"Balanced"

General Discussion
Hero Winrate

Chen 42.6%
Tassadar 43.3%
Malthael 43.6%
Hanzo 44.6%
D.Va 44.7%
Medivh 44.9%
Garrosh 45.2%
Greymane 45.4%
Malfurion 45.7%
Probius 46.0%
Deckard 46.1%
Kel'Thuzad 46.2%
Junkrat 46.3%
Genji 46.6%
Gul'dan 46.7%
Blaze 46.9%
Abathur 47.0%
Uther 47.0%
Li-Ming 47.0%
Yrel 47.1%
The Lost Vikings 47.1%
Valeera 47.2%
Illidan 47.4%
Stitches 47.4%
Maiev 47.7%
Lt. Morales 47.9%
[quote]
09/30/2018 03:49 PMPosted by CrazyLegs
Hero Winrate

Chen 42.6%
Tassadar 43.3%
Malthael 43.6%
Hanzo 44.6%
D.Va 44.7%
Medivh 44.9%
Garrosh 45.2%
Greymane 45.4%
Malfurion 45.7%
Probius 46.0%
Deckard 46.1%
Kel'Thuzad 46.2%
Junkrat 46.3%
Genji 46.6%
Gul'dan 46.7%
Blaze 46.9%
Abathur 47.0%
Uther 47.0%
Li-Ming 47.0%
Yrel 47.1%
The Lost Vikings 47.1%
Valeera 47.2%
Illidan 47.4%
Stitches 47.4%
Maiev 47.7%
Lt. Morales 47.9%


Anyone can go to hotslogs and copy and paste a bunch of win rates. This data has nothing to do with balance. And the data from hots logs is inconsistent. This is due to the fact that many people either 1. don't upload or 2. the site simply wont let them upload do to their silence policy. And please and I mean PLEASE tell me that hots logs is a credible site because I would LOVE to argue why it isn't.
This data has nothing to do with balance.


Lol, okay. Brilliant take.
09/30/2018 03:59 PMPosted by CrazyLegs
This data has nothing to do with balance.


Lol, okay. Brilliant take.


Oh YOU DID! NICE!!!!

Ok here were go. This data you got from hots logs which I'm sure you got from the front page. Not only doesn't cover certain leagues it doesn't even cover the entire player base in the game. So you are getting bad data that doesn't represent the entire population. Now this might be hard for you to understand I'm sure. But since it DOESN'T core the entire player base, it doesn't represent changes that need to occur to specific heroes. Like your little Lucio problem you posted earlier. So these win-percents you posted are not credible because it doesn't reflect the entire playerbase.
09/30/2018 03:59 PMPosted by CrazyLegs
...

Lol, okay. Brilliant take.


Oh YOU DID! NICE!!!!

Ok here were go. This data you got from hots logs which I'm sure you got from the front page. Not only doesn't cover certain leagues it doesn't even cover the entire player base in the game. So you are getting bad data that doesn't represent the entire population. Now this might be hard for you to understand I'm sure. But since it DOESN'T core the entire player base, it doesn't represent changes that need to occur to specific heroes. Like your little Lucio problem you posted earlier. So these win-percents you posted are not credible because it doesn't reflect the entire playerbase.


In statistics, you just need a large enough sample, not all the numbers.

You are right in that those numbers don't actually matter though. Balance in this game is top down, since higher skilled players have more consistent game play. There are exceptions, but for the most part, the vast majority of games aren't applicable to "balance" issues.
09/30/2018 03:59 PMPosted by CrazyLegs
This data has nothing to do with balance.


Lol, okay. Brilliant take.


Anything else about non-credible data from a biased site you want to talk about crazylegs? I'm all for it.
[quote]<span class="truncated">...</span>

Oh YOU DID! NICE!!!!

Ok here were go. This data you got from hots logs which I'm sure you got from the front page. Not only doesn't cover certain leagues it doesn't even cover the entire player base in the game. So you are getting bad data that doesn't represent the entire population. Now this might be hard for you to understand I'm sure. But since it DOESN'T core the entire player base, it doesn't represent changes that need to occur to specific heroes. Like your little Lucio problem you posted earlier. So these win-percents you posted are not credible because it doesn't reflect the entire playerbase.


In statistics, you just need a large enough sample, not all the numbers.

You are right in that those numbers don't actually matter though. Balance in this game is top down, since higher skilled players have more consistent game play. There are exceptions, but for the most part, the vast majority of games aren't applicable to "balance" issues.


Going to just repost this from my other form to let you understand why this isn't credible.

The samples in hotslogs contains highly biased data. It doesn't count toward accounts that have been silenced or those who don't upload. It's highly inaccurate for those reasons alone. And since it DOESNT contain ALL of the playerbase, it's not relavent because only people that do upload are the ones being represented. If I just did a study on those who uploaded I would have inconsistent information about the representative population which is the data that the original poster blindly copied and pasted.
09/30/2018 04:13 PMPosted by Firall
You are right in that those numbers don't actually matter though.


I did not say numbers didn't matter. I said the hots logs data that the original poster copied and pasted didn't matter. Numbers do matter in statistics. You just said you need a large enough sample. You do know you get a sample from a population right? And the entire population isn't in hotslogs it's just those who UPLOAD the game i.e., selective games that they win for example. So using this data to correlate it to balance changes would be incredibly insane. I'm merely remarking upon the fact that the original poster is trying to spread false information to the public and needs to be addressed.
09/30/2018 04:13 PMPosted by Firall
You are right in that those numbers don't actually matter though.


