|
|
On August 24 2012 01:53 Canas wrote:
Also what happened with TLIH, I was just feeling like putting my tryhard hat on and the thread's not stickied anymore? Did it just die from inactivity or something?
pretty much what happened, im counting on a surge of popularity once everyone with a steam account can play.
I dont think it will ever be as high quality as it was before though, there will be many more noobs like me when its active again i bet :p.
|
On August 24 2012 01:36 Kipsate wrote:
Can we not have LoL vs Dota arguments here, it shits up the thread/forum. Keep LoL out of this please, interested to hear results anyway.
How is it LoL vs DotA when someone showcases a method someone used on one game and recommends using it for another one? Like, I'm aware that people hate admitting when someone else is doing something better or worse but this isn't even about that. It's about a damn method trying to analyze a game. You can do exactly the same for any game with resources out there. -_-
To get back on topic, a huge problem with this "method" is that it probably confuses correlation with causation. You need way more data to remove the skill gap between teams. I'm pretty sure it's more like that the better team had the lead at a certain time and THAT'S why they also won. Not because it's impossible to come back, but because the better team rarely gets into the position where they're behind in the first place.
|
Netherlands45349 Posts
On August 24 2012 02:01 r.Evo wrote:Show nested quote +On August 24 2012 01:36 Kipsate wrote:
Can we not have LoL vs Dota arguments here, it shits up the thread/forum. Keep LoL out of this please, interested to hear results anyway. How is it LoL vs DotA when someone showcases a method someone used on one game and recommends using it for another one? Like, I'm aware that people hate admitting when someone else is doing something better or worse but this isn't even about that. It's about a damn method trying to analyze a game. You can do exactly the same for any game with resources out there. -_- To get back on topic, a huge problem with this "method" is that it probably confuses correlation with causation. You need way more data to remove the skill gap between teams. I'm pretty sure it's more like that the better team had the lead at a certain time and THAT'S why they also won. Not because it's impossible to come back, but because the better team rarely gets into the position where they're behind in the first place.
things like Highlighting that comebacks are much more prevelent and capable in Dota than LoL. are pretty subtle(or nonsubtle) jabs at LoL
but either way let it go, no point in arguing about it.
|
It's because dota, unlike lol, is a game with dynamic asymmetry in which hero roles and effectiveness, as well as farm allocation and effectiveness, greatly different from game to game and points between a game. The "state" of a dota game cannot be summarized by the gold count the way lol can be, because there's a far larger variety in hero lineups and their effectiveness. A 4-1 antimage lineup can be down 25% gold by 20 minutes to an all-in push lineup and people who know things about DotA might say they're winning, which is almost never the case in lol.
In general, lol meta and heroes are picked in such a way that 2 teams "match up" in heroes of essentially equal strength and scaling with symmetrical lanes, which makes the "snowball effect" far more prevalent because if 2 heroes function and scale essentially the same way, any side with even a slight gold lead will have the advantage in both lane and engagements.
Basically, what the data really shows isn't that comebacks are more prevalent in either game, but that the game state variance is different and less accurately represented by a single number
|
On August 24 2012 02:07 Kipsate wrote:Show nested quote +On August 24 2012 02:01 r.Evo wrote:On August 24 2012 01:36 Kipsate wrote:
Can we not have LoL vs Dota arguments here, it shits up the thread/forum. Keep LoL out of this please, interested to hear results anyway. How is it LoL vs DotA when someone showcases a method someone used on one game and recommends using it for another one? Like, I'm aware that people hate admitting when someone else is doing something better or worse but this isn't even about that. It's about a damn method trying to analyze a game. You can do exactly the same for any game with resources out there. -_- To get back on topic, a huge problem with this "method" is that it probably confuses correlation with causation. You need way more data to remove the skill gap between teams. I'm pretty sure it's more like that the better team had the lead at a certain time and THAT'S why they also won. Not because it's impossible to come back, but because the better team rarely gets into the position where they're behind in the first place. things like Show nested quote +Highlighting that comebacks are much more prevelent and capable in Dota than LoL. are pretty subtle(or nonsubtle) jabs at LoL but either way let it go, no point in arguing about it.
or it's just the conclusion of the limited data using the same method for both games. I didn't know data had ulterior motives and agendas towards certain games. How would you suggest one describes a 26% comeback percentage vs a 6%? I see no issue in saying that the preliminary data using the above method show combacks happen more often in Dota, because, well, that's exactly what the data shows.
