|
Czech Republic18921 Posts
On September 27 2014 07:27 Fleetfeet wrote:Show nested quote +On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: ![[image loading]](http://i.imgur.com/uOGZddJ.png) . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing.
How is 67% even close to "irrelevant"? That's a very sizeable win percentage.
|
On September 27 2014 07:29 cecek wrote:Show nested quote +On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage.
Yeah, that's a pretty strong statistic. And you can't underrate how much not adjusting to the patch affects that: you know what a team with a 5k lead at 10 minutes will usually play like. They have to give some stuff away to lose after that kind of lead. Before they could and it would just not matter as much, as evidenced by the .81 stats.
|
On September 27 2014 07:29 cecek wrote:Show nested quote +On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage.
Let's put it this way.
The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1
A) 67% of the time
B) 82% of the time
Considering what we understand of that advantage, which seems more reasonable?
Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. It seems stupid to push a team with an obvious, earned, and substantial advantage more towards losing the game, simply because it makes matches more interesting.
|
A 5K xp lead isn't a mind blowing amount either if you divide it up among 5 heroes(which it's not going to be due to it being so close to the laning phase at 10 minutes) its about a level and a half or so per hero. Its alot, but I don't believe it should be as insurmountable at this data is suggesting. If that gave you a 82% win rate in the last patch, it was one of the many reasons teams were GGing in the first 15 minutes of the game at TI4.
|
I'm still not sure why we're talking about XP leads when gold is the thing that was the thing that mattered more in 6.81 and the biggest change from one patch to the next. XP is nice and all, but those mid game deathballs were because of items and the advantage they got from having super fast meks and the like from towers.
Doubt we'll get it for a while, but still is something to think about IMO.
|
Czech Republic18921 Posts
On September 27 2014 07:36 Fleetfeet wrote:Show nested quote +On September 27 2014 07:29 cecek wrote:On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage. Let's put it this way. The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1 A) 67% of the time B) 82% of the time Considering what we understand of that advantage, which seems more reasonable? Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that.
What are you even talking about? Where did you get the team spawning at level 3 versus a team spawning at level 1?These table shows the xp difference at minute 10, when on average a team has 15k xp, so a 5k xp lead will on average look something like 17k to 12k xp. Not 5k vs 0 like your weird example, that's a completely different case.
|
There's arguments for both sides but I think the most important thing is you could get xp/gold leads without not necessarily playing "better" than your opponent. Mostly via good lanes/draft, while strategy should be important there's been to much focus on it imo.
Now even if you're very behind the chance to outskill your opponent is higher.
|
Czech Republic18921 Posts
On September 27 2014 07:41 LeLoup wrote:
I'm still not sure why we're talking about XP leads when gold is the thing that was the thing that mattered more in 6.81 and the biggest change from one patch to the next. XP is nice and all, but those mid game deathballs were because of items and the advantage they got from having super fast meks and the like from towers.
Doubt we'll get it for a while, but still is something to think about IMO.
The xp and gold formulas are exactly the same, it shouldn't be much different, I think. There are kill streak bounties for gold, but not for xp, but it still should still be very close. The guy couldn't do gold, because the dota 2 api apparently doesn't show gold graphs or something.
|
On September 27 2014 07:41 cecek wrote:Show nested quote +On September 27 2014 07:36 Fleetfeet wrote:On September 27 2014 07:29 cecek wrote:On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage. Let's put it this way. The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1 A) 67% of the time B) 82% of the time Considering what we understand of that advantage, which seems more reasonable? Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. What are you even talking about? Where did you get the team spawning at level 3 versus a team spawning at level 1?These table shows the xp difference at minute 10, when on average a team has 15k xp, so a 5k xp lead will on average look something like 17k to 12k xp. Not 5k vs 0 like your weird example, that's a completely different case.
Whoa. Apparently you didn't even understand that the example was an example. A level advantage is relevant at all stages of the game, and while a level 3 hero laning against a level 1 hero isn't quite the same impact as a team consisting entirely of level 8/9 heroes fighting a team of level 7/8 heroes, it's much easier to imagine my provided example than an actual situation, because there's less shit flying around.
|
The Dota 2 API does have gold graphs. You have to have DB+ to see them (and even then I can only see the advanced stats for my games and a few others) but they do exist afaik.
And yes, the differences are pretty big. What looks like back and forth on the gold graph is often skewed back to 0 in various ways where the gold graph tends to be more favored to one side (thanks to objectives taken which is often never accounted for in these graphs).
