Rioter Comments
: Learn Statistics and Data Mining with League of Legends Data - Principal Component Analysis
This is a test case for my Reddit bot. Please ignore. ***** Picture test: http://media2.giphy.com/media/ic9U7hTpHIycw/200_s.gif http://media2.giphy.com/media/ic9U7hTpHIycw/200_s.gif_ http://media2.giphy.com/media/ic9U7hTpHIycw/200_s.gif _ media2.giphy.com/media/ic9U7hTpHIycw/200_s.gif_ media2.giphy.com/media/ic9U7hTpHIycw/200_s.gif ***** Bold test: This is **bold**. Normal asterisks are like this: \*, \*\*, \*\*\* ***** Italics test: This is _italics_ Normal underscores are like this: \_, \_\_, \_\_\_ ***** Strike-out test: This is ~~strikeout~~ Normal tildes are like this: ~, ~~, ~~~ ***** Combination test; ** boldbold _bold_ ~~bold~~** _Italics Italics **Italics** ~~Italics~~_ ~~strikeout strikeout **strikeout** _strikeout_~~ ***** Links test: [Link here](https://www.reddit.com/r/sufficiencybot/comments/3bnxfc/riotsboardmessageincommentbot/) ***** Champions test: {{champion:266}} {{champion:103}} {{champion:84}} {{champion:266}} {{champion:103}} {{champion:84}} Invalid champion: {{champion:999}} ***** Summoner spells test: {{summoner:21}} {{summoner:2}} {{summoner:13}} Invalid summoner: {{summoner:99}} ***** Items test: {{item:3001}} {{item:3133}} {{item:1327}} Invalid item: {{item:9999}} ***** Stickers test: {{sticker:slayer-jinx-catface}} {{sticker:zombie-brand-clap}} ***** Numbered list test; 1. 1 2. 2 3. 3 ***** Unnumbered list test: * 1 * 2 * 3 ***** Quotation test > To be or not to be, that is the question With names: > [{quoted}](name=Insufficient RAM,realm=NA,application-id=2XjzURgc,discussion-id=xuEZY6bG,comment-id=,timestamp=2015-05-13T04:31:38.585+0000) > > Blah blah and > [{quoted}](name=William NOMIS,realm=EUW,application-id=cIfEodbz,discussion-id=Oy9l7HgR,comment-id=0000,timestamp=2015-12-11T13:28:23.755+0000) > > Very interesting ! :) broken quote1 > [{quoted}](name=William NOMIS,realm=EUW,application-id=cIfEodbz,discussion-id=Oy9l7HgR,comment-id=0000,timadestamp=2015-12-11T13:28:23.755+0000) > > Very interesting ! :) broken quote2 > [{quoted}](name=William NOMIS,realm=EUW,application-id=cIfEodbz,discussion-id=Oy9l7HgR,comment-id=0000,timadestamp=2015-12-11T13:28:23.755+0000 > > Very interesting ! :)
Rioter Comments
Rioter Comments
Rioter Comments
Rioter Comments
Rioter Comments
: Did it, neat stuff! I notice it takes into account ARAM games as well. If I may suggest to filter those out as you do not have control over that champion choice if possible? Otherwise, good job! I do love me some Lucian games. Note: response time ~ 3 minutes. Not bad. Edit: Oh and I would also suggest re-posting this at around 2:00-6:00pm EST to get a bit more exposure. The boards are a bit slow this time of day. Hell, I shouldn't even still be awake....lol
It does take ARAM into account. I could either take a player's ranked games (can be tracked back many games), or look at the most recent any 10 games. I chose the latter since it allows everyone to participate, not just the players with ranked records. I'll consider revising this though... factoring in ARAM doesn't really make sense! As far as response time is concerned, it's probably because the bot (or reddit as a whole) was a bit overloaded... when I was testing the bot, the response was within 5-10 seconds. Thanks for trying it!
Rioter Comments
TehNACHO (NA)
: Solution: Smite Gromp and Chicken, go gank, repeat. Don't power farm so much that you'd need to Smite Krugs.
I already do that. But essentially I can't Smite the golems... which is super awkward. It also forces me to start on the blue side since red side start is simply not viable (again, due to Heavy Hand).
