By quincy0191

Part One in this series (which wasn’t initially intended as a series, but that’s just part of this funny thing we call life) focused on which heroes were actually really strong, and not just by counting the whiny comments after someone loses a game. I was going to be mildly more scientifically rigorous and count only the whiny comments with expletives.

But then I figured that would take too long, I already have this big spreadsheet with numbers, and I’m lazy. Though evidently not lazy enough to turn the concept for one article into three, or maybe that is in fact validation of my laziness. I’m getting off track.

This time out we’re interested in the worst heroes, specifically so we can point and laugh at the Thai cosplayers for dressing up like such a bad hero and mask the fact that they’re way more attractive than we are. Like before, we’re looking for a hero that has played as many games as possible to reduce the impact of a few chance losses, but has also failed as miserably as possible as much as possible. Here’s our formula:

Games * (Wins-Losses)

Last time we wanted the highest numbers, and in the fashion of everything being flipped upside down, this time we’re looking for the biggest negatives. Heroes with an even record will be punished, as net wins will be close to zero, bringing their scores closer to zero. Heroes with a low number of games played and a low win% won’t be doing well either, since they won’t be maximizing the games played portion.

The Five Worst Heroes

Leon casino,



Whaaaatt?! That is cray cray. Sure, the fact that they’ve played a lot of games is helping them get on here, but Torturer (#8), Rhapsody (#4), and Bubbles (#5) are all in the top 10 most played heroes, and it seems like they’ve been five of the worst over HoN Tour Season 3. That’s really because they have; Torturer blows everyone out of the water with a miserable 40.8% win in 103 games – teams have a startling dedication to picking a hero that just does not win. After ten or twenty games you could say he’s just having a bad run, but 100 games and a forty percent win rate is extremely odd.

It’s not like one team is doing all the damage here, either, as he’s been put in the hands of 17 different clubs, and his #1 fans in BMG have a winning record in their 26 games (double #2 Dawn’s 13 tries, though they’re at just 30.8% win). SynC have stayed away from Torturer, with just a 3-1 record, so he’s not getting help from the big dogs. It might be time we face the reality that Torturer is just not a good hero.

I refuse to accept that for the other support, though. Rhapsody is the best hero in the game, and I don’t care what the stupid numbers say. They don’t know her like I do. But it’s pretty difficult to argue against 154 games of subpar performance, and it’s really hammered home when you realize that she topped the charts in Cycle 2 at 68.2% and was played a ridiculous 45 times in the first two cycles with a 62.2% win rate. Since then, she’s mostly hovered between 35-45% with a low of 27.3% in Cycle 6. A 4-1 record the next cycle didn’t signal a turnaround, as she dropped back to 36.4% for Cycle 8. Rhapsody’s just had a miserable go of it since Cycle 2 ended, seeing far fewer games and much rarer wins. But that’s probably just a series of coincidences, guys, it’s not like Torturer! She’s a great hero! I’m not in denial!

Bubbles also hasn’t actually been good. As I’ve mentioned a few times on stream, though, HoN’s #1 team rarely plays The Bubs, and that’s not very helpful – it may seem like a cop-out to continuously talk about SynC and their influence on these, but it’s highly relevant when a team like that does or does not play a hero given their insane winning percentage. Bubbles’ lack of consistent victories is probably significantly influenced by iNsania’s prior lack of interest in picking him, and though that does say something about him, his versatility should be enough to keep him as a consistent pick.

The Masters are next up, and it’s pretty surprising that Puppet appears here, though you’ll notice he’s not nearly as bad in points as Bubbles and almost three times better than Torturer. He’s also the first hero that hasn’t been played a ton, which lends more credence to the idea that he’s just not that good, or he’s not being used in the right role (support PM for president!), or he’s not being used in the right games. Whatever the problem is, he isn’t winning.

