The final part of our predictions. The tournament predictions are probably the easiest, since, for the most part, they are less dependent on the meta and choosing from <10 categories is easier than picking one of 117 heroes, one of 90 players or one of 18 teams. Do not hesitate to change your own predictions if our reasoning seems flawed.
It was “45 to 49” a year ago and “45 to 49” was also the correct answer in The International 2017. The tournament format hasn’t changed and any speculations regarding relative power levels of the participating teams are just that—speculations.
101+ heroes is probably the best bet. Meta is very diverse, many strategies are viable and teams are going to experiment on the brink of elimination against a tougher opponent. It was also the correct answer the last two years and we feel Dota improved over this period.
91 to 100 is a safe bet. It makes sense historically and logically, since most heroes are worthy of a pick in one situation or another, if only to surprise the opponent, but not all of them are going to get banned.
This is pretty much a guess as you will never know whether a game will turn into a bloodbath. What we do know is that this TI we have the extra benefits from third and fourth Roshan, which theoretically should help close the games. This means slightly lower amount of kills. We advise to pick around 100 kills total, going above if PSG.LGD and TNC Predator are in the same group and lower otherwise.
70:00 to 79:59 is going to be our prediction for this tournament. There are more benefits to controlling the map, compared to last year. There is more XP on the map and slow starts are going to be rarer. Gold now comes faster as well, meaning power levels of the heroes increase faster, if they are uncontested. All of that coupled with a potential “free” gold injection from the third Roshan makes us believe that this category is probably the safest bet.
This is historically correct figure and there is no reason for us to believe it is going to change. Perhaps the changes to deny XP might allow teams to stand their ground for a little longer, but it is still quite unlikely.
Around 25 is usually the safest bet, at least historically. Sometimes one player just seems to take off and start killing enemies left and right. We are going to go just below 25, since, as discussed previously, games should be theoretically shorter and less bloody this patch.
Once again, this is historically correct answer and no amount of balance updates can change the fact that once in a while a support player can absolutely desynchronize with the game and go on a feeding spree. Go one category lower, if you feel like the level of competition is significantly higher this year, compared to the previous two years.
It is always 36+ and there are no reasons for it to change, especially with TNC.TIMS attending the tournament.
Alchemist is probably going to be picked at least a couple of times and will have at least one decent game. This is enough of a reason to go for 1000+ GPM, especially since ranged creeps now ramp up their gold bounty slightly faster, compared to last year.