Sakeri (NA)
: Zed's a girl?? Confirmed by Riot Blaustoise!
: > [{quoted}](name=Riot Blaustoise,realm=NA,application-id=yrc23zHg,discussion-id=Q0W4UFT6,comment-id=0004,timestamp=2018-01-28T20:11:17.512+0000) I dont know you, but I already know I love you, {{sticker:sg-lux}} I mean, ahem, we have so much in common. Riot Blaustoise really ~~Hydro~~ Pumped me with his words which I can truly relate with.. {{sticker:sg-jinx}} I'm all Wet now, but np Afterall I am well **aware** that there is a endless **supply** of Flames out **there** that I will use to warm myself and **dry.** ~~Yey, It Rhymes~~
> [{quoted}](name=LifelessAngel,realm=EUW,application-id=yrc23zHg,discussion-id=Q0W4UFT6,comment-id=00040000,timestamp=2018-01-29T01:37:49.023+0000) > > I dont know you, > but I already know I love you, > {{sticker:sg-lux}} > I mean, ahem, we have so much in common. > Riot Blaustoise really ~~Hydro~~ Pumped me with his words which I can truly relate with.. > {{sticker:sg-jinx}} > I'm all Wet now, but np > > Afterall I am well **aware** > that there is a endless **supply** > of Flames out **there** > that I will use to warm myself and **dry.** > > ~~Yey, It Rhymes~~ ... thank you? O__O
: Riot Employees
Hey a couple of things here: (1) holy yes I get camped out of my mind on my Riot account (lowkey I appreciate it, I think it makes me play smarter). (2) I'm not always that "unemotional". I love league, I'm very emotionally invested in it, and I certainly get... passionate at times. I would be lying if I said I haven't had games where I was unsportsmanlike, too passive aggressive with my teammates, or gave up a winnable game. Even on my Riot account. But that's the nature of League, it's a roller coaster of up and down emotions, and I think the best we can ask of ourselves is not let the "downs" bring our teammates down. I've been guilty of that, and it genuinely feels bad when I reflect on it. (3) While I don't think I "flame", I do love me some spicy banter with the enemy team... I'm probably one of the Rioters most willing to engage in trashtalk. But I keep that for the enemy to tilt them. Plus if I'm on my Riot account I only do it if they start it first - if you through down the banter gauntlet then I think it's fair game ;)
: When A riot member doubts your off meta pick....
DUDE THIS IS ME! Okay I was also confused because there was a duo in the room, and you picked Bard, so I thought you might be swapping. Also my dude, Bard is one of my favorite champions. I used to one-trick him before I became a toxic Yasuo main. You played that shit top so well. I actually don't think he's terrible top right now? I mean you can scale into a monster with the stacking changes to meeps. GG WP M'dude. Also some Bard love for you: http://na.leagueoflegends.com/en/page/bards-big-birthday-datapalooza
: Really interesting to me that Malphite is in the babysitters category. My guess would be because he pretty much plays a defensive lane until level 6, and doesn't roam until after that point. Still, interesting to see where some of these champs fall.
Yeah I replied to a comment about this on reddit, but I think it's a mix of (1) playing defensive until he's 6 and then (2) typically his first ult is used with his lane partner for some kind of all-in. His 2nd ult probably isn't ready until ~11-13 minutes which is just as laning phase starts to end.
: I like this article, it's pretty eye-opening. I know this is not an exact science but I want to point out the positioning of the minions on the Tempered Fate graph is not entirely true. This might affect the calculations. The minions were not in a straight line & Gangplank slowed the 1st one (which is the reason it was caught in the ultimate). Just wanted to point that out. All in all I appreciate what you're doing. Keep being awesome.
(1) True he positioning in the _diagram_ of the play is not 100% accurate because of the gangplank slow, however... (2) for the sake of the _actual calculation_, the last two minions in the line (the ones that matter for the calculation) weren't effected be the gangplank slow. The only movement in the end of the minion line was the 2nd to last minion gets pushed up a _little bit_ by the Quinn. I don't think it was that impactful, and like you said, not perfect science ;)
: "hard" versus "soft" CC is kinda ambiguous. Especially, are Roots and Silences hard or soft, when they let you cast spells/move resp. That being said, even a VERY basic "inflicted CC time" and "amount shielded", even with errors and not completely relevant measures (I mean, the current "amount healed" is pretty sh*t too, no offense) would feel very gratifying for support players, and could help gauge your performance on the same champ from game to game. I'm not talking game analytics here, more players feels and satisfaction ;) So probably not your garden, but if you can forward that to the relevant person...