Rioter Comments
Sargonas (NA)
: We've actually got a whole thread tinkering around on this over in the Dev Forums. The TL;DR is that the match history uses the X-Y cords from the game, and the in-game map is actually a bit larger due to the "off the map" space on the borders in the sky boxes beyond the normal game play area. The map image however is cropped down (and different dimension ratios), so the numbers don't line up. We've been looking at ways to make this less painful for folks but the reality is there is no easy solution, even Match History has to do some magic shenanigans with the numbers to make it look ok, and even then it isn't perfect.
As far as I can tell through my reverse engineering (simple linear algebra), the official match history's map use the Riot API X-Y cordinate subject to a scaling multiplier. No noticable cropping/translations that I can detect (beside flipping the Y-axis because on the webpage the Y-axis points downward). I am not sure if it has any magic to it except the dots are huge so it's hard to see that they are misplaced. But anyway, I am happy to hear that this is being looked into!
Sargonas (NA)
: This is really slick! I love it! Gonna dig around and poke at it more to see if I have any useful feedback for you, but at first glance is this pretty cool. I love it when something new comes alive from our Community Developers! One thing I did notice.. it threw an error saying there was no Match history for my name however, the error where it suggests you check your realm. On a hunch, after that happened I waited a minute or so and tried again and it worked fine. Likely some kind of data caching issue on the first ever lookup of a Summoner name I presume? Might want to see about adding in a custom return for those scenarios.. but nothing critical.
While you guys are at it, can Riot provide a better map? Riot's official matchhistory and this website both use this image as the map: http://i.imgur.com/RXRo2Wo.png But this map has some problems. Not only is it wrong (just look at the dragon pit), it does not actually correspond to the actual map coordinates from Riot API that well either - you can't just use the X-Y coordinate from Riot API (which ranges from 0 - 15000) then divide by a constant (for scaling) and place it on that map because everything is a little off. It would be nice if Riot can provide an official version of the map that matches the coordinates from Riot API (subject to scaling) and is fairly accurate.
: Nemesis an interesting game mode. It's almost like an anti-meta experiment, and I'm sure it'll shake things up. The other day four friends and I played a Nemensince game against a team who gave us five ADCs. We gave them three tanks before we realized what they were doing, and decided to give them two more tanks to see what happened. They ended up buying five thornmails into our five black cleavers and beat us pretty bad. But it was an interesting game. Is this the type of game you guys hoped to see when you setup this game mode? What other weird (i.e., non-meta, non-traditional) team comps do you thin would be fun? After Nemesis is over, could you guys release some statistics about the games played? OKCupid writes some interesting articles about the statistics from their online dating website, and I think it would be fun to see similar articles written by Riot about these unique game modes.
Shameless plug if you are intetested http://www.goldper10.com/article/881-nemesis-draft-win-rates-and-analysis.html
Krosan (NA)
: Are you guys using data from Nemesis Draft as a data point on the general perceptions of champions? It's not a big surprise that I see champions like Eve, Yorick, and Urgot given out frequently, but it *was* fairly shocking for me the first time I realized exactly how badly people at my level undervalued support picks. I've ended up against teams with four supports, then got blamed for the loss because I gave the enemy Rumble or someone else without useful CC.
http://www.goldper10.com/article/881-nemesis-draft-win-rates-and-analysis.html
Rioter Comments
RiotChun (NA)
: Mordekaiser 5.4 changes
I like these changes. I think this is the way to go. While you are at this, can you guys look at the ghost's interaction with items/certain passives? Currently, the ghost's items sometimes work (notably Frozen Mallet) but not others (Runaan's Hurricane); simiarly, the ghost's passive sometimes work (e.g. Rengar's Leap) but not others (Yasuo's shield, Kalista Martial Poise). I am sure there are other bugs/features out there for ghosts, but as a player it is very difficult to test them (e.g. does penetration work on ghosts? Does it use physical or magical penetration? Does Spellsword (the mastery) actually work on the ghost? Does Nashor's Tooth work?
Rioter Comments
: I do not play Ahri often, did she need a buff?
