With The International 11 starting on October 15, 2022, Ben “Noxville” Steenhuisen shares his Compendium Predictions. As always, he looks at the data behind the game and combines that with his superb game knowledge. In this article, he shares his insights on the heroes and tournament statistics.
Note: this article was written mid-way through the TI11 Last Chance Qualifiers (LCQ) event, so there might be some new and useful information on teams and players to come, which could affect his predictions. Also, many of the team and player predictions rely on the group stage draw. Therefore, “Noxville” will stick to analysing the heroes and tournament statistics since those don’t rely on individual matchups as much.
Remember that for all of these predictions, you are able to bet real money on them here at Pinnacle.
The Last Chance Qualifiers give us a unique international LAN insight into the metagame just before TI11 itself kicks off. We do need to consider that the Last Chance Qualifier teams might have separate preferences to the already qualified teams – but due to them all scrimming together for a while now – it’s probably much closer than many would expect.
What are my hero predictions?
Undying has been picked in nearly half of the LCQ games, with Shadow Fiend a small distance behind. Marci would be a contender here if she wasn’t banned so frequently. If Undying starts being banned a bit more, Tusk or Earthshaker could get picked more and more. However, I’ll begrudgingly go with Undying for Most Picked Hero.
Marci has been the most banned hero (and the most overall picked or banned hero – not once ignored in a draft), followed by Enigma and Batrider. Almost every top team has an excellent Enigma player, so I would imagine this is more of a Marci/Enigma battle. I’ll go for Enigma as the Most Banned Hero - this hero invokes such fear for so many teams, I’m sure it will remain highly banned.
Choosing which hero will have the Highest Win Rate With Five+ Games is always a risky matter – the logic is that you want some busted hero which you expect to be banned a lot (like Enigma and Marci) or some third-phase gotcha pick (Slark/Ursa/Pudge) – but there aren’t that many really. I’ll go with Pudge, who’s 2-0 in the second phase and 7-1 in the third phase. Other niche picks perhaps are Anti-Mage or Necrophos – both have been banned but remain unpicked in the LCQs so far.
For highest kills average, I like Ursa, Slark, Morphling, and Ember. For kills in a single game, I like Tinker, Sniper, or once again Ember. I’ll go with Ember for Highest Kill Average, and Sniper for Most Kills In A Single Game.
Average Assists is a tricky category, there are so many viable options. Disruptor has been doing well, but is also a favourite of many of the better LCQ teams. Skywrath Mage, Dark Willow, Chen, and Spirit Breaker represent good value also. Io and Clockwerk in longer games can alsogain plenty of assists. I’m going with Clockwerk, a commonly picked hero who can earn many kills throughout the game and is reasonably commonly picked.
For Lowest Death Average, once again, there are a few decent options: Slark and Morphling are my favourites but I could be convinced to get behind Lifestealer or Lycan. I’ll trust the statistics on this and go with Slark.
For Highest Average Last Hits, if Medusa were more commonly picked she might be viable, but this seems more of a decision between Terrorblade and Templar Assassin for now.
Terrorblade is more likely to win, but only marginally.
For Highest XPM, my choice has got to be some ancient-killing fiend. Templar Assassin, Ursa, and a more active Lina could take this but Templar Assassin has the edge. Tinker could be a stompy hero who I could back – but I don’t think it will be picked five times (as good Tinker players will have it banned against them too much). Templar Assassin is, therefore, my pick.
Most Last Hits In A Single Game has got to go to Terrorblade, who has appeared in five of the top eight Last Hit records in this patch, and three of the top eight in the LCQ. Any epic game can upset this record, but you’ve got to play the odds.
What can we expect when it comes to the tournament itself?
The playoffs are formatted as four best-of-one, 17 best-of-three, and one best-of-five matches. The PGL Arlington Major averaged 2.7 games per best-of-three, putting us on the precipice of 50-54 and 55-60 depending on the finals being 3-2 or not. This is always difficult but I’m going to stick with 50-54 for the Total Number Of Games Played and hope there are a few more clean sweeps.
In just 81 games, the Last Chance Qualifiers have already seen 105 heroes picked and 97 heroes banned. My gut is that in the 230+ games of TI11 overall we will see both picks and bans get to 101+.
Combined Total Kills In A Game is a really a risky prediction to make, due to how many long sluggish games there are (often games start off as very one sided and then get very close). 91-100 is a relatively middle of the road value, but it could be one step up (101-110) or down (81-90).
So far in this current patch (with around 730 games played so far) there’s been just 11 games shorter than 20 minutes, and just one of those (Hokori vs. Tempest) features TI11 teams. Historically, 15:00 to 19:59 is the safer bet, but I’m betting on one step up, 20:00 to 24:59, for the Shortest Game Of The Tournament. Twenty-six of the games in this patch (3.57%) have been longer than an hour, but only two of those have been longer than 80 minutes. I think the current patch is more difficult for longer games, so I’ll go against conventional wisdom and go for the shorter time of 70:00 to 79:59 for Longest Game Of The Tournament and hope there’s no epic stalemates.
So far in this patch, there have been seven games where a player has had 26+ kills. This will very likely be 26+, or one category below (23-25), but I’m going big with 26+ kills for Most Kills By A Hero In A Game. Last year had just two games with 26+ (Monet and Nightfall); however, there are around 36 more games this year.
Similarly, there’ve been 14 games already this patch with 36+ assists, so surely that’s the favourite option here and I’m going with it for Most Assists By A Hero In A Game. Last year, this was set by Vitalie "Save-" Melnic with 38 on Bounty Hunter and there have been musings about Bounty Hunter tactics from Singaporean scrim insiders!
Compared with previous years, there are slightly more cases of extreme feeding, across a wide variety of support heroes (Tusk, Skywrath Mage, Clockwerk, and Crystal Maiden to name a few). When looking at the ESL One Stockholm Major this year, only twice did a player die more than 17 times (both times were Stinger). I’m going to go under here, with 15-17 for the Most Deaths By A Hero In A Game and hope that at TI11, the feeding will be more controlled.
There have been seven instances of 1000+ GPM By A Hero In A Game this patch, on four different heroes. I’m backing that number to go up by at least one during TI11.
What about team and player predictions?
The rest of the Compendium Predictions will be released right after the TI11 groups are drawn, since a lot of them are affected by individual matchups. I can reveal that the team I have as the overall predicted winner for TI is PSG.LGD. That said, the odds that we’ll have a repeat winning team or player are relatively good. OG are a formidable team so they could win a record third TI title, or perhaps Team Spirit can make it two in a row to match OG’s TI8-TI9 run. Both have shown they have the pedigree to beat the best and are also likely finalists, but as of now I’m sticking to my PSG.LGD pick.
I also believe that these three teams will be the top teams from their regions (OG from Western Europe, Team Spirit from Eastern Europe, and PSG.LGD from China), along with BOOM Esports from Southeast Asia. South America and North America are tricky; it could be close between TSM/Evil Geniuses and Thunder Predator/beastcoast depending on the form of the teams.