Predictions are especially hard this year. Few DPC tournaments, problems with international competition in general and multiple small patches after the last big event all make guesses unreliable at best. We will still try to do our best to provide the reasoning behind our picks and oftentimes discuss the alternatives.
The most picked category often belongs to a hero that is strong and flexible, but not necessarily overpowered. Shadow Demon fits this description extremely well: he was among the most picked hero in the last several international tournaments, is often ignored as a ban and generally fits a lot of drafts simply because he is among the strongest stacking supports in the game.
When it comes to most banned heroes, we feel like it is going to be either Beastmaster or Batrider. Two BKB-piercing disables that always find their game and can have a massive impact regardless of the state of the match. Once again, it is The International and safe, comfort picks are going to be a priority. Between the two, we feel like Beastmaster is more likely to take this category: recent statistics support it and his scouting ability is close to unparalleled.
Highest win rate is most likely a cheese last pick hero. A hero that doesn’t appear too often, but who is absolutely impossible to deal with in the right game. Given the state of heroes like Meepo and Broodmother, we feel like this category will probably belong to either Tinker or Arc Warden. We lean towards the former, especially given how multiple participating mid players have been training the hero in pubs.
Highest kill average shouldn’t be too dissimilar to the previous category and in all fairness, it could once again be Tinker, who has problems closing games and pushing, but has no problems killing heroes over and over again. Given the change to the hero’s playstyle, it is also likely he’s going to be played in a very active manner. According to stats, however, there is an even better fit here: Ember Spirit, who can consistently set up and finish off enemy characters.
In the Assists category, one can’t go wrong with Silencer in the current meta. Global spell hero who rarely takes kills for himself is probably a good pick. Depending on how you feel about the state of Nature’s Prophet, he can also be a viable choice, as are typical Spectre and Zeus.
Lowest deaths is almost always Anti-Mage and why would it change? An elusive character who doesn’t want to fight at all and rarely joins risky endeavours. It is not uncommon for the hero to lose the game with zero deaths as well.
When it comes to last hit averages, this category is usually occupied by either Naga Siren or Anti-Mage. We are almost certain there are going to be at least five Anti-Mage games, but not so sure about Naga. Personally, we feel like Naga Siren is tactically weak right now and is heavily outclassed by most carries in the game, illusion or otherwise, but she is a very strong macro hero and there are going to be teams who will successfully leverage her almost global presence.
XPM average is weird in the current patch, since kills and teamfights is where most XP comes from. Picking a dedicated position one farming hero in this category feels risky in 7.30. Instead, a hybrid teamfight and farming core like Templar Assassin is going to be our choice. You also can’t go wrong with Anti-Mage and he is ahead of TA stats-wise, but we feel like the hero won’t be as powerful and as reliable at the highest level of competition.
The highest kills in a game belong to a snowball hero who in the right game can start killing early and never stop. This is Tiny, in our opinion. His burst damage and late-game scaling allow him to stay relevant throughout the game and in a good match he can go on an absolute rampage, before tossing his teammate to Tier 4 towers and ultimately losing.
Finally, the most last hits in a game is once again Naga Siren. This one is even safer than the last hit average category, since there is going to be at least one Naga game, this is almost certain.
The first pick is, realistically, between 45 and 54 and depending on how you feel the teams are matches this time around, you can’t go wrong with either 45-49 or 50-54. We will give our preference to the latter: 50 to 54. In part because we feel like the teams are matched somewhat evenly and there are a lot of full Bo3 possibilities. In part, because after missing a year, we want as many games as humanly possible.
Second category is 101+. Compared to competition, Dota doesn’t produce the same hero time after time, using one of roughly five predefined templates. Every hero in Dota is unique and every one has a reason to appear in a professional game.
Similarly, 101+ is a good guess for heroes banned as well. Last TI every single hero was either banned or picked. Now, there are four extra heroes in the game, but the number 101+ stayed the same, so we think it is a safe bet.
Combined total kills is trickier. We are going to go for 91-100 for a “slightly above average” prediction, but one above and one below are both viable choices.
For the longest game, we are going to go with 70:00 to 79:59. The pick above (80 to 89) is historically more accurate, but so far The International was played without the neutral items. Tier 5 items are a big deal and they should allow the winning team to better translate their advantage into victory.
The shortest game, in our opinion, is going to be between 15:00 and 19:59. This is what happens historically and there are no reasons to deviate.
For kills, going 26+ might not be safe, but we feel like there is always at least one game where a hero just gets completely out of control and simply never stops killing. Given how we predict Tinker to be a big deal this tournament, there is no reason not to go for 26+.
There are times where players get out of control in the opposite direction as well, but we feel like professional players never stay tilted for long. 15 to 17 is our pick for most deaths.
36+ assists is honestly just your average long and active game and 1000+ GPM is just your average Alchemist victory. The latter will probably appear at least a couple of times, but if you don't feel so, going 900 to 999 can be a viable pick as well.
Feel free to share your opinions in the comment section below and return tomorrow when we discuss the team and player predictions.