Though he’s rarely seen in the professional scene, Riki continues to dominate pubs at a Spectre-like level across all skill brackets, sitting at a 59% overall win rate in the past month. He has flourished even in Divine and Immortal brackets, where players are keen to punish Riki’s weaknesses.
We’ve harped on here before about the strength of invisibility heroes in lower skill brackets. Riki's win rate drops to 56% in Ancient and 53% in Divine/Immortal, a respectable margin that's higher than heroes like Necrophos, Ursa, Phantom Lancer, and Silencer--the meta picks of high MMR pub games.
While Riki’s win rate has hovered between 58-59 the past month, his pick rate has increased 4.5%, the 2nd highest next to Pudge.
There was much speculation abound when Riki’s ultimate, Tricks of the Trade, got reworked in 7.12 six months ago. His win rate, in the days after the change, hopped from 52% to nearly 58% overall. Though it declined in higher MMR brackets, he was still above 50% in Divine and higher games.
Predictably, he was shortly nerfed in the ensuing patches. In 7.14, Valve increased Tricks of Trade’s attack interval, and in 7.16, increased its cooldown as well as Cloak and Dagger’s Fade delay. The nerfs didn’t stop there, as Smoke Screen’s cooldown increased from 11 to 17/15/13/11 in 7.18, and Blink Strike’s cast range reduced from 800 to 500/600/700/800.
Throughout it all, two overall changes became Riki’s saving grace. First, his base attack damage increased from 4-8 to 11-15 from patches 7.17 and 7.18. Second, Blink Strike’s damage increased from 55/70/85/100 to 100 across all levels. With the damage from Blink Strike being applied before his attack, Riki could essentially last hit and still remain invisible. He won’t be able to use it as frequently as Phantom Assassin’s Dagger to farm safely, but it let him salvage a bit of his weak laning stage.
Roaming was already the norm, but the introduction of Bounty runes became a boon for greedy position 4 support heroes. The success of roaming heroes now didn’t only hinge on the outcome of their ganks. They benefited from just moving around the map.
However, changes to the frequency of bounty runes—from 2 minutes to 5 minutes—as well as the removal of its experience bonus has essentially nerfed the roaming role in Dota. Compounded with the increased focus of the laning stage--how a lane can be swayed by just the amount of ranged creep denies—support heroes now had to stay put. With Riki being a weak laner, he had few opportunities as a support, and much less as a core.
Riki was picked just 3 times of the entirety of TI8. Liquid was the only team who won with the hero, with Miracle on it in the 1 position. He surfaced 8 times during the Kuala Lumpur qualifiers, with a 37.50% win rate. Compared to Spectre, Riki, once shutdown, has few windows to comeback into the game. He has no global ultimate where he can reap some teamfight gold. Those opportunities, however, are plentiful in pubs.
It’s not surprising that invisible heroes have historically dominated lower skill level pubs, but Valve has dampened their impact overall multiple patches. Shadowblade at one point used to have a cooldown of 18 seconds instead of 28. Sentry Wards (and Observer Wards) can be bought in singles and can stack in your inventory—a much needed quality of life change that made life easier for supports. In as recent as 7.16, Sentries got buffed, increasing its duration from 4 to 6 minutes.
Overall, Dota players have grown smarter over the years. Invisibility doesn’t always mean invincibility. Even in a 3K pub, you may see a sentry lane ward against an enemy Treant and Riki.
So, the way Riki takes advantage across the board in skill brackets may not just be the rigor of detection in pubs, but rather the pace of play. Games just last longer in pubs. They're are less coordinated, and hero synergies are determined on the fly than at the outset.
What we also got an object lesson watching OG during TI8 was that the meta isn’t solely about the lane. Some spectacular mid game play by Ceb’s Axe brought OG back from a disastrous early game. And OG’s Treant pick in game 1 ensured that they’d be able cut their early game losses and delay the game just enough.
Yes, Riki has some strong skills that seem to verge on being imbalanced. His 25 Tricks of The Trade Talent, already on a short cooldown, makes it a nightmare to participate in a teamfight as a support. His designed weaknesses are supposed to offset this. Keep him down once, and shut him down for good. And yet in pubs, mistakes don’t necessarily compound as they do in competition, because they’re plenty to be made by both teams. That's the kind of room Riki needed to takeover pubs.