Jesus, people are so sensitive about LoL.
|
On August 24 2012 02:15 crms wrote:Show nested quote +On August 24 2012 02:07 Kipsate wrote:On August 24 2012 02:01 r.Evo wrote:On August 24 2012 01:36 Kipsate wrote:
Can we not have LoL vs Dota arguments here, it shits up the thread/forum. Keep LoL out of this please, interested to hear results anyway. How is it LoL vs DotA when someone showcases a method someone used on one game and recommends using it for another one? Like, I'm aware that people hate admitting when someone else is doing something better or worse but this isn't even about that. It's about a damn method trying to analyze a game. You can do exactly the same for any game with resources out there. -_- To get back on topic, a huge problem with this "method" is that it probably confuses correlation with causation. You need way more data to remove the skill gap between teams. I'm pretty sure it's more like that the better team had the lead at a certain time and THAT'S why they also won. Not because it's impossible to come back, but because the better team rarely gets into the position where they're behind in the first place. things like Highlighting that comebacks are much more prevelent and capable in Dota than LoL. are pretty subtle(or nonsubtle) jabs at LoL but either way let it go, no point in arguing about it. or it's just the conclusion of the limited data using the same method for both games. I didn't know data had ulterior motives and agendas towards certain games. How would you suggest one describes a 26% comeback percentage vs a 6%? I see no issue in saying that the preliminary data using the above method show combacks happen more often in Dota, because, well, that's exactly what the data shows. Jesus, people are so sensitive about LoL.
People are terrified of it more like.
|
On August 24 2012 02:10 Kupon3ss wrote:
It's because dota, unlike lol, is a game with dynamic asymmetry in which hero roles and effectiveness, as well as farm allocation and effectiveness, greatly different from game to game and points between a game. The "state" of a dota game cannot be summarized by the gold count the way lol can be, because there's a far larger variety in hero lineups and their effectiveness. A 4-1 antimage lineup can be down 25% gold by 20 minutes to an all-in push lineup and people who know things about DotA might say they're winning, which is almost never the case in lol.
In general, lol meta and heroes are picked in such a way that 2 teams "match up" in heroes of essentially equal strength and scaling with symmetrical lanes, which makes the "snowball effect" far more prevalent because if 2 heroes function and scale essentially the same way, any side with even a slight gold lead will have the advantage in both lane and engagements.
You're absolutely right about the hero lineups. It bothers me to no end when dota casters are talking about how far behind X team is because they're down 6k gold, meanwhile team Y is all early game pusher/gankers etc., and team X has a big hard carry and have successfully defended their T2/T3 towers. We all know what's going to happen if team Y doesn't rax within 20-25minutes, but casters just explode over the gold graphs and pay no attention to anything else on the field.
|
Casters will get better over time, no worries.
|
Northern Ireland22213 Posts
Agreed, there was no hidden meaning behind what he said Kipsate :S
|
The thing is that the study fails because he assumes a 10% gold advantage at 12min means you are equally ahead in LoL and in Dota, when it simply is not true. In Dota you may even be behind, depending you your strategy and team composition. Plus 12mins in LoL may be a diferent stage of the game than in Dota, but I have no idea how gold balance really works in LoL. Even the distribution of gold between your heroes in Dota is a big factor to consider. 10% is just not that much at that stage of the game.
In the end, it only says what everyone should already know and only serves to bring useless discussion between Dota and LoL. Of course a gold advantage in LoL means more than in Dota, since teams are ussualy more "balanced" while in Dota you can have lategame strats, early game strats, try to pool money on your carry, etc. That's not even a positive side of the game, since many LoL players will defend the scaling of all heroes and other things that make the game more symmetrical. It's more a matter of preference with upsides and downsides. Of course it will look bad if you focus on the downsides.
You could easily say what thestudy proved was that being 10% ahead in gold in LoL 12mins is a bigger advantage than than the same situation in Dota. But that conclusion would be meaningless and would generate a lot less views. Would people accept the study if it used 15% in Dota and 10% in LoL? Probally not, but perhaps being 15% ahead in Dota means, on average, you have the same advantage a team in LoL has with 10%. It's really hard to quantify these things, or even define what "advantage" means.
|
Well stated criticism of motbob's data.
While LoL and Dota are remarkably similar they are also vastly different. I'm not sure what an analogous situation would be in DOTA to a 10% gold lead in LoL. Perhaps 1 set of rax down by X time or something. It'd be very hard to figure something like this out.