On September 27 2014 07:48 Fleetfeet wrote:Show nested quote +On September 27 2014 07:41 cecek wrote:On September 27 2014 07:36 Fleetfeet wrote:On September 27 2014 07:29 cecek wrote:On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage. Let's put it this way. The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1 A) 67% of the time B) 82% of the time Considering what we understand of that advantage, which seems more reasonable? Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. What are you even talking about? Where did you get the team spawning at level 3 versus a team spawning at level 1?These table shows the xp difference at minute 10, when on average a team has 15k xp, so a 5k xp lead will on average look something like 17k to 12k xp. Not 5k vs 0 like your weird example, that's a completely different case. Whoa. Apparently you didn't even understand that the example was an example. A level advantage is relevant at all stages of the game, and while a level 3 hero laning against a level 1 hero isn't quite the same impact as a team consisting entirely of level 9 heroes fighting a team of level 7/8 heroes, it's much easier to imagine my provided example than an actual situation, because there's less shit flying around.
Your example was stupid and hyperbolic. The difference between a 2500% XP lead and a 30% XP lead are significant enough to call it bullshit.
|
![[image loading]](http://i.imgur.com/08pYDjj.png)
So I was Dire and we won this game. Idk, my last three games were like this. Like stomps then huge swings in favor of the stomped.
|
Czech Republic18921 Posts
On September 27 2014 07:48 Fleetfeet wrote:Show nested quote +On September 27 2014 07:41 cecek wrote:On September 27 2014 07:36 Fleetfeet wrote:On September 27 2014 07:29 cecek wrote:On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage. Let's put it this way. The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1 A) 67% of the time B) 82% of the time Considering what we understand of that advantage, which seems more reasonable? Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. What are you even talking about? Where did you get the team spawning at level 3 versus a team spawning at level 1?These table shows the xp difference at minute 10, when on average a team has 15k xp, so a 5k xp lead will on average look something like 17k to 12k xp. Not 5k vs 0 like your weird example, that's a completely different case. Whoa. Apparently you didn't even understand that the example was an example. A level advantage is relevant at all stages of the game, and while a level 3 hero laning against a level 1 hero isn't quite the same impact as a team consisting entirely of level 9 heroes fighting a team of level 7/8 heroes, it's much easier to imagine my provided example than an actual situation, because there's less shit flying around.
It's the combination of a completely made up example and actual percentages from the table I posted that confused me. :D I still don't really know what you're saying. It is a matter of opinion and preference how "swingy" or "snowbally" anyone likes the game to be, neither is objectively better than the other. In the end it all comes down to the creative vision of the developers, doesn't it?
|
The Dota 2 API does have gold graphs. You have to have DB+ to see them (and even then I can only see the advanced stats for my games and a few others) but they do exist afaik.
The Dota2 API definitely has the data but I'm pretty sure what we see on DB+ is the result of their own parsing of said data, not a graph provided by the API (that wouldn't even make sense tbh).
@FleetFeet sorry but your example really is weak because five level 3s going against level 1s would completely overpower them unless they just don't have spells that mean anything.
|
On September 27 2014 07:55 cecek wrote:Show nested quote +On September 27 2014 07:48 Fleetfeet wrote:On September 27 2014 07:41 cecek wrote:On September 27 2014 07:36 Fleetfeet wrote:On September 27 2014 07:29 cecek wrote:On September 27 2014 07:27 Fleetfeet wrote:On September 27 2014 07:13 cecek wrote:Guy on reddit made a post where he analysed the winrate depending on how much xp ahead a team is at 10 minutes in the past versions. Basically, this table is the important part: . I definitely like the state of it now more than in 6.81. Being just 2k xp ahead and having a 67% winrate at 10 minutes seems pretty ridiculous, now it's just 56% for the same lead at 10 minutes. It is a huge change and will take time to get used to, but it is definitely positive, imo. Some kind of a middleground is probably better and I assume IceFrog will tweak the formulas a bit, because this bitching from the community is just absurd. source It's not like the bitching is unfounded. You say yourself that it'll take time to get used to. Some people are just not used to it more vocally than others. And a 5k exp lead only amounting to a 67% chance to win (even skewed as that is by people not "playing this patch right" yet) is gross, to me. It's not like that 5k exp lead wasn't earned or given in the first place, I don't understand how making it much closer to irrelevant is a good thing. How is 67% even close to "irrelevant"? That's a very sizeable win percentage. Let's put it this way. The team that spawns with all five of their heroes starting at level 3 (a 2500 experience advantage) will beat the team that starts with all of its heroes at level 1 A) 67% of the time B) 82% of the time Considering what we understand of that advantage, which seems more reasonable? Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. What are you even talking about? Where did you get the team spawning at level 3 versus a team spawning at level 1?These table shows the xp difference at minute 10, when on average a team has 15k xp, so a 5k xp lead will on average look something like 17k to 12k xp. Not 5k vs 0 like your weird example, that's a completely different case. Whoa. Apparently you didn't even understand that the example was an example. A level advantage is relevant at all stages of the game, and while a level 3 hero laning against a level 1 hero isn't quite the same impact as a team consisting entirely of level 9 heroes fighting a team of level 7/8 heroes, it's much easier to imagine my provided example than an actual situation, because there's less shit flying around. It's the combination of a completely made up example and actual percentages from the table I posted that confused me. :D I still don't really know what you're saying. It is a matter of opinion and preference how "swingy" or "snowbally" anyone likes the game to be, neither is objectively better than the other. In the end it all comes down to the creative vision of the developers, doesn't it?
