| Forum Index > LoL General |
|
Sufficiency
Canada23833 Posts
On March 10 2014 07:46 iCanada wrote: Why would they even delete it? Dont make sense... No idea. My other post which I did things manually reached the front page. I don't see why this one is bad. But hey, at least lol reddit is moderated, where as here Cheep can just go around insult anyone at will. So it's a trade off. Anyway, new post here: http://www.reddit.com/r/leagueoflegends/comments/1zzzu3/rp_calculator_figure_out_how_to_get_to_0_rp/ | ||
|
Eiii
United States2566 Posts
On March 10 2014 05:42 chalice wrote: can you filter for league/division? would be interesting to see who the top 10-20 d1/challenger players are in categories like gold per minute. although i wonder if a stat like that somewhat self-filters, excepting smurfs, i would expect there would be a strong correlation between GPM and ELO. Shit, this is totally possible but my data collection wasn't set up for it. When I keep toying with this in the future, division info'll be the first thing I add. On March 10 2014 05:59 iCanada wrote: What champions tend to do the most damage and take the most damage? What kind of stats are there in terms of winning percentage? Just so you know the database's large enough that I sort of have to take a subsample of it-- so instead of 1mil games we're looking at 10k or so, so these results are about right but will change every time I run it. I can get better averages later but I'm not that patient for now lol + Show Spoiler + Top ten champs by average TOTAL damage dealt: +---------------+-------------+ Top ten champs by average damage dealt to champions: +---------------+------------+ Top ten champs by average damage received: +---------------+------------+ Top ten combined (dmg dealt to champs + dmg taken): +---------------+------------+ On March 10 2014 05:44 Alaric wrote: I'd be pretty interested if you have informations regarding wards, but I'm not sure how to judge it. Average wards per game? Average wards per minute? Average number of wards out on the map at the same time? Here's what I can do! + Show Spoiler + Average ward stats: +------------------+------------------+ Average ward stats for winning teams: +------------------+------------------+ Average ward stats for losing teams: +------------------+------------------+ | ||
|
iCanada
Canada10660 Posts
Like... I actually just straight up dont get it. Maybe its just bad golds being bad, but like... the number of times I've seen kids die to level 3 monkey ganks blows my mind. The guy will like walk through their ward, and then they just die. Any other junglert they walk through the ward and they just back and play like a not retard. I just... I dont understand. Clearly once it goes invis it stops existing. | ||
|
Eiii
United States2566 Posts
+ Show Spoiler + +---------------+------------+ | ||
|
Sufficiency
Canada23833 Posts
On March 10 2014 09:00 Eiii wrote: And just for fun, here's the bottom ten champs by total damage dealt + taken: + Show Spoiler + +---------------+------------+ This is really interesting. I assume you can get the statistics of each match, including the champions played in each match? It would be interesting to do some statistical analysis to figure out which champions counter which. | ||
onlywonderboy
United States23745 Posts
| ||
|
Eiii
United States2566 Posts
You could do some weird machine learning stuff but this dataset is so huge I don't even wanna try. It's theoretically possible I think, though I only have a little bit of knowledge about ML! | ||
|
Alaric
France45622 Posts
Amusing how people on average only buy one ward but drop seven, trinket OP.  | ||
|
Ketara
United States15065 Posts
On March 10 2014 07:59 Sufficiency wrote: No idea. My other post which I did things manually reached the front page. I don't see why this one is bad. But hey, at least lol reddit is moderated, where as here Cheep can just go around insult anyone at will. So it's a trade off. Anyway, new post here: http://www.reddit.com/r/leagueoflegends/comments/1zzzu3/rp_calculator_figure_out_how_to_get_to_0_rp/ It's funny because it's true. | ||
|
Eiii
United States2566 Posts
On March 10 2014 09:13 Alaric wrote: That's way higher than I'd have thought (for the wards). As people said, I'd be curious to see it cut by divisions. Do you know how's the repartition at least (eg. if there's 40% bronze games or if it's more or less balanced, for example), or nothing at all? Amusing how people on average only buy one ward but drop seven, trinket OP. nothing at all, that's something I wasn't even collecting. I'm not convinced I could get a great sample size for the higher divisions anyway, though. I could expand my current dataset to include that info, but it'd take about ten days thanks to the rate limit | ||
|
Zess
Adun Toridas!9144 Posts
At least on Reddit you get banned for making it your personal blog. | ||
|
Sufficiency
Canada23833 Posts
On March 10 2014 09:05 Eiii wrote: PS I copied the wrong table into that post, it's updated now and correct. You could do some weird machine learning stuff but this dataset is so huge I don't even wanna try. It's theoretically possible I think, though I only have a little bit of knowledge about ML! I don't think it;s machine learning. It's just statistics. I think if GEE is applied to this dataset we can find out the the synergies and counters of a particular champion (provided there is enough data). | ||
|
chalice
United States1945 Posts
i think you are unfairly judging csheep for his past behavior, rather than being objective about the present. | ||
|
killerdog
Denmark6522 Posts
On March 10 2014 07:59 Sufficiency wrote: No idea. My other post which I did things manually reached the front page. I don't see why this one is bad. But hey, at least lol reddit is moderated, where as here Cheep can just go around insult anyone at will. So it's a trade off. Anyway, new post here: http://www.reddit.com/r/leagueoflegends/comments/1zzzu3/rp_calculator_figure_out_how_to_get_to_0_rp/ You disabled being able to enter text/things like 10e99, but entering decimals still makes it die | ||
|
Sufficiency
Canada23833 Posts
On March 10 2014 10:43 killerdog wrote: You disabled being able to enter text/things like 10e99, but entering decimals still makes it die Oh god please  who is going to enter decimal lol.I think entering things like 1e4 works as intended, though. | ||
|
HazMat
United States17077 Posts
| ||
|
oneofthem
Cayman Islands24199 Posts
it could be interesting though | ||
|
Sufficiency
Canada23833 Posts
On March 10 2014 11:04 oneofthem wrote: earlier someone mentioned applying sports analytics to league. we need better data in order to do that. stuff like damage done, good fights initiated etc. some sort of accurate reflection of actual plays rather than kda and such. it could be interesting though I looked at the Riot API. It seems that you need to crawl your way to get the match history, because the API itself only provides the match history per USER... and since there is a limit to how many times you can get someone's match history, you have to space out the requests and it can potentially take months. But, like I said, one of the easiest things that can be done is synergy/counter analysis. All that is needed is the match history for the last X months to do a fairly good analysis. | ||
|
SagaZ
France3460 Posts
Nice driveby btw I like it | ||
|
Roffles
Pitcairn19291 Posts
On March 10 2014 11:04 HazMat wrote: The Maknoon that is streaming does not look like Maknoon. Did he gain a lot of weight or something? What did he do when he was retired? Looks like he ate a lotta cheeseburgers. | ||
| ||
|
|
CrankTV Team League
WardiTV Qualifier
Epic.LAN
Big Brain Bouts
SHIN vs Elazer
Percival vs Nicoract
Reynor vs Lambo
Replay Cast
RSL Revival
Clem vs Lambo
Scarlett vs Cure
CranKy Ducklings
Epic.LAN
IPSL
Dragon vs Hawk
RSL Revival
Classic vs Trap
herO vs SHIN
[ Show More ] |
|
|
EPT
