I'm just curious as to why HSReplay data is considered ''the meta" when it only shows a small % of total games played in the actual meta.
I know it's the only data people have access too, but to use it to state things like "x card isn't used in the meta" because it isn't on HSReplay, is a bit wrong?
And isn't the data from HSReplay inherently inaccurate anyway? source
All I can say is once I started making decisions based on HSReplay and not VS, I was finally able to at least achieve legend rank. I don't think one can say what HS presents is the definitive picture of what is happening out there but since the data is being gathered from actual games it does at least give you some indication of which match ups to target and what you can expect.
HS Replay data is biased. Bias is a statistical term. The data collected from a game between opponents, one of which is using HS Replay, one is not, is not symmetrical. And it's reasonable to assume that those two players might have different skill levels for multiple reasons.
However, the basis of statistics is that we can infer some information from a sample that is a small fraction of the population. I wouldn't worry so much about the sample size, but the bias almost certainly leads to inflated win rates in the specific deck section. They use some kind of math magic (Bayesian statistics possibly) to smooth out the win rates between different archetypes.
All I can say is once I started making decisions based on HSReplay and not VS, I was finally able to at least achieve legend rank. I don't think one can say what HS presents is the definitive picture of what is happening out there but since the data is being gathered from actual games it does at least give you some indication of which match ups to target and what you can expect.
Interesting topic. Any knows a database that collects the same? Like im testing many Off-Meta cards now and im Interesting about the statistics. Or this data is blizz only lol?
All I can say is once I started making decisions based on HSReplay and not VS, I was finally able to at least achieve legend rank. I don't think one can say what HS presents is the definitive picture of what is happening out there but since the data is being gathered from actual games it does at least give you some indication of which match ups to target and what you can expect.
and you think VS data is not from actual games ??
No, I understand it is, I didn't mean VS is not gathered from stats trackers. I just find that the VS report isn't helpful to me, but HS filtering and match up info has been useful.
If u blindly trust 1 source of information u will have a hard time on the internet. In hearthstone use checks stats and decklist from many places and combine others ideas with what u have learned yourself. 1 card difference is huge in alot of decks.
If u arent using a tracker u are not taking this game seriously, i look at my own stats alot more then i look at others.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
I'm just curious as to why HSReplay data is considered ''the meta" when it only shows a small % of total games played in the actual meta.
I know it's the only data people have access too, but to use it to state things like "x card isn't used in the meta" because it isn't on HSReplay, is a bit wrong?
And isn't the data from HSReplay inherently inaccurate anyway? source
All I can say is once I started making decisions based on HSReplay and not VS, I was finally able to at least achieve legend rank. I don't think one can say what HS presents is the definitive picture of what is happening out there but since the data is being gathered from actual games it does at least give you some indication of which match ups to target and what you can expect.
HS Replay data is biased. Bias is a statistical term. The data collected from a game between opponents, one of which is using HS Replay, one is not, is not symmetrical. And it's reasonable to assume that those two players might have different skill levels for multiple reasons.
However, the basis of statistics is that we can infer some information from a sample that is a small fraction of the population. I wouldn't worry so much about the sample size, but the bias almost certainly leads to inflated win rates in the specific deck section. They use some kind of math magic (Bayesian statistics possibly) to smooth out the win rates between different archetypes.
and you think VS data is not from actual games ??
Interesting topic. Any knows a database that collects the same? Like im testing many Off-Meta cards now and im Interesting about the statistics. Or this data is blizz only lol?
No, I understand it is, I didn't mean VS is not gathered from stats trackers. I just find that the VS report isn't helpful to me, but HS filtering and match up info has been useful.
If u blindly trust 1 source of information u will have a hard time on the internet. In hearthstone use checks stats and decklist from many places and combine others ideas with what u have learned yourself. 1 card difference is huge in alot of decks.
If u arent using a tracker u are not taking this game seriously, i look at my own stats alot more then i look at others.