EPTInteresting that both get released (probably) within a patch of each other :p
On the other hand, two new delayed skillshot based squishies to be fotm, a good time for lb :D
[Patch 4.2] Xerath/Skarner General Discussion - Page 17
| Forum Index > LoL General |
|
killerdog
Denmark6522 Posts
Interesting that both get released (probably) within a patch of each other :p On the other hand, two new delayed skillshot based squishies to be fotm, a good time for lb :D | ||
AsmodeusXI
United States15536 Posts
On February 12 2014 06:55 Ketara wrote: It's just such a dated joke. I mean, look at the last champ who came out. He doesn't have a self rez! In fairness, I mostly wanted to say that line. | ||
|
Fusilero
United Kingdom50293 Posts
On February 12 2014 06:55 killerdog wrote: If you ignore the ults, then xerath and vel'koz are the same champion. (xerath q = velkoz w, xerath w (with stun) = velkoz e, xerath e = velkoz q (with stun)) The only differences are the ults, and small differences in how much damage the skills do. Interesting that both get released (probably) within a patch of each other :p Yaaay more sieging mid laners, can't get enough of those. | ||
|
Alaric
France45622 Posts
| ||
|
wei2coolman
United States60033 Posts
| ||
|
dae
Canada1600 Posts
On February 12 2014 06:53 Goumindong wrote: So a little probability lesson for today When discussing probabilities in that manner the only thing that is generally being looked at is the calculated probability for an individual segment. Not the probability for what actually happened. People who get mad at streaks are generally exhibiting an observation bias. E.G. Lets suggest that we play 9 games in a row and we want to know the probability of seeing, in those 9 games a situation where we get FP/LP on our duo 3 times in a row. Well there are many more combinations than is immediately obvious as to how many combinations exist and the individual probability of each one, while very low, all gets added together. For instance there are 16 combinations in which any particular streak exists, there are 5 potential streaks but some of the streaks overlap and are the same combination. Assuming independence this puts a maximum value on the likelihood at 8%(it will actually be slightly smaller than this). As you play more games the probability of a streak of 3 increases. So its unlikely but not too unlikely.... Until you take it to its logical conclusion. Suppose that every TL' general played those same 9 games and suppose furthermore than those which did not see an uncommon situation did not come into the thread to report it, but only some fraction of the people who did see an uncommon situation did. So now we have say 50 people, each looking at a 7% probability that this 3 streak or more happens. The probability that none of the 50 people who report to the chat encountered the completely random situation in which they got a streak of 3 in a row is about 2%. Now here is the real thing. If you were getting the first/last before the patch then there would be zero games in which you also did not get the same first/last after the patch if nothing was changed. So testing whether or not its working is fairly easy. Has anyone duo'd and not gotten first/last with people they know have divergent MMR? If yes, its probably just that you got unlucky Given the time of day, and the fact that the patch just came out, and my respect for the original posters intelligence, I made the logical assumption that those 3 games were the only 3 games he had duoq'd since the patch. With that assumption, and the fact that riot does make stupid mistakes sometimes, and that it would only take 1 person to have duoq'd without this observation to disprove it, it's pretty obvious that there was a quite large chance that riot screwed up. Say theres a couple hundred people that post in this thread. Say that there is 50 that have played since the patch. The chance that 1 out of those 50 had a 1/1000 chance in the first few games he duoq'd is pretty low. And that no one has come out to say that it didn't happen for them, just makes it more likely. And, I find it funny you posted a wall of text in response to a basic "what are the chances this happened" post. | ||
|
Gahlo
United States35142 Posts
TSM TSM TSM | ||
|
Zess
Adun Toridas!9144 Posts
On February 12 2014 06:53 Goumindong wrote: So a little probability lesson for today When discussing probabilities in that manner the only thing that is generally being looked at is the calculated probability for an individual segment. Not the probability for what actually happened. People who get mad at streaks are generally exhibiting an observation bias. E.G. Lets suggest that we play 9 games in a row and we want to know the probability of seeing, in those 9 games a situation where we get FP/LP on our duo 3 times in a row. Well there are many more combinations than is immediately obvious as to how many combinations exist and the individual probability of each one, while very low, all gets added together. For instance there are 16 combinations in which any particular streak exists, there are 5 potential streaks but some of the streaks overlap and are the same combination. Assuming independence this puts a maximum value on