I kind of like the changes. I like it because it essentially reverts her back to the 'old Ahri' without the 'must land E' business. I think she will have a lot of potential and enjoy a playstyle similar to in S2 with Rylai's being core.
Riot Jag (NA)
: [5.2] Live Gameplay Patch Forecast
Let me preface with this: I think DFG needs fixing. However, in my opinion, removing it outright is too drastic (even if you decide to replace it with something else) and there are other approaches. First of all, I want to point out that Lich Bane actually had the same problem before. Lich Bane was: 1. Skews power of many champions (I can get nerfed, but I can always buy Lich Bane to normalize my power). 2. Only a handful of champions can utilize it optimally. This is similar to DFG, which is an item that only a few champions can use (since you need to unload in less than 4 seconds) and normalize the power of those champions. Yet Lich Bane survived by just tweaking its numbers. I don't see why DFG cannot simply be adjusted. For example, the following can be done. First get rid of the flat 15% max health damage - which IMO is fundamentally broken and very out of place - and change it to flat damage. Secondly, instead of damage amplification for 4 seconds, make it store damage for 4 seconds then unleash it as DoT after 4 seconds. It'll be kind of like a Zed ultimate, but provides the victim with many ways to stay alive instead of being instantly wiped off the map. Lastly, I want to point out that DFG has been in the game FOR FIVE YEARS. When you remove something which many players feel is fundamental about the game for so long, it's actually a betrayal to the player base. I am sure a compromise can be made about this. DFG does not need to just simply go away.
Rioter Comments
Worre (EUW)
: There's some problem with the buttons/forms and their JS on the normal calculator that causes it to not work on Chrome. the advanced calc is fine so you can cross reference the JS and see what's wrong.
> [{quoted}](name=Worre,realm=EUW,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=000d,timestamp=2014-12-25T02:14:25.221+0000) > > There's some problem with the buttons/forms and their JS on the normal calculator that causes it to not work on Chrome. the advanced calc is fine so you can cross reference the JS and see what's wrong. I can copy the code from the advanced calculator to the one that does not work on Chrome - now you would think that would fix things, but surprisingly it does NOT fix it. The issue is with the page setup itself. I am very convinced the issue is with Google Site or Chrome.
Rioter Comments
Odenmaru (NA)
: And so began the age of the great war. The war against the tyranny of RiotAmes.
> [{quoted}](name=Odenmaru,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=0009,timestamp=2014-12-24T04:18:51.424+0000) > > And so began the age of the great war. > > The war against the tyranny of RiotAmes. It actually began 10 months ago when I made and published the first version of this calculator!
Kakwane (NA)
: Ames, I challenge you! I have achieved 0RP!
http://boards.na.leagueoflegends.com/en/c/GD/JycNIgdp-the-0-rp-calculator-it-tells-you-how-to-get-to-0-rp ^___^
Arkara73 (NA)
: It seems to be working on everything aside from Chrome
> [{quoted}](name=Akali is SO HOT,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=0007,timestamp=2014-12-24T03:48:14.888+0000) > > It seems to be working on everything aside from Chrome Actually, the "secret page" works on Chrome too. https://sites.google.com/site/the0rpcalculator/calc/i-know-what-i-am-doing Which is basically the same thing except it has two extra buttons. So I think it is fixable, but I did a compare and failed to find anything alarmingly different. I kind of blame Google Site at this point for compiling my Javascript file into a jumble.
: > [{quoted}](name=Bad Morde Bot,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=00030000,timestamp=2014-12-24T03:34:31.501+0000) > > Try it on Safari? Let me know. Working on Safari.
> [{quoted}](name=If I were Jesus,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=000300000000,timestamp=2014-12-24T03:40:00.451+0000) > > Working on Safari. I think I can reproduce the issue now. Let me try to fix it. Thank you.
: Not working. I'm on Chrome and Mac if that means anything. Cool otherwise for the people that it works for.
> [{quoted}](name=If I were Jesus,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=0003,timestamp=2014-12-24T03:33:54.966+0000) > > Not working. > I'm on Chrome and Mac if that means anything. Cool otherwise for the people that it works for. Try it on Safari? Let me know.
