Whenever there is a discussion on a hero/build/talent, there are always people that like to point at (usually unfiltered) data on https://www.heroesprofile.com/ and then assume the "argument high-ground", because their position is "supported by actual data", then get upvoted to the heavens. Even if the data difference is 1-2%, which can quite often be below the statistical margin of error.
On the other hand, whenever there is a discussion about playerbase skill and knowledge level, the commonly stated idea is that "it's terrible".
But aren't those two kind of contradictory, when looking at pick rates? Referring to Heroes Profile's pick rates is basically depending on the majority player base's opinion. There is also the "herd effect", which is if everyone is looking at the Heroes Profile pick rates, then pick talents that have the highest pick rates, it will result certain talents having abnormally high pick rates in relative to how much value they provide. Similarly, online guides on Youtube/Icyveins/Reddit can also have a significant effect on the pick rates, even if said guides are years old.
Win rates are also quite often misleading as single talents can have higher/lower win rates not just because of the raw value they provide. Defensive talents seem to have a lower pick and win rate, since they're usually picked in losing situations when the player is getting deleted too quickly in teamfights and trying to stay alive (if there was a way to filter to those situations, probably then these talents would have by far the highest win rate with that filter). High skill-floor talents also usually have a lower pick rate and win rate, however those that master the hero will quite often find that they get the most value out of them in comparison to other options.
I think a lot of problems could be resolved if people started using the filters, especially Hero Level to filter to players who have the hero level above "X" (e.g. 10+), to show how people that are experienced with a hero play it. This removes a lot of the herd effect on the pick-rates and the low-experience effect on the win-rates. Combine this with also filtering to higher ranks (e.g. Platinum+) with the Player Rank filter and you can get a lot more accurate data on "How valuable is talent/hero X when properly utilized?". Sometimes, because there isn't enough data (<200 for lowest pick talent), for a lot of heroes this can make the values unusable, so for those, you need to drastically increase the Timeframe (add more patches). If there isn't enough patches to add without including a patch that changes the hero, then the data simply isn't reliable enough to back claims.
So what I'm saying is that Heroes Profile without filters is useless? Oh hell no. It's a great tool to show how you can get the most value if you're new as it shows you the safe options for people with low experience. If you're new to a role, pick the hero with the highest winrate among those that have 20%+ popularity (e.g. Rehgar for Healer or E.T.C. for Tank). If you're new to a hero, pick the talent build with the highest pickrate (e.g. for Rehgar, Lighting Shield build with Cleanse and Ancestral).
TL.DR.: Use the Hero Level and Player Rank filters on Heroes Profile to get a more accurate "value of talent/hero" data. If there isn't enough data, significantly increase the Timeframe. If you can't without including patches that changed the hero(es) in discussion, then there isn't enough data to back claims.
Source: Original link
© Post "Referring to Heroes Profile data to validate arguments is way too often abused on this sub" for game Heroes of the Storm.
Top 10 Most Anticipated Video Games of 2020
2020 will have something to satisfy classic and modern gamers alike. To be eligible for the list, the game must be confirmed for 2020, or there should be good reason to expect its release in that year. Therefore, upcoming games with a mere announcement and no discernible release date will not be included.
Top 15 NEW Games of 2020 [FIRST HALF]
2020 has a ton to look forward to...in the video gaming world. Here are fifteen games we're looking forward to in the first half of 2020.