The compendium predictions are always tough. Even if you analyze past data, cross-reference it with the current meta, add to it common Dota sense and adjust for TI having slightly different stakes and different priorities than most other tournaments, you are still left with <30% chance of being correct at best. So take these and all predictions with a grain of salt and read the text provided to understand our thought process.
Versatility and general power level are the two main contributing factors when it comes to the most picked hero of any given tournament and TI is no exception. The most picked hero is going to be someone who can fit into most lineups and is powerful, but not offensively enough so that they are banned in every game.
Before the last couple of patches our vote would surely go to Ember Spirit, but the hero did receive multiple nerfs and while he is still being played, it is unlikely he will remain as popular as he was during the qualifiers.
Centaur Warrunner is in a similar boat. The hero was incredibly popular before the nerfs, but haven’t been tested on an international stage after them, so it can be a rather risky choice. The same, but to a lesser extent goes for Grimstroke.
Low popularity of some of the most TI-speicific heroes, such as Batrider and Beastmaster makes the prediction even harder, so our educated guess is a lot more of the latter with only a hint of the former.
Our pick is going to be Spirit Breaker, since we are feeling risky and the hero was slowly getting more and more popular, without receiving many nerfs. He can also be played as both position four and position five support, so there are more ways for teams to fit in this BKB-piercing disruptive support into their lineup.
This will probably be either Chen or Io.
Points for Io: it is TI, best teams are attending, Liquid is looking strong and global presence allowing for insane tempo should never be underestimated.
Points for Chen: it is TI, best teams are attending, Secret is looking strong and global presence allowing for insane tempo should never be underestimated.
We believe that Chen is played by more players across more teams, so he is going to be our choice, since he will get banned more often.
People think this category will most likely be won by a last-pick hero. “Typical Broodmother, Huskar, Meepo, Arc Warden cheese works only under specific conditions, but if it works it is more or less unstoppable” is the theory.
But then we take into account what happened last year: Meepo wasn’t even picked five times, Broodmother and Arc Warden were overused as a cheese start and had their win rate lowered to sub-70%, while Huskar turned out to be simply bad.
The story was more or less the same the year before and we have no reason to believe the same thing won’t happen again. Going for a pocket strategy hero that is undervalued by most teams and is not widely understood as a “cheese” hero is probably a better bet. Or we can go even further and nominate a hero that is perfect for balancing out janky pocket strats and allow them to win.
We nominate Dragon Knight, because the hero received a lot of sizeable buffs over the last several patches, but no one truly explored him in a LAN environment. He won’t be a popular hero, he isn’t a flashy hero, but he is a hero that balances out any wonky lineup, providing reliable disable, target to show the enemy, push potential and decent late game scaling.
Your nomination can and probably should differ, but the general guideline stays the same: don't pick a cheese hero and go with something you feel is undervalued instead. Alternatively go for Meepo, if you think the hero will be played more than five times.
Clinkz is a safe, reliable bet that was in top10 top killers in the last couple of tournaments and, perhaps equally important, in the last two TIs. But the hero is also kind of getting a bit too popular—more games means more potential weak games where he can be shut down and completely fall out, potentially ruining the average.
Broodmother, as discussed previously, is a “cheese” hero and will mostly be picked in games, where she fits in naturally. She isn’t the most reliable win condition, but her playstyle generally involves annoying the enemy through constant push and jungle farming and this annoyance frequently leads to mistakes and unnecessary deaths.
But then there is Bloodseeker, who many are sleeping on and who is similarly a surprise pick. We believe that some teams are going to try the hero and find out that he works really well as a tempo core that runs across the map getting kill after kill. And since he is not going to be picked unless the team is certain they can fit him in, the probability of bad games is lower.
Go for Clinkz if you are not feeling risky, Broodmother if your risk tolerance is moderate and Bloodseeker if you are ready to ride or die with us on the riskiest of rollercoasters.
Both heroes were in top5 in terms of assists in the previous two TIs and their popularity and power level are around the same this year as well. Both are probably really good choices, but we will go with Spirit Breaker since he is simply a more versatile hero, is capable of assisting from the start of the game and is less likely to deal the last hit.
Flash farming heroes who play the avoidance game for the majority of the match are what we are looking for in this category and both Anti-Mage and Naga Siren fit the bill perfectly. These heroes rarely show on the map and do so only when they are sure they can’t die. Given how Naga is probably a slightly better laner we will go with her for this category.
Naga Siren and Anti-Mage are here for the same reason they were a part of the previous category. They hit creeps and then they hit the throne—this is pretty much the whole gameplan. Sven is kind of similar, but he does all that in roughly 80% of the time, which makes him a worse fit for the category that involves longer games.
And once again, since Naga Siren is just slightly better at stalling and finding farm than Anti-Mage, even in the worst case scenarios, we believe she is going to be the best fit here. Alternatively, depending on how you evaluate the hero, you can also go for Medusa.
Meepo, if you believe the hero is going to be played more than five times or Anti-Mage otherwise.
We are looking for a tempo core that can absolutely dismantle the enemy team in teamfights but can’t really push too well to capitalize on this and win the game too early. Whereas our average kills per match prediction of Bloodseeker is extremely risky, this one is a lot less so. Once again, go for Broodmother or Clinkz if you don’t feel like Bloodseeker is going to show up at TI as we do.
Sure, the hero is kind of bad right now, but there is going to be at least this one game, where Tinker is going to have a good time in lane and, being slightly on a weaker side, a bad game overall. And he will stall, split-push, teleport and blink all over the place and vacuum the whole map, taking all three lanes and maybe a part of the jungle. We are unsure whether he will win this game, but he is most likely to break some records, just like last year.
Dotabuff is continuing to develop new Dota products and would like your feedback to help us shape them! Tell us a little about yourself and your Dota experience and you may be selected to do a short video call with us, and receive a $20 steam gift card for your time.