The comeback potential in LoL vs DOTA has always been debated and both sides have reasons for believing why it's better/worse in each game. I think this information is nice and offers severeal inferences you could draw from it even if it doesnt' necessarily solve any sort of fanboy debate.
|
On August 24 2012 02:57 crms wrote:
Well stated criticism of motbob's data.
While LoL and Dota are remarkably similar they are also vastly different. I'm not sure what an analogous situation would be in DOTA to a 10% gold lead in LoL. Perhaps 1 set of rax down by X time or something. It'd be very hard to figure something like this out.
The comeback potential in LoL vs DOTA has always been debated and both sides have reasons for believing why it's better/worse in each game. I think this information is nice and offers severeal inferences you could draw from it even if it doesnt' necessarily solve any sort of fanboy debate.
I don't believe it's useless, I'm just again't saying "LoL had 10%, Dota had 26%, that means Dota has 2.5 times more comebacks than LoL." It's just not that simple. You could have a good discussion about this, but then I don't think general discussion is the right place, it would need a lot of talk and would just clutter the thread. Here, people are just going to look at the results, comment without really thinking and end up with the same old LoL vs Dota discussion that helps noone. Perhaps a thread about it could be created, but again I'm not sure anything new would come out of it. It ussually ends up being the same shitty kind of thread we have seen plenty of times. It's like discussing religion, sometimes it's just better to ban it alltogether and be safe.
Edit: Someone should do the same with experience, if anything just to compare the results and see if anything funny comes out of it.
|
stole 46int in a 40~min game. yolo
|
Brazil1429 Posts
I'm dealing with data all the time because of the site and I watch a fuckton of games since Dota 1, and I can say that gold difference isn't decisive at all in that stage, unless we're talking of a really massive-gigantic difference. I saw soo many situations where the gold lead was artificial and not telling you the truth of the position. A considerable gold difference matters nothing if you can lose everything in a teamfight. The thing to look is to discover when the advantage is 'definitive' and solid, instead of illusory
|
Ok so I would seriously appreciate if someone could watch this game I just played and point out what I did wrong, if anything.
The match ID is 35075978. I was playing suicide lane as puck, and I was still level 1 by 6 minutes. Should I even have bothered staying bot? Should I just have gone mid to help DK instead? Was there something I could've done better, because it really didn't feel like it? The one thing I can think of off the top of my head was level 1 phase shift instead of illusory orb to help trade harassment with rubick, but I'm not even sure about that.
There was quite a bit of flaming in teamchat telling me I lost the game and such, but it REALLY didn't feel like that was why we were losing, Am I wrong?
|
Didnt watch the replay but one thing you can do when going hard lane and having no chance to get xp is using your Orb to blink into the forest and just stay hidden leeching xp.
|
Netherlands45349 Posts
watched the replay a bit, nothing you can do except double jungle with axe(which in pubs theyl go fucking nuts over)
Your Dragon Knight had no clue how to lane mid and top had no clue how to pull so you didn't get a good tradeoff, apart from that you guys were not only outpicked but in general they were better players.
|
Regarding the "LoL is snowbally" argument: i think it's not caused by the snowball effest as much as by the fact that IMO in LoL there is higher skill difference between teams than in DOTA (in DOTA there are more teams with similar skill levels than in LoL, in which there are gaps in skill levels of maany teams). When someone who knows LoL scene sees the game between 2 teams coming he will predict the winner with no problem. In DOTA you see way more upsets than in LoL and honestly even though i'm watching DOTA2 competetive scene since like 9 months (+TI1) i still get surprised quite often by the outcomes of some games.
In other words - currently in LoL it's not about the "team X had gold lead so they won", it's about the "team X was better so they had gold lead and then they won".
|
On August 24 2012 04:16 Kipsate wrote:
watched the replay a bit, nothing you can do except double jungle with axe(which in pubs theyl go fucking nuts over)
Your Dragon Knight had no clue how to lane mid and top had no clue how to pull so you didn't get a good tradeoff, apart from that you guys were not only outpicked but in general they were better players.
Yeah that's what I figured, thanks. Guess I should stop suicide laning if I'm in a game with random people because 99% of the time it's impossible to carry a game from that position.
Only thing is that typically baddie pubs who suicide lane end up like 0/8 10 mins into the game, so idk.
|
Hm I randomed a lycan last game and I usually never play him. MatchID 35065993. What could I have done better to turn the tide? I found it very difficult to push towers vs a tinker, veno and furion and my team yelled at me why I'm not pushing.
Which decisions should I have made to turn the pushing tide of the radiant? :p
|
|
|
|
|
|
EPT