: You started with 1888 RP. To get to 0, you need to: Buy 650 RP 2 time(s). Buy 1x 750RP Skin. Buy 1x Mystery Gift (490RP). Buy 4x 975RP Skin on sale (487RP). 1888 + 2 * 650 - 1 * 750 - 1 * 490 - 4 * 487 = 0 Have fun! Huh, cool!
> [{quoted}](name=Best Furry NA,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=0001,timestamp=2014-12-24T03:29:42.780+0000) > > You started with 1888 RP. To get to 0, you need to: > Buy 650 RP 2 time(s). > Buy 1x 750RP Skin. > Buy 1x Mystery Gift (490RP). > Buy 4x 975RP Skin on sale (487RP). > 1888 + 2 * 650 - 1 * 750 - 1 * 490 - 4 * 487 = 0 > Have fun! > > Huh, cool! Oh man. The asterisks are clashing with the formatting for Riot's board. I changed the asterisks to "x".
: Not working. Clicking calculate and nothing is happening.
> [{quoted}](name=Reaperinator,realm=NA,application-id=mNBeEEkI,discussion-id=JycNIgdp,comment-id=0000,timestamp=2014-12-24T03:26:49.034+0000) > > Not working. Clicking calculate and nothing is happening. Oh man. Can you tell me which browser you are using? I only tested this for IE ~~and Chrome~~ (looks like Chrome no longer works for some reason...)... if it doesn't seem to work, try a different browser. Also you need to be able to run Javascript.
Rioter Comments
Zielmann (NA)
: This is strictly a guess, but I could see it being that way because ARAM players are used to the viewing angle on the map. The zoom and angle is slightly different than what most players will be used to from SR. And it would make sense that the blue side is easier for players to adapt to, since it has better vision. People who have played a lot of ARAM would be more used to the angle you get on the purple side, so I can see it making some sense that they fare a little better on purple in ARAM than all these non-ARAM players who are experiencing that camera angle for the first time now.
> [{quoted}](name=Zielmann,realm=NA,application-id=LFfTlAky,discussion-id=tXZIVZty,comment-id=000000000000,timestamp=2014-12-18T03:55:40.140+0000) > > This is strictly a guess, but I could see it being that way because ARAM players are used to the viewing angle on the map. The zoom and angle is slightly different than what most players will be used to from SR. And it would make sense that the blue side is easier for players to adapt to, since it has better vision. People who have played a lot of ARAM would be more used to the angle you get on the purple side, so I can see it making some sense that they fare a little better on purple in ARAM than all these non-ARAM players who are experiencing that camera angle for the first time now. Yes I can definitely see this being the case. What I am even more curious about is that this game mode is heavily reliant on hitting a very long and narrow skillshot - the Poro Toss. It would be interesting if we can see chance of landing it from the blue side and from the red side.
Zielmann (NA)
: Seasoned ARAM players could have told you this same stuff ages ago. I got to play two Sona games today. I was shocked she wasn't banned at all. And more shocked when the other team didn't even pick her both games. Our team won both games in about 11 minutes. Oh, and I believe in ARAM, it's about 55/45 in favor of blue side, so that's not that surprising, either.
> [{quoted}](name=Zielmann,realm=NA,application-id=LFfTlAky,discussion-id=tXZIVZty,comment-id=0000,timestamp=2014-12-18T03:21:42.155+0000) > > Seasoned ARAM players could have told you this same stuff ages ago. I got to play two Sona games today. I was shocked she wasn't banned at all. And more shocked when the other team didn't even pick her both games. Our team won both games in about 11 minutes. Oh, and I believe in ARAM, it's about 55/45 in favor of blue side, so that's not that surprising, either. Yes I recall around 55% win rate for blue side on ARAM as well. Not sure about the extra 3% on this mode though.
Rioter Comments
: That sounds amazing. Would you mind if I sent you a file and discussed how I'd like to incorporate the API? I would love some guidance.
> That sounds amazing. Would you mind if I sent you a file and discussed how I'd like to incorporate the API? I would love some guidance. I think you can just take a look at his wrapper, or someone else's. It's pretty straight forward. The wrapper will read out the JSON for you. The rest is just putting things into the database.
: I'm curious on the specifics, I've tried messing around with the API a bit for stats stuff. My skills aren't quite up to par at the moment.