I did not say numbers didn't matter. I said the hots logs data that the original poster copied and pasted didn't matter. Numbers do matter in statistics. You just said you need a large enough sample. You do know you get a sample from a population right? And the entire population isn't in hotslogs it's just those who UPLOAD the game i.e., selective games that they win for example. So using this data to correlate it to balance changes would be insane. I'm merely remarking upon the fact that the original poster is trying to spread false information to the public and needs to be addressed.


Those numbers.

Meaning hotslogs.

Again, statistics just need a large enough sample. Unless people are saying, I'm only going to upload games where X hero loses, and it's a significant amount of the people uploading, then these numbers should be relatively close. In fact a while back blizzard released internal numbers on healer win rates, and all but on or two of them were within 1-5% hotslogs stats for the same time period IIRC.

If you're using it to compare your skill with others, especially specific people, that's where the type of selective uploading you're talking about is going to skew things.
[quote="207681582381"][quote]<span class="truncated">...</span>

Those numbers.

Meaning hotslogs.

Again, statistics just need a large enough sample. Unless people are saying, I'm only going to upload games where X hero loses, and it's a significant amount of the people uploading, then these numbers should be relatively close. In fact a while back blizzard released internal numbers on healer win rates, and all but on or two of them were within 1-5% hotslogs stats for the same time period IIRC.

If you're using it to compare your skill with others, especially specific people, that's where the type of selective uploading you're talking about is going to skew things.[quote]

From an unbiased population. This is entirely voluntary response. You just said it was voluntary therefore your argument is invalid.
09/30/2018 03:49 PMPosted by CrazyLegs
Hero Winrate

Chen 42.6%
Tassadar 43.3%
Malthael 43.6%
Hanzo 44.6%
D.Va 44.7%
Medivh 44.9%
Garrosh 45.2%
Greymane 45.4%
Malfurion 45.7%
Probius 46.0%
Deckard 46.1%
Kel'Thuzad 46.2%
Junkrat 46.3%
Genji 46.6%
Gul'dan 46.7%
Blaze 46.9%
Abathur 47.0%
Uther 47.0%
Li-Ming 47.0%
Yrel 47.1%
The Lost Vikings 47.1%
Valeera 47.2%
Illidan 47.4%
Stitches 47.4%
Maiev 47.7%
Lt. Morales 47.9%


Its not possible to have 50% every where .
Players got different playing styles and players arent all the same caliber of skill.
It just shows out a portion of people how many of them play it well enough to win .
The statistics from hotslogs are valid. The sample size is more than big enough to draw statistics. However, the question is, what does winrate what to do with anything? Balance? Sure, but it doesn't really tell the full story.

Personally, what Riot does for LOL is very accurate. It's not just winrate, but pick rate and ban rate that must be taken into account. If hero has a low winrate, is it because he is hard to master? Do players who play the champion (main them) start to get closer to the 50% as they start to master it?

Based on the information we have, there are quite a few champion that can be considered underpowered. Sure, we can't say with absolute certainty but the data that we have available (not just winrate) suggest that they are slight weaker.

Anyone suggesting that the sample size is not good enough is nuts. You don't just try to take the stats from 1% of the top player base. You need to balance out stuff that are bad in the the lower tiers of the player base as well. No, it doesn't mean it will become OP in the top tier. You can min/max heroes balance. It's important to flush out the stuff that the lower tier players are also suffering from.

How do i know this? Riot, unlikely Blizzard actually does give a lot of this information out. They have done minor changes to a champion which increase the winrate/pickrate at the bottom tier (low ranked players) which keep the numbers better at the top tier.
09/30/2018 05:51 PMPosted by MCHammer
The statistics from hotslogs are valid. The sample size is more than big enough to draw statistics. However, the question is, what does winrate what to do with anything? Balance? Sure, but it doesn't really tell the full story.

Personally, what Riot does for LOL is very accurate. It's not just winrate, but pick rate and ban rate that must be taken into account. If hero has a low winrate, is it because he is hard to master? Do players who play the champion (main them) start to get closer to the 50% as they start to master it?

Based on the information we have, there are quite a few champion that can be considered underpowered. Sure, we can't say with absolute certainty but the data that we have available (not just winrate) suggest that they are slight weaker.

Anyone suggesting that the sample size is not good enough is nuts. You don't just try to take the stats from 1% of the top player base. You need to balance out stuff that are bad in the the lower tiers of the player base as well. No, it doesn't mean it will become OP in the top tier. You can min/max heroes balance. It's important to flush out the stuff that the lower tier players are also suffering from.

How do i know this? Riot, unlikely Blizzard actually does give a lot of this information out. They have done minor changes to a champion which increase the winrate/pickrate at the bottom tier (low ranked players) which keep the numbers better at the top tier.


Again, its biased and also voluntary so its not credible data. And I'm not talking about Riot or league I'm specifically talking about hots logs data and how inaccurate it is.
The statistics from hotslogs are valid.


They aren't not even remotely valid.
@MCHammer

Don't respond to Arms31312. He's a troll who uses sophistry to endlessly argue with people. Just report him and move on.
09/30/2018 06:21 PMPosted by CrazyLegs
@MCHammer

Don't respond to Arms31312. He's a troll who uses sophistry to endlessly argue with people. Just report him and move on.


I'm sorry that you are taking things personally. But you posted false information on this post and it needs to be addressed.
Quick match? Hero League? Bronze winrates? Or Master? Last week or the whole history?

SO MANY QUESTIONS.

Join the Conversation

Return to Forum