Right I mean healing has the biggest problem of not distinguishing self and ally healing. It's mostly reflective of ally healing until you get champions like Trundle or TK who have massive self healing. If you had comprehensive "CC Measures" you could give specific times and categories of CC inflicted. For instance, Soraka players could have numbers like "X healing to allies, y damage to enemies, z seconds slowed, q seconds silenced, and j seconds rooted". Still there are a lot of non-measurable impacts that supports (and all champions) bring to a lane and to a game. Basically the concept of a Blitz who hasn't used his hook exerting pressure in a lane but magnified to tons of different abilities, CC, and utility (again in the Soraka case, the power of her E is greatly in the zoning it creates and less-so in whether you land a root or not).
: > [{quoted}](name=Riot Blaustoise,realm=NA,application-id=Ag8jgd8Q,discussion-id=YAHNpven,comment-id=00180000,timestamp=2016-03-21T19:30:27.685+0000) > > What I think you might also be interesting in is a **total output index** that measures pure numerical output. That scale would be **damage + healing** so on the upper end you'd have pure number output champs and on the bottom end you'd have more utility based champs. It would look something like... > > **TOTAL OUTPUT INDEX** (damage + healing) > {{champion:16}} 27,215 > {{champion:63}} 24,172 > {{champion:37}} 23,559 (low on the BBI, but it's because she does both a lot of healing _and_ damage) > {{champion:143}} 23,230 > > ... > > (bottom end; all utility champs including a lot of shielding with Janna & Lulu) > {{champion:201}} 11,452 > {{champion:40}} 11,079 > {{champion:412}} 10,851 > {{champion:117}} 10,472 Well, shielding should be measured and have a numerical value, but that's a whole other debate, and not one for today :p And yes, Sona (and Nami) were the people I had in mind that would be grouped with guys like Leona/Thresh when it's not exactly relevant. I think the scatter-plot+clustering approach would also identify fairly well the "numerical output" champs vs the "utility champs": total output is the distance from the origin, so "utility" champs would be gathered close to the origin, while the "numerical" champs would be very far, either close to one axis (Soraka/Brand) or close to the diagonal (Sona) but still far from the origin anyway. Random Rambling: I'm curious to see if an unguided method would yield clusters with an intuitive meaning for humans (like Healers vs Damage vs Utility) or if statistical relations would create strange groups. It's a common problem with classification algorithm/machine learning that they reach good solutions but can't explain why they are good. Like how, on a whole different scale, AlphaGo manages to beat the world champion but cannot **teach** you why its moves are good. Anyway, that was a very interesting article+forum thread, hope there will be more like this one or like the warding one soon ;) Stuff like Flash-key are fun but more anecdotal, I think ^^'
Hey Deathburs7 I got you fam ;) http://imgur.com/t12RKuq
: Here's a fun thought exercise - supposing support champions all have a 50% winrate (in practice, not quite true but fairly close), this means that they provide similar overall 'value' to a team. So this means that the difference on this total output index can be interpreted as the value of their non-damage oriented skills. For example, we can assume that Lulu contributes 17k worth of non-damage utility since she averages that much less damage. It's also worth noting that a lot of the big numbers come from AOE. Soraka gets a global heal and Sona gets aoe damage and heals, and Brand and Zyra can slum dunk an entire team with their ult. Their AOE abilities focus on damage, whereas someone like Thresh or Braum has their AOE oriented around CC.
Yeah I love this line of thinking. Theoretically we have some measure of CC (although it includes any sort of slow), but currently we can only measure 'value' as healing, damage, and utility (utility being really nebulous as it means CC, shields, and spells like Thresh Lantern). Would be really cool to have CC broken down into soft vs hard, measure shielding, also buffs or enhancements to allies, and then have 'utility' mean tougher to measure skills like Thresh W or Bard E. You could create a massive index of a support's 'value' based off all those metrics.
: And yet we still have trouble calculating the velocity of an unladen swallow.