> I'm curious on the specifics, I've tried messing around with the API a bit for stats stuff. My skills aren't quite up to par at the moment. I am using a Java wrapper by TheEnigmaBlade. I save the results with a postgreSQL database that runs locally.
: How did you mine the data to get all of the League names? Really well done, by the way.
> How did you mine the data to get all of the League names? Really well done, by the way. The data is mined using the Riot API and saved onto my local database.
: Good work, but you have quite a bit of time on your hands.
> Good work, but you have quite a bit of time on your hands. Ha, I don't think you know how much effort I actually spent on it.
: I know this isn't a serious statistical analysis, but here's the thing: when you're dealing with p-values and what is statistically significant or not, **do not tell me if something is statistically significant or not**. Tell me the p-value. I will make my own judgment of whether it's significant or not. It's slightly pedantic since I'm virtually certain you used an alpha of 0.05. But it's basically the same thing as Riot telling you, "yeah, you can't build AD on Malz because it doesn't fit our subjective paradigm of what is fun for you." Fuck that. I'll make my own decision on what's fun for me. Likewise, I will make my own decision of what I consider to be significant or not. Honestly, I would have probably gotten away with an alpha as high as 0.2 because this is a very non-serious thread. Of course, if you actually used an alpha of 0.003 or used a one-sided test and the p-value was only 0.01, I'd be like, "nah bro, that's totally significant." It's a minor thing because even if it were significant, the practical difference between the max and min of the data is so small it doesn't really matter. My comment is mostly for future reference.
> I know this isn't a serious statistical analysis, but here's the thing: when you're dealing with p-values and what is statistically significant or not, **do not tell me if something is statistically significant or not**. Tell me the p-value. I will make my own judgment of whether it's significant or not. It's slightly pedantic since I'm virtually certain you used an alpha of 0.05. But it's basically the same thing as Riot telling you, "yeah, you can't build AD on Malz because it doesn't fit our subjective paradigm of what is fun for you." > > Fuck that. I'll make my own decision on what's fun for me. Likewise, I will make my own decision of what I consider to be significant or not. Honestly, I would have probably gotten away with an alpha as high as 0.2 because this is a very non-serious thread. Of course, if you actually used an alpha of 0.003 or used a one-sided test and the p-value was only 0.01, I'd be like, "nah bro, that's totally significant." > > It's a minor thing because even if it were significant, the practical difference between the max and min of the data is so small it doesn't really matter. My comment is mostly for future reference. Hi, This is fair criticism. But due to my time constraint I was not able to publish all statistical analysis results including p-values. I will keep this in mind and present my findings in vigourous, scientific manner in the future. I know I can do better than this - I have a master degree in statistics with experience in composing articles for peer-reviewed publication. Should public interest arises I will be more than happy to amend the gaps in the original post as you outlined. Cheers, Sufficiency
: okay so whats the tl;dr basically? your against sexism? orrrr… because this sort of confused me
> okay so whats the tl;dr basically? your against sexism? orrrr… because this sort of confused me TLDR: actual research is done, but this post is as serious as Riot's claim that they are an equal opportunity league namer.
llamasx (NA)
: Dang Sona's 5 above average. Did you get who had the most and least leagues?
> Dang Sona's 5 above average. Did you get who had the most and least leagues? On top of my head, Kassadin has 47. A couple of Shadow Isle champions have sub-30 leagues. But do keep in mind that I might have missed a few leagues despite my best efforrt. If a Rioter can provide an official list (Riot DBA please) it would be awesome.
Rioter Comments
: Amusingly enough, those are relatively tame numbers.
> Amusingly enough, those are relatively tame numbers. Unpopular opinion time! I think Riot should allow EXP gain (but not IP gain) in passworded custom games post level 15. Botting in matched games is a lot more damaging than botting in passworded custom games.
Rioter Comments
Rioter Comments
Rioter Comments
Rioter Comments
Rioter Comments
: Damn. I have Rune Page 1 and Rune Page 2 (which isn't even finished) http://www.lolking.net/summoner/na/43472416#runes proof
Show more

Insufficient RAM

Level 34 (NA)
Lifetime Upvotes
Create a Discussion