| Forum Index > LoL General |
|
CatharsisUT
United States487 Posts
On February 18 2013 13:39 De4ngus wrote: super inaccurate. its just based on teir/division and has nothing to do with mmr. Agreed. I was at 1395 before the league change, have won 2 and lost 1, and it thinks shows 1379 for me. So not that accurate. On the other hand the Elobuff calculation seems to be fairly accurate in my case. | ||
|
Zess
Adun Toridas!9144 Posts
On February 18 2013 13:49 CatharsisUT wrote: Agreed. I was at 1395 before the league change, have won 2 and lost 1, and it thinks shows 1379 for me. So not that accurate. On the other hand the Elobuff calculation seems to be fairly accurate in my case. That's unfortunate that they don't have a better estimate They had enough information to pretty much emulate true MMR. Use the guys with a ton (150+?) of games in preseason 3 as seeds with a marker for stable elo with maybe k=12 and calculate elo based on match histories. Then as you update k factors and elo, you can use the marked players as an extra check. If someone who's elo is supposed to be 1300 but played 10 games in a row with 1500 player who have marked elo as stable, then you know your system was too restrictive on the k factor compared to riot's MMR engine and can update the player in question followed by updating the emulator itself with the new information | ||
|
obesechicken13
United States10467 Posts
On February 18 2013 14:01 xes wrote: That's unfortunate that they don't have a better estimate They had enough information to pretty much emulate true MMR. Use the guys with a ton (150+?) of games in preseason 3 as seeds with a marker for stable elo with maybe k=12 and calculate elo based on match histories. Then as you update k factors and elo, you can use the marked players as an extra check. If someone who's elo is supposed to be 1300 but played 10 games in a row with 1500 player who have marked elo as stable, then you know your system was too restrictive on the k factor compared to riot's MMR engine and can update the player in question followed by updating the emulator itself with the new information If they tried to reverse engineer a formula based on previous data, then the simulator may not use the same formula as Riot's. Eventually Riot's ELO and Lolking's score might fall completely out of sync and since matchmaker uses Riot's Elo I guess this is the best they could do. I think if they're basing your ELO on tiers and divisions then the score will be accurate enough. It'll be close over the long run as no one can stay in one division while their elo drops or rises by 200 points except for the challenger people. Still, it convolutes the whole system even more than before. Riot probably hired a person to develop a better matchmaking system and this was their solution. | ||
TheYango
United States47024 Posts
| ||
|
Nos-
Canada12016 Posts
| ||
|
wei2coolman
United States60033 Posts
On February 18 2013 14:44 Nos- wrote: Not sure if you guys read the brief history through chinese dota article up but, the metas that chinese dota went through feel very much mirrored by what league went through or is going through right now. Difference is, meta changes in LoL are influenced by player base reacting to Riot's changes to LoL, as opposed to Dota, which is playerbase figuring shit out for themselves. | ||
|
cLutZ
United States19574 Posts
On February 18 2013 15:10 wei2coolman wrote: Difference is, meta changes in LoL are influenced by player base reacting to Riot's changes to LoL, as opposed to Dota, which is playerbase figuring shit out for themselves. I just read it. I dont recall Lol ever having a lategame meta. We had a couple of teams play lategame comps, and some teams exploited laneswapping before it was understood properly to get fed AD Carries, but nothing like average games being even 40+ mins. | ||
|
smOOthMayDie
United States997 Posts
| ||
|
BlackPaladin
United States9316 Posts
On February 18 2013 15:16 cLutZ wrote: I just read it. I dont recall Lol ever having a lategame meta. We had a couple of teams play lategame comps, and some teams exploited laneswapping before it was understood properly to get fed AD Carries, but nothing like average games being even 40+ mins. ? Season 1 pro games were essentially "farm for 30-40 mins then group and fight." fb being 20+ mins in was typical. | ||
|
zodde
Sweden1908 Posts
On February 18 2013 15:16 cLutZ wrote: I just read it. I dont recall Lol ever having a lategame meta. We had a couple of teams play lategame comps, and some teams exploited laneswapping before it was understood properly to get fed AD Carries, but nothing like average games being even 40+ mins. Wasn't CLG.eu's playstyle something that could be compared to a lategame meta? They picked a team that wouldn't ever lose a late game 5v5 team fight if played correctly. They lost early, but just turtled and when it came to team fights, they still won. I don't think you can compare the game lengths directly between DotA and LoL, it's better to compare relative game lengths. CLG.eu could have games where they were in the lead the whole game, but it still took 40+ minutes to close it out, where the more aggresive teams would end the games in 20-25 minutes. While they are both mobas, DotA and LoL are still vastly different games, but I still see some similarity to the history of DotA and what we're seen in LoL the past years. On what wei2coolman wrote, I think that Riot should make fewer changes to the game, just to give the pro scene a chance to develop their own styles. A more static game, in term of game changes/patches, would be good for the e-sport side of LoL. But I guess Riot cares too much about their casual playerbase? | ||
