by JR Cook - 8 years ago     show comments

Anytime a new Hero is release, it seems that much of the conversation about them centers around how low their win rate is. This trend has continued with the most recent Hero, Greymane. Most of the Heroes released over the last few months have ended up with a win rate in the upper 30 percent range (36-39%) after their first week of play. Greymane managed to just break the 40 percent mark, coming in at 40.9%, but he is right in that same neighborhood. People use these numbers, which are taken from the front page of HOTSLogs, to argue that these new Heroes are bad. Other people read/hear these comments, and then repeat them. There are two problems here.

The first is that the people taking the information from HOTSLogs often don’t fully understand where the numbers come from, or what they really mean. They simply see the new Hero close to the bottom of the list for win rate, and assume that means they aren’t good. They take a very limited amount of data out of context, and form a definitive opinion based on it. This leads to the second problem, which is that other people take these opinions, and repeat them as if they are a certainty. They probably haven’t even seen the actual numbers, they just know that a poster on Reddit or their favorite streamer said that such and such Hero has a lousy win rate. Sadly, many of the people saying that these new Heroes are bad haven’t even played them. Since only a fraction of the player base purchases a Hero immediately upon release, they only know that the Greymane in their last game died 14 times and that a website says that he has the third lowest win rate in the game. Of course Greymane is a bad Hero!

…or is he?

For starters, let me explain what I mean when I say they take the data out of context. That 40.9% win rate I mentioned earlier came from the front page of HOTSLogs seven days after Greymane was released. If you hear a person talking about the win rate of a certain Hero, this is probably where that information came from. The first thing that we want to be aware of is that the numbers on the front page are based on Hero League games with a Hero level of 5+. This is an important distinction, and it is something that people often don’t take into consideration, even though it is noted at the top of the page. Knowing where these numbers originate, does it change how we view them? It certainly should. For one thing, the sample size is small, and for another it is limited to one specific situation. You can’t tell how good of a student someone is by just looking at their history grade, or how well they did on their first test. The same is true for new Heroes.

Let’s take a look at a few numbers and see what we can deduce from them:

• In Hero League, Greymane had a 40.9% win rate and was the 18th most popular Hero during that week
• In Quick Match, Greymane had a 45.9% win rate and was the 5th most popular Hero during that week
• There were roughly three times as many Greymane games played in QM as there were in HL during that week

The first thing we can see see is that judging Greymane, or any other new Hero, by simply looking at the front page of HOTSLogs only tells part of the story. This shouldn’t be a big surprise now that we know these numbers are only based on Hero League games. Not only are Greymane players still learning how to play their Hero, but their teammates are also still figuring out how to play with a Greymane on their team. This puts that team at a double disadvantage, and the results reflect that. Hero League win rates can be a great tool for measure established Heroes, but they aren’t the best way to measure new Heroes. The sample size is just too small, and the nature of HL doesn’t lend itself well to brand new Heroes.

In my opinion, Quick Match is a much better gauge for judging new Heroes, and Greymane’s numbers in QM were quite respectable. For one thing, there were a lot more games played in QM, which means more data. More data means less variance, and less variance means a more accurate picture. There is also a higher chance in QM that there will be someone else who is also playing a Hero who is new to them, which helps lessen the downside of playing a newly released Hero. He still has a ways to go to reach that magical 50% win rate, but 45.9% seems like a reasonable starting point given the learning curve for a new Hero.

It is natural to expect win rates for a new Hero to increase as players have more time to figure out how to play as that Hero, and with that Hero. Greymane certainly seems to be following this path, as his percentages have gone up slightly in both HL and QM since I originally gathered these numbers on Tuesday. In fact, they have been going up consistently since he was released, but this doesn’t stop some people from claiming his win rate is too low. New Heroes don’t usually win a lot. That doesn’t mean they are bad, it just means that people are still figuring them out. People often complain that the most recent Heroes don’t have the same immediate impact on the game that a Kael’thas, Johanna, or Leoric did. What they forget, or choose not to mention, is that all of those Heroes ended up having to be nerfed. You can’t take a Hero that is over tuned and use it as evidence that another Hero is under tuned. These are not fair comparisons, but people make them anyway.

I love numbers, but when it comes to new Heroes they don’t tell the whole story. The best way to form an opinion on them is to either play them, or talk to someone who has. While not every Hero is going to fit you play style or flavor wise, you should at least be able to tell if they are viable. When it comes to Greymane, the comments that I have heard from people who have actually played him are generally positive. The consensus seems to be that he is a solid Hero who will have a place in the meta. The fact that he can play as either ranged or melee should make him a good choice in a wide range of comps. He also has a good fun factor, so he should see play across all skill levels. He has quickly become one of my favorites, and I can definitely see him having a spot in my Hero League line up.

HOTSLogs is a great tool that provides a wealth of information, both about your game and the game as a whole. I would encourage everyone to upload their games and analyze the resulting numbers. Don’t forget to use the ‘Additional Filters Available Here’ option to adjust how you filter the data that you are presented with. There is a saying that 80% of statistics are made up on the spot. Even when they aren’t made up, it is easy to find numbers that back up whatever point you would like to make. Keep that in mind the next time someone tries to use statistics to give you their opinion and pass it off as fact. They might be right, or they might be wrong. You have to judge for yourself. Good Luck and Have Fun!