On September 27 2014 07:36 Fleetfeet wrote:
Obviously neither answer is objectively better than the other. However, I feel like B is more reasonable than A, and you can't really fault me for that. It seems stupid to push a team with an obvious, earned, and substantial advantage more towards losing the game, simply because it makes matches more interesting.
Holy dickbutts it's like we're saying the same damn thing.
And yeah I can see how me using actual percentages from the graph can be confusing now, haha. I just didn't wanna throw 100% entirely made up numbers in there.
|
Czech Republic18921 Posts
|
The real problem is that late game team comps are now vastly superior to early-mid game teams. Why strategize when you can just pick Spectre/Medusa/Void, lose the early game horribly, and then win anyway?
dumb as hell imo.
|
On September 27 2014 08:01 Carwash wrote:
The real problem is that late game team comps are now vastly superior to early-mid game teams. Why strategize when you can just pick Spectre/Medusa/Void, lose the early game horribly, and then win anyway?
dumb as hell imo.
Maybe because according to the data if you lose the early game horribly you're still massively favored to lose.
|
On September 27 2014 08:02 FHDH wrote:Show nested quote +On September 27 2014 08:01 Carwash wrote:
The real problem is that late game team comps are now vastly superior to early-mid game teams. Why strategize when you can just pick Spectre/Medusa/Void, lose the early game horribly, and then win anyway?
dumb as hell imo.
Maybe because according to the data if you lose the early game horribly you're still massively favored to lose.
that's only if you pick non-late game oriented heroes. check the stats. all hard carries are WAY up in win rate and all mid game heroes are way down. hmmmmmmm wonder why that is??? i mean shit, i don't have to defend this. Spectre is at a 62% win rate lmao. It's a joke.
|
On September 27 2014 08:05 Carwash wrote:Show nested quote +On September 27 2014 08:02 FHDH wrote:On September 27 2014 08:01 Carwash wrote:
The real problem is that late game team comps are now vastly superior to early-mid game teams. Why strategize when you can just pick Spectre/Medusa/Void, lose the early game horribly, and then win anyway?
dumb as hell imo.
Maybe because according to the data if you lose the early game horribly you're still massively favored to lose. that's only if you pick non-late game oriented heroes. check the stats. all hard carries are WAY up in win rate and all mid game heroes are way down. hmmmmmmm wonder why that is??? i mean shit, i don't have to defend this. Spectre is at a 62% win rate lmao. It's a joke.
Yep this is way worse than what people were complaining about before: hard carries being made irrelevant by the 6.81 meta. Definitely not an intended indirect buff to hard carries (going along with direct buffs to them) and definitely no skew to the data caused by people thinking early leads make them nigh-invincible.
I'm sure when players stop throwing bodies at buildings, I dunno, some time after the first 36 hours of the patch they've had to adjust, it will make no difference in the win rate of hard carries. Nor will everyone knowing that carries are back and they should make sure they have one instead of trying to have the game essentially won at 10-15 while the rest of the game is a formality. None of that will happen.
Omniknight's at 64%
Hero savers have gone way up in value.
And bloodseeker isn't even that close to 50%, despite people declaring him OP.
Old Bloodseeker players: HA HA I RUN AROUND AND KILL YOUR SUPPORTS HA HA
Doesn't work the same way anymore, which is fine. Fuck those guys. Most of them aren't smart enough to adjust to this patch quickly. I think he's still stupid-strong but you have to play him very differently to get the most out of it.
|
Omniknight's at 64%
And bloodseeker isn't even that close to 50%, despite people declaring him OP.
So like... iunno.
...dazzle and venge barely moved, and Chen is down a bunch. I don't think "Hero savers have become more valuable" is a solid explanation.
|
|
|
|
|
|
EPT