|
Gahlo
United States35172 Posts
| ||
|
zulu_nation8
China26351 Posts
| ||
|
Two_DoWn
United States13684 Posts
| ||
|
wei2coolman
United States60033 Posts
On February 18 2013 15:32 zodde wrote: Wasn't CLG.eu's playstyle something that could be compared to a lategame meta? They picked a team that wouldn't ever lose a late game 5v5 team fight if played correctly. They lost early, but just turtled and when it came to team fights, they still won. I don't think you can compare the game lengths directly between DotA and LoL, it's better to compare relative game lengths. CLG.eu could have games where they were in the lead the whole game, but it still took 40+ minutes to close it out, where the more aggresive teams would end the games in 20-25 minutes. While they are both mobas, DotA and LoL are still vastly different games, but I still see some similarity to the history of DotA and what we're seen in LoL the past years. On what wei2coolman wrote, I think that Riot should make fewer changes to the game, just to give the pro scene a chance to develop their own styles. A more static game, in term of game changes/patches, would be good for the e-sport side of LoL. But I guess Riot cares too much about their casual playerbase? I disagree about CLG.eu's playstyle being considered lategame meta. Though they always stalled to lategame, they didn't have good lategame strategy. Essentially, they're only lategame power was their teamfighting ability, no actual lategame powertiming/objective control. | ||
|
TheYango
United States47024 Posts
On February 18 2013 15:32 zodde wrote: On what wei2coolman wrote, I think that Riot should make fewer changes to the game, just to give the pro scene a chance to develop their own styles. A more static game, in term of game changes/patches, would be good for the e-sport side of LoL. But I guess Riot cares too much about their casual playerbase? Does anyone really believe "the casual playerbase" will quit over a longer patch cycle? It seems like a really weak argument to me. Hell, you could make the argument that a longer patch cycle would even drive up RP sales as players might be more willing to invest in champs/skins for a popular champ without fear of that champ being nerfed. | ||
|
zodde
Sweden1908 Posts
On February 18 2013 16:20 wei2coolman wrote: I disagree about CLG.eu's playstyle being considered lategame meta. Though they always stalled to lategame, they didn't have good lategame strategy. Essentially, they're only lategame power was their teamfighting ability, no actual lategame powertiming/objective control. Yeah I guess you're right. My understanding of DotA, and what lategame meta even mean in DotA is very limited. On February 18 2013 16:21 TheYango wrote: Does anyone really believe "the casual playerbase" will quit over a longer patch cycle? It seems like a really weak argument to me. Hell, you could make the argument that a longer patch cycle would even drive up RP sales as players might be more willing to invest in champs/skins for a popular champ without fear of that champ being nerfed. People (casuals) whine so much on the forums about "OP champs", that would certainly increase if a really "OP" champion wasn't patched in several months. | ||
|
Chrispy
Canada5878 Posts
Seriously. | ||
|
HazMat
United States17077 Posts
On February 18 2013 16:29 Chrispy wrote: Why are there so many goofballs out there not banning Nidalee 100% of the time? Seriously. Plz use the appropriate thread for ur rage ty. | ||
|
cLutZ
United States19574 Posts
Those are lategame Strategies employed by A team/in an instance. What was described, at least as far as I could tell was a meta that would look like this in LOL: Nunu=Must ban status. Vayne and Kog Maw are highly contested. Midlane Consists on Anivia vs. Karthus vs. Cassio vs. Caitlyn/Corki. Toplane is a Jax/Irelia festival with characters like Kayle Sometimes making an appearance. Then you attempt to max out your items and ward everywhere before doing much at all. EHome simply controlled the map and maxed out three heroes before even attempting to go onto LGD's high ground. Its like S2 CLG EU playing itself, | ||
|
Nos-
Canada12016 Posts
| ||
| ||
StarCraft: Brood War Dota 2 League of Legends Counter-Strike Other Games |
|
Escore
The PondCast
WardiTV Invitational
Zoun vs Ryung
Lambo vs ShoWTimE
Big Brain Bouts
Fjant vs Bly
Serral vs Shameless
OSC
Replay Cast
CranKy Ducklings
RSL Revival
SHIN vs Bunny
ByuN vs Shameless
WardiTV Invitational
Krystianer vs TriGGeR
Cure vs Rogue
uThermal 2v2 Circuit
[ Show More ] |
|
|
EPT
