|
|
Space prison is eternal by definition. For toplane I'd also like to see something like K+A per teleport and for support something with wards.
|
Cayman Islands24199 Posts
well if an adc plays poke champs all the time he'll get up on the damage charts.
chaotic games that go late also tend to have higher dmg/time given the scaling of stats
|
Space really does most damage? I knew he was doing fantastic this season, but that's still surprising. Pray is 1 dmg/min behind lol, and I see him doing tons of damage every game, something I've always noticed about him.
Also a bit weird to add dmg done and taken for top laners, it means we can't discern the difference between someone who likes to play damage heavy carries and who plays tanks.
|
On July 02 2015 19:16 DarkCore wrote: Space really does most damage? I knew he was doing fantastic this season, but that's still surprising. Pray is 1 dmg/min behind lol, and I see him doing tons of damage every game, something I've always noticed about him.
Also a bit weird to add dmg done and taken for top laners, it means we can't discern the difference between someone who likes to play damage heavy carries and who plays tanks. That's intentional, since metas aren't always defined by one or the other.
|
On July 02 2015 18:37 Fildun wrote: Space prison is eternal by definition. For toplane I'd also like to see something like K+A per teleport and for support something with wards.
It does seem a bit too simplistic but it's an interesting concept. Would like to see it expanded on further.
|
When space is the highest scorer clearly there is something wrong.It is interesting but like all statistics are too dependant on champion/team picks.Don't think it is possible to make a single statistic like this and be even close to true.
|
GrandInquisitor
New York City13113 Posts
On July 02 2015 22:13 nafta wrote: When space is the highest scorer clearly there is something wrong.It is interesting but like all statistics are too dependant on champion/team picks.Don't think it is possible to make a single statistic like this and be even close to true.
The perfect is not the enemy of the good. Just because you can't come up with a single number that summarizes everything about what it means to "carry", doesn't mean we shouldn't do the best we can with the information we do have.
Here, the formula is simple and intuitively useful. Space has done more damage per minute than any other ADC in LCK. That's good.
|
Baa?21242 Posts
these stats have shortcomings as most do, but they're a lot better than the stats we get in LCS, which tend to not take into account game time/total kills at all.
|
My one frustration is that these stats aren't of a similar metric. You have some at high numbers, other percentages.... Imo, they should do what they do currently, set the highest at 100 and grade all other scores based upon the highest.
|
Baa?21242 Posts
Comparing across different roles is too difficult. Rather than trying to find a way to artificially standardize at 100, having numbers that make sense within a role is preferable imo.
|
On July 02 2015 22:44 GrandInquisitor wrote:Show nested quote +On July 02 2015 22:13 nafta wrote: When space is the highest scorer clearly there is something wrong.It is interesting but like all statistics are too dependant on champion/team picks.Don't think it is possible to make a single statistic like this and be even close to true.
The perfect is not the enemy of the good. Just because you can't come up with a single number that summarizes everything about what it means to "carry", doesn't mean we shouldn't do the best we can with the information we do have. Here, the formula is simple and intuitively useful. Space has done more damage per minute than any other ADC in LCK. That's good. which means nothing since you need context.What champs has he played/how many times was he in a laneswap/was his team ahead or behind more often and countless more.Basically it doesn't tell you anything about how he performed unless you watched the games.
The lcs statistics are about as useful as the support "carry rating".
|
Surely the score for the support position should factor in the average wards placed per game
|
I think Fury damage stats are pretty good for how often Samsung lose
|
On July 03 2015 00:34 jack342342 wrote: Surely the score for the support position should factor in the average wards placed per game
or like, sweeping differential would be cool, I'm not sure if that can be calculated w/o going back and watching replays though
like cheep said all stats have flaws and so do these but looking at the way it turned out, only Duke's absence and the supports seem really disconnected from reality(wolf for instance is bad but plays on a team with a lot of kills so he racks up assists).
|
Cayman Islands24199 Posts
it's definitely interesting data but given the small sample size and the pretty sizeable effect champ choice/game types have on these stats it's also easy to read too much into these numbers.
this is more useful at assessing playing style/priority within individual teams than cross team player comparisons
|
On July 03 2015 00:34 jack342342 wrote: Surely the score for the support position should factor in the average wards placed per game
I'm a bit more surprised at that the support position formula lacks a denominator of any kind.
|
On July 02 2015 22:44 GrandInquisitor wrote:Show nested quote +On July 02 2015 22:13 nafta wrote: When space is the highest scorer clearly there is something wrong.It is interesting but like all statistics are too dependant on champion/team picks.Don't think it is possible to make a single statistic like this and be even close to true.
The perfect is not the enemy of the good. Just because you can't come up with a single number that summarizes everything about what it means to "carry", doesn't mean we shouldn't do the best we can with the information we do have. Here, the formula is simple and intuitively useful. Space has done more damage per minute than any other ADC in LCK. That's good.
Problem is that these ratings are gigantically flawed. The simple KDA metric is actually a better indicator of performance than these ratings, You can basically just go ham all the time, die a lot and get rewarded for it, since there is no punishment for dying too often. All roles should definitely take into account deaths as well.
Next step (in an advanced rating) is to take the champion into account. E.g. a top lulu player shouldn't be punished for not tanking a lot of damage.
|
On July 05 2015 22:04 Hider wrote:Show nested quote +On July 02 2015 22:44 GrandInquisitor wrote:On July 02 2015 22:13 nafta wrote: When space is the highest scorer clearly there is something wrong.It is interesting but like all statistics are too dependant on champion/team picks.Don't think it is possible to make a single statistic like this and be even close to true.
The perfect is not the enemy of the good. Just because you can't come up with a single number that summarizes everything about what it means to "carry", doesn't mean we shouldn't do the best we can with the information we do have. Here, the formula is simple and intuitively useful. Space has done more damage per minute than any other ADC in LCK. That's good. Problem is that these ratings are gigantically flawed. The simple KDA metric is actually a better indicator of performance than these ratings, You can basically just go ham all the time, die a lot and get rewarded for it, since there is no punishment for dying too often. All roles should definitely take into account deaths as well. Next step (in an advanced rating) is to take the champion into account. E.g. a top lulu player shouldn't be punished for not tanking a lot of damage.
I would never say KDA is a good indicator of performance (then again, I also don't think any single number is a good indicator). There are a million reasons why you could have a high or low Kill count: split pushing, drawing attention, suiciding to kill/lockdown the carry, playing a supportive champion, playing an assassin with one goal. And then team strategies like trying/staying away from team fights, early aggression, safe laning, trying to stall blah blah blah.
There's too many factors in League to be able to deduce any sort of meaningful statement from one single number. You can say 'this guy had a great game, look at his KDA' but most of the time you'll be saying that because you watched the game, and know why he has that KDA. I've seen supports get MVP in the LCK where their KDA, or even number of assists alone like this stats page, was not particularly high, but they got the award for a few very vital plays.
|
I would never say KDA is a good indicator of performance (then again, I also don't think any single number is a good indicator). There are a million reasons why you could have a high or low Kill count: split pushing, drawing attention, suiciding to kill/lockdown the carry, playing a supportive champion, playing an assassin with one goal.
There are a millions of reasons yes, but it doesn't have a signficiant bias towards aggressive or passive teams/players. When you can get rewarded for feeding by suciding over and over (as you deal and take damage in the proces), you know the rating is as good as useless.
Someone could play a passive long game with few engagements, and you would have a shit rating even if you played every single engagement flawlessly.
So until this rating can take into account deaths, it's inferior to KDA (which also is flawed).
|
|
|
|