3 Interesting Pub Trends

I observe a lot of different Pub Trends, but some of them don’t warrant an entire article to themselves. So here is a 5-in-1 interesting stat fueled trends that have been happening over the last few months.

Lightning Slow?

Leshrac’s Lightning Storm has received a bunch of changes over the last few patches, including the addition of a brief slow. However these don’t seem to be paying off as Leshracs who level Lightning are more likely to lose than those who don’t.

The trend is that Leshracs who opt for Attribute Bonus at levels 9, 10, 12, 13 and/or 14 win more than people who get Lightning Storm at that level. Additionally, the Stun + Edict build at the early levels also happens to beat out the Stun + Lightning build. This is very interesting, since a third of Leshracs go Lightning. Cause vs Correlation is a massive argument for Skill Builds (aka, Do people get Lightning more when they are losing?) One factor is that Leshrac has no massive win rate difference between skill brackets, so there’s no false positives from this. Another is that Leshrac builds are determined at Level 2; before most teams have a big advantage.

Statistically one of the better build paths for Leshrac. Note that Level 3/4 are interchangeable.

Statistically a very optimal Leshrac build, with levels 3/4 being mostly interchangeable (light green).

The Lawnmower Leshrac build is incredibly powerful; you press a couple buttons and then run around mowing everyone down with incredible aoe damage. It also allows you to get crucial early game towers. This suggests to me that 6.82 should buff Lightning, since the stats say the skill doesn’t measure up just yet. The recent buffs improved Lightning for the third of people who chose it, but not to the point where they win more than Team Edict. Hmm…. now all we need is a Team Edict t-shirt.

A recent game where I transition into a Semi-Carry after a successful start.

A solid early game thanks to Edict often results in you transitioning into a powerful Semi-Carry

Tusks Are Improving

When Snowball was reworked in 6.81 it effectively raised the Skill Floor of the hero, resulting in a massive drop for his pub win rate to 39%. Luckily this win rate was not forever. Over the weeks the average Tusk player improved as they became more used to the ability. A few minor buffs in 6.81b to Tusk also helped push him along. The trend however is very clear: Tusks are getting ever so slightly better every week.

Tusk had about a 43.6% win rate in June, going to 44.8% in July, and for the week of July 28 – Aug 3, his win rate was 45.1%. Even better, his win rate is higher in the Very High skill bracket. For the week mentioned above he has a 47% win rate in Very High, which is also steadily increasing. His pick rate has remained steady over the last few fortnights, so no shenanigans there.

It snakes up and down, but overall it’s a slow and steady upward inclination. I am very interested to see where the Tusk train will stop, he had about a 47.5% win rate before the rework. Can he overtake it?

Necrophos To Boot

Necrophos has an insanely high pub win rate, but there’s also a wide variety in boot choices for the hero. However what is popular isn’t always the best, so I ran the numbers, ignoring Boots of Speed + Travel as always since the data is too skewed for those options. Boot choice is a great item to see the effectiveness of, since people chose them very early into the game and are less impacted by a hero winning/losing in that game (except Speed/Travels). Radiance has really high win rate because people almost always get Radiance when they are winning (such as on Spectre, Centaur, Bristleback, etc) the 4 main boot options however don’t generally fall to into this trap.


You can see that Power Treads are the predominate boot choice in both Skill Brackets, but decline in Very High as they mostly flow into Phase Boots. I compared the boot’s win rate in that bracket, to the win rate of the hero in that bracket . Interestingly, Necrophos with Arcanes wins more than those with Power Treads. A lot more. Which means I can say that the statistics strongly suggest that Arcane Boots are the best choice for the hero for how he is played in pubs. Which also lead me to saying that whatever role and/or play style Necrophos is playing in games that he gets Arcane Boots in, is the statistically superior way to play Necrophos.

Necrophos uses a lot of mana and being able to safely have enough mana to use your abilities is vital. People underestimate how much mana he uses, especially with a fast Mekanism. Relying on Sadist is often an unreliable option, perhaps even causing you to Scythe someone for Sadist, and not for the skill itself. Which is more important now that Scythe adds +30% respawn timer. I assume that Necrophos with Arcanes are less likely to get levels of Sadist early on, which is important as Skill Builds that skip early levels of Sadist also have a slight win rate edge. Like with Pugna, Arcane Boots also help your entire team with Deathballing, which is when a Necrophos level advantage and constant heals come in handy. These few issues are my two cents on the problem with treads as evident by their winrate.

One thought on “3 Interesting Pub Trends

  1. I think an important stat that you left out is that Arcane Boots always have a higher winrate than other boots on any hero. The exceptions being the usual Bloodstone builders.
    I would be interested to see how much better Arcane Boots are on Necrophos than on other heroes, as well as how the generic winrate for Arcane Boots are in Normal, High, and Very High Skill brackets fare.
    Additionally, I’d like to see the same test done on Dark Seer and Enigma, both of which have highest build rate with Arcane in addition to a high build rate of Soul Ring, even though the item is almost never built on them in the competitive scene.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s