Actually it's about 24 MPH or 11 meters per second {{champion:34}} {{item:3285}} {{item:3009}}
: Also would it be possible to see roval data on the other supports?
Made a comment in the main thread that has the breakdowns of both the Rovals and the BBI ^__^
Jolg (NA)
: Interesting. Though I think {{champion:48}} might be an anomaly due to his passive and W. That is probably all self healing.
Yup. Definitely from his passive. Healing counts self-healing as well (hence why Tahm Kench is pretty low too).
: I'm really perplexed by the BBI... The fact that it's a difference means that people who heal AND burn for 10K each are grouped with people who NEITHER heal or burn at all, if you want to use it as a comparative tool, so it's not really relevant. If you use it only to evaluate what's the "main" role of a champ **independently of other champs**, it's less strange, but I'm not sure I understand the point, because the fact that Bard deals more damage than he heals doesn't really mean he is allowed to misplay its shrines, they are still an important part of its power budget, even if less than the half. My gut feeling here is that a scatter plot with amount healed on one axis and damage dealt on the other, and then a very basic clustering method (maybe K-means with manually trying 2 or 3 values? You have only 25 champs after all, it's fast) would be much more useful and not much more complicated, so a better return on time spent overall. (I hope this doesn't come out as arrogant or anything ^^ I'm just really interested in this kind of data-analysis, so I try to participate, just my 2 cents ;) ) EDIT: On a related note, how did you clustered the 4 kinds of supports (Baby-sitter, care-taker, play-maker, bard) in the roaming part? EDIT2: Ah, this was already answered: 4 buckets of 8. My bad.
This doesn't come off as arrogant - I actually agree with your points! The point of the BBI is just to understand a support's primary output in terms of damage and healing. Ofc Bard's shrines are still really important to his laning. What I think you might also be interesting in is a **total output index** that measures pure numerical output. That scale would be **damage + healing* so on the upper end you'd have pure number output champs and on the bottom end you'd have more utility based champs. It would look something like... **TOTAL OUTPUT INDEX** (damage + healing) {{champion:16}} 27,215 {{champion:63}} 24,172 {{champion:37}} 23,559 (low on the BBI, but it's because she does both a lot of healing _and_ damage) {{champion:143}} 23,230 ... (bottom end; all utility champs including a lot of shielding with Janna & Lulu) {{champion:201}} 11,452 {{champion:40}} 11,079 {{champion:412}} 10,851 {{champion:117}} 10,472
Dessem (EUW)
: There were three things I took away from this: Holy SHIT that play was amazing. There's a BBI for Supports, and Janna actually heals more than she does damage... despite her ult being her only heal. **Riot has tools to track the position of champions at any point in the game?! :OOOO**
sp00ky right? :OOOOO And yeah, honestly that play was the most impressive Bard play I've seen. Sure the great-Bard-chases-&-clown-fiestas are always fun to watch, but there was something so simple and poetic about this one ^__^
: tHIS WAS AN AMAZING READ ty, [bard noises]
Bioxic (NA)
: As a math nerd and support main, this article made me happy. I also want to pick up Bard now. Good article, Blaustoise!
Anytime bud! Also if it helps picking up Bard... here's what I'm running these days: Standard: {{item:2049}} {{item:3158}} {{item:3092}} {{item:3110}} {{item:3190}} {{item:3222}} {{item:3065}} Split push: {{item:3512}} {{item:3060}} Light em up: {{item:3094}} {{item:3285}} {{item:3100}} {{item:3115}}
: Bard's Big Birthday Datapalooza
I posted the full breakdown of the BBI in a comment in this thread, and since a few people have requested the Roval data and where other supports fall here it is: **ROVAL RANK BREAKDOWN** {{champion:432}} 18.28 million units^2; red side Roval Playmakers {{champion:98}} 17.80 {{champion:53}} 17.08 {{champion:48}} 16.76 {{champion:78}} 16.54 {{champion:63}} 16.20 {{champion:1}} 15.72 {{champion:412}} 15.71 {{champion:12}} 15.36 Caretakers {{champion:43}} 15.32 {{champion:111}} 15.27 {{champion:143}} 15.19 {{champion:99}} 15.16 {{champion:26}} 14.95 {{champion:89}} 14.52 {{champion:25}} 14.29 {{champion:117}} 14.25 Babysitters {{champion:54}} 14.17 {{champion:201}} 14.11 {{champion:223}} 14.03 {{champion:37}} 13.95 {{champion:267}} 13.73 {{champion:40}} 13.59 {{champion:44}} 12.81 {{champion:16}} 12.54
: I'd just like to see what data you came up with to be honest. I understand that your research was very Bard-centered, so your numbers on other champions would be secondary. Still, It would be interesting to see where different supports fall on the BBI. If you ever felt up to it, maybe you could do an X/Y graph with BBI on one axis (as raw damage vs. healing) and have the other axis be utility focused. You could do seconds in a game supports locked down an enemy with CC (offensive utility) and how many seconds in a game supports were buffing allied champions with damage/movespeed/shields/resistances/etc (defensive utility). I'm not sure if you would be able to pull data like that, but it would be interesting to see. How do you think his clear speed as a jungler would be with Runic Echoes? Or is the primary problem just his first/second clear?