the likelihood at 8%(it will actually be slightly smaller than this). As you play more games the probability of a streak of 3 increases. So its unlikely but not too unlikely.... Until you take it to its logical conclusion. Suppose that every TL' general played those same 9 games and suppose furthermore than those which did not see an uncommon situation did not come into the thread to report it, but only some fraction of the people who did see an uncommon situation did. So now we have say 50 people, each looking at a 7% probability that this 3 streak or more happens. The probability that none of the 50 people who report to the chat encountered the completely random situation in which they got a streak of 3 in a row is about 2%. Now here is the real thing. If you were getting the first/last before the patch then there would be zero games in which you also did not get the same first/last after the patch if nothing was changed. So testing whether or not its working is fairly easy. Has anyone duo'd and not gotten first/last with people they know have divergent MMR? If yes, its probably just that you got unlucky edit: Another way to think about it is by looking at the lottery. The probability of winning the lottery is something like 1/200,000,000. But this does not mean that no one wins the lottery because individually its extremely unlikely. It also doesn't mean that the actual probability isn't 1/200,000,000 if you(or your buddy, or your work mate, or someone in your city) wins. I hope for edification of those in your vicinity that your current employment is wholly unrelated with education in statistics and/or probability. | ||
|
cLutZ
United States19574 Posts
On February 12 2014 07:04 Gahlo wrote: http://esportsexpress.com/2014/02/team-solomid-acquires-heroes-of-newerth/ TSM TSM TSM HoN still exists? | ||
|
Goumindong
United States3529 Posts
On February 12 2014 07:10 xes wrote: I hope for edification of those in your vicinity that your current employment is wholly unrelated with education in statistics and/or probability. Why? What was the issue with the post? Say theres a couple hundred people that post in this thread. Say that there is 50 that have played since the patch. The chance that 1 out of those 50 had a 1/1000 chance in the first few games he duoq'd is pretty low. And that no one has come out to say that it didn't happen for them, just makes it more likely. About 5%, so not terribly low. The idea that people haven't come out to say that they have is probably because the likelihood of people looking at the thread is low compared to the people who go to post an experience. | ||
|
Fusilero
United Kingdom50293 Posts
There are dozens of players. At any time there may be a literal plurality of matches. Serious note though from what I've heard it's still popular in SEA | ||
|
cLutZ
United States19574 Posts
On February 12 2014 07:15 Fusilero wrote: There are dozens of players. At any time there may be a literal plurality of matches. Serious note though from what I've heard it's still popular in SEA http://www.quickmeme.com/meme/3pk5la ? | ||
MoonBear
Straight outta Johto18973 Posts
On February 12 2014 06:53 Goumindong wrote: [SNIP] Okay.
I am going to kill this stats discussion right here and now because it is so bad. | ||
|
Celial
2602 Posts
| ||
|
thenexusp
United States3721 Posts
User was warned for this post | ||
|
Gahlo
United States35142 Posts
On February 12 2014 07:17 Celial wrote: Runes/Masteries for Aatrox and Darius? For Nocturne I'm going to try Ninjakens guide. When I played Darius it was typical bruiser ArPen/armor/mr/AD. AD marks are always an option. | ||
|
canikizu
4860 Posts
On February 12 2014 07:15 Fusilero wrote: There are dozens of players. At any time there may be a literal plurality of matches. Serious note though from what I've heard it's still popular in SEA People play more HoN in SEA than Dota2 in Korea. Not to mention Tencent just got distribution right for HoN in China. Apparently there will be a new blind-pick mode, you can't see the other team picking, you don't lose gold when dying, and everybody has one free dedicate courier. http://vimeo.com/80659594 | ||
|
Zess
Adun Toridas!9144 Posts
On February 12 2014 07:19 thenexusp wrote: All of stats nonsense could be solved by a couple other duo queuers (of significantly different ranks) going into duo queue and posting their results. one non first/last pick would prove that riot didn't screw up, and even a couple more first/last picks would greatly strengthen the hypothesis that riot did screw up. Not necessarily, before this patch I had duoq that were in the middle of the pick (it happens very rarely though). So you'd have to first sample for the prior, then assess whether the posterior is statistically significant. User was warned for this post | ||
|
TheYango
United States47024 Posts
On February 12 2014 07:21 canikizu wrote: Not to mention Tencent just got distribution right for HoN in China. Apparently there will be a new blind-pick mode, you can't see the other team picking, you don't lose gold when dying, and everybody has one free dedicate courier. Doesn't HoN already give your team a free courier? Or do you mean per-player? | ||
|
oneofthem
Cayman Islands24199 Posts
| ||
| ||
