HS Replay statistics can't really be taken too literally. Just ask all the people who thought Rafaam was good in Zoo.
Rafaam never had a good "played" winrate in zoo, but the decks with him still did better than the ones without because he would turn some certain losses into wins. He is also a meta call.
I don't even look at the Played Win Rate column. Leeroy must have one of the highest played win rates in HS, but it's not that good. Etc.
But more to point, a bug in HSR stats led people to believe that Rafaam was a keep in the mulligan vs Warrior. How did they screw up this bad? They explain it in a recent article on the site. But you probably know all this.
There is a difference between data and information. Statistics allows one to transform data into information, but there are a minimal set of requirements that have to be followed in order to do this. The data on HSR does not adhere to these requirements. And this is assuming that it is being calculated correctly.
I did not know about that bug, I will check it out.
Working with the crafting guide, I have spent a substantial amount of time turning HSreplay number into crafting advice. The most common "bug" is that cards get carried to a high deck winrate by a strong archetype they can be used in while actually making that archetype worse. I think comparing winrates of decks with and without a certain card can sort most of that out, which is what I did trying to evaluate Rafaam.
The most problematic using hsreplay for choosing a deck is that your opponents at higher levels will do the same and counter you. Pretty often, there is a lag which makes the best decks weaker in the real world. As for "played" winrates they often give useful information, especially for lategame cards that are not combo pieces or finishers (like leeroy) and tech cards (which usually perform worse than expected.)
Control warrior is a very strong deck atm, 65% winrate is very high for a popular deck, even just after a release.
for me https://www.hearthpwn.com/decks/1295144-gloops-token-druid-legend-450
token druid both before and after SoU, the new cards was generous on vanilla token druid!
before SoU 40-50 after 450
proof
I would hope you would be able to et 12 wins in brawlesium with that contro lwarrior deck.
look how many legendaries in it and the dust cost of that deck. You're average player isn't gonna be able to make those kinds of decks.
and some people say hearthstone isn't a p2w game....
I did not know about that bug, I will check it out.
Working with the crafting guide, I have spent a substantial amount of time turning HSreplay number into crafting advice. The most common "bug" is that cards get carried to a high deck winrate by a strong archetype they can be used in while actually making that archetype worse. I think comparing winrates of decks with and without a certain card can sort most of that out, which is what I did trying to evaluate Rafaam.
The most problematic using hsreplay for choosing a deck is that your opponents at higher levels will do the same and counter you. Pretty often, there is a lag which makes the best decks weaker in the real world. As for "played" winrates they often give useful information, especially for lategame cards that are not combo pieces or finishers (like leeroy) and tech cards (which usually perform worse than expected.)
Control warrior is a very strong deck atm, 65% winrate is very high for a popular deck, even just after a release.
Editor of the Heartpwn Legendary Crafting Guide:
https://www.hearthpwn.com/forums/hearthstone-general/card-discussion/205920-legendary-tier-list-crafting-guide