I just replied to a comment with the full breakdown of the BBI ^__^ And yeah I think it's his first/second clear that gives jungle Bard problems {{champion:432}} {{summoner:11}} {{item:3070}}
: Really interested on...basically everything about this that doesn't involve Bard, really. (My primary thought on Bard is something similar to the old 'Your Vayne, their Vayne' videos.) Are those terms 'Babysitters, Caretakers, Playmakers' official subgroups of supports, or just something you came up with? Do you feel like sharing the full BBI? If nothing else, I am really glad that you're sharing these kinds of analytics.
'Babysitters, Caretakers, Playmakers' was just a simplified subgroup for supports just to make the article lighter. Since we I took the 25 "top support game" champions, it was simple to have Bard + 24 other champions and just divide the other 24 into buckets of 8. As for sharing the overall BBI... sure, why not! Damage > Healing {{champion:63}} 21273 {{champion:143}} 20694 {{champion:99}} 17356 {{champion:1}} 14271 {{champion:43}} 14189 {{champion:25}} 12196 {{champion:54}} 11002 {{champion:26}} 10551 {{champion:78}} 8016 {{champion:117}} 7982 {{champion:53}} 7636 {{champion:111}} 7435 {{champion:432}} 6817 {{champion:412}} 5763 {{champion:89}} 5610 {{champion:98}} 5493 {{champion:201}} 4897 {{champion:37}} 4345 {{champion:223}} 4057 Healing > Damage {{champion:16}} 18010 {{champion:44}} 6159 {{champion:48}} 2722 {{champion:12}} 2275 {{champion:267}} 1098 {{champion:40}} 469
Tafiir (NA)
: IT'S CHIME TIME GUYS {{champion:432}}
http://i.imgur.com/g2wKkCL.gif
: Two things! One, can we get a mini article discussing the entire breakdown of your general support analysis? Seems like some really interesting data that could provide a lot of insights into champion roles and play styles. Two, do you play Bard in other roles, or just support?
I'm not sure about the first thing, but feel free to ask any questions if you're curious about specifics! For instance, I responded to someone else in the thread with another way that you can use healing + damage to compare supports based on raw numerical output versus utility + buffs + CC. I've only really played him support. I took him in the jungle when he first came out but his clear speed it pretty weak. I have been known to occasionally build him AP as a carry in a botlane or top though{{item:3285}} {{item:3100}} {{item:3089}}
: Stats, graphs, and equations, oh my! This was great! As a Leo main, I'm left wondering where she fell on roams and BBI? {{champion:89}} PRAISE THE SUN {{champion:89}}
Leona falls at 16 out of 25 on the BBI! She's just behind Thresh and just ahead of Shen. Another way we can use the BBI those is _add damage and healing_ to get total output. Leona's overall damage + healing output is really small but that makes sense because her core strength is hard CC. Other champs that have low overall output are Thresh, Braum, Janna (since shields don't count), etc. On the other end you have Soraka, Brand, Sona, Zyra, etc. It's sort of a measure of how much utility that support typically has versus raw numerical output.
: As a bard main, this article was a pleasure to read! Many thanks.
As another bard main to a bard main, it was my pleasure ^__^

Riot Blaustoise

Level 227 (NA)
Lifetime Upvotes
Create a Discussion