinking about this before I read your comment and, logically, if for every kill there is a death, then theoretically the average of every k/d in every match ever should come out to 1.0. Then I had a good laugh when I remembered all the times I fell off the map and died but no one killed me, so the k/d has to be less that 1.0. (Not sure the math is entirely correct, but it's a funny thought. :) )
-
[quote]logically, if for every kill there is a death, then theoretically the average of every k/d in every match ever should come out to 1.0[/quote] This is incorrect. KD is a ratio and not a 'number', so averaging them like this doesn't work. Consider a 1v1, Player A gets 3 kills before dying once. Player B then has one kill to his 3 deaths. That gives Player A a KD of 3.00 and Player B a KD of 0.33. The average between the two KDs is 1.6665 despite there being four kills and four deaths. Also, remember that with KD, the first death is 'free'. Going 25 and 0 gives the same KD as going 25 and 1. This means that players that go unbroken, or just join late in the game, or leave early skew the results further. TL;DR - SBMM still wouldn't trend the median KD to 1 because KD isn't a 'number'.
-
Kd bottoms out at zero and tops out at, theoretically, unlimited. Consider how many low kds would have to exist to justify a single 3.0 if the [i]median[/i] was 1. You'd need way more people with low kds. Which means most people don't trade 1 for 1. That may be one reason why Bungie went to kda. It makes people feel better.