We round this out with DM, who has long been on my list of bad heroes, and appears here again. A 31% win rate is spectacularly bad, the second-worst in the game of any hero who’s been played at least 20 times (Ravenor’s at 28.6% and climbing), and he’s exactly the sort of hero this is meant to identify: one that teams just should not pick. Drunken doesn’t do anything well enough to merit a selection; he’s not a good mid, suicide, or carry, being sort of all of those things and not really any of them. It’s not like his underlying stats are bad, either; his K/D/A is fine, and his GPM is good. He just doesn’t win.

The Next Five: Pharaoh, Wretched Hag, Prisoner 945, Kraken, Madman
Pharaoh has the lowest GPM of any non-support at 253. When he gets even mediocre farm, 275 GPM, he rockets to a 92.3% win. Is he good because he often wins with what should be considered entirely attainable farm for a mid/suicide, or bad because he apparently rarely gets it? Only Mini knows. Wretched Hag has seen a precipitous drop in use since the first few cycles of HTS3, and though she does still show up here and there, a 38.6% win suggests she probably shouldn’t. Hag really has the same problem as Drunken: what does she do that another hero doesn’t do better?

Prisoner wins about as often as Torturer, he’s just played less, and feels like he really doesn’t belong on this list. But in over 40 games, he hasn’t performed, and would be the most surprising inclusion were it not for Kraken. Rejoice, however, because Kraken simply suffers from being the game’s second most popular hero at 162 games and sitting just below an even record at 49.4% win. He’s one of the cases where the formula doesn’t work all that well; Kraken is a middle-of-the-road hero that could flip to the complete other side with a short burst of wins. Rounding out the top 10 is Madman, yet another hero caught between roles; he doesn’t do well as a suicide, and doesn’t do well as a carry, but is about equally bad at both. So he gives you options, which has value in itself, they just are both poor ones, and that kind of removes most of the reason to pick him.

Top Five (Non-SynC games): Kraken, Magmus, Torturer, Bubbles, Rhapsody
Once again, we’ll see what happens when SynC steps to the side, and this time it’s an episode of When Methods Go Bad. These are just five of the top six most played heroes excluding SynC games – Glacius is #3 – and they all have records that are just under 50%, with only Torturer below 45%. They’re being penalized for getting played a lot and losing more than they win, which does say something, but not really anything meaningful. None of them are terrible, they’re just a little south of perfectly average, and very popular.

This time round we might have actually identified something of value. The first thing that stands out is the presence of two support heroes on the “worst” table and zero on the “best” table. There are two conclusions to be drawn from that; the first is that jungle strategies are better. When a jungler faces two supports, it wins more often, thereby driving overall support win% down (one win for support on the jungler’s team, two losses for the supports opposing). The second is that oft-picked supports lose a little bit more often, and that rarely-picked supports win a little more often. Apparently there’s something to be said for being creative with your support picks, and following the crowd when it comes to cores.

Second, there are heroes here that really should be picked less often than they are, though in general things “should” be a lot worse; teams aren’t stupid, so they don’t keep picking heroes that lose, and that means you generally won’t see high numbers of games coupled with really low win rates. The other side of that coin is that teams will imitate strategies that are winning and consequently lower the value of those heroes as they get diluted, but that means high win% heroes are pushed to larger games played totals and usually retain records over 50% at least thanks to their initial success.

The point is that a) this list is bound to contain a few heroes that are fine because the truly bad won’t get enough games to show up and b) it will contain a few heroes that you didn’t expect to see because you’re not used to seeing heroes that lose a lot get picked every game, so it’s harder to form an idea of them as bad. Maybe that makes it more valuable than the previous list, maybe less valuable. Either way, stop picking Torturer.

quincy0191
quincy0191 is a HoN veteran with a focus on the numbers. Fascinated by quantifying and valuing human beings, he shunned the world of finance in favor of sports and competition. When not on Honcast you can find quincy playing games like Civilization and Pokemon, or watching movies, writing, pondering, or catching a game of baseball.