No menu items!

    Dota 2 – Manila Major Compendium Predictions

    The Manila Major is just around the corner and as with every Major, you can enter predictions into your Battle Pass on the outcome of the tournament. Among “simple” predictions such as the tournament winner or the final prizepool [EDIT: for TI6], you can also predict more intricate topics such as the amount of total games played, heroes picked or which player will highest GPM average.

    If you might be lost on what kind of predictions to make, but wish to take on the gamble for some sweet Battle Points anyway, we have you covered. A lot of community renowned personalities have done their own take on these predictions and we will guide you through them with a list of alternatives that you can and maybe should pick at your own leisure.

    A list of predictions can be found here, compiled by reddit user Kainaabz. Links to the individual prediction videos can be found at the bottom of the article.


    When looking at any predictions, it is vital to look towards the most recent tournament trends, especially the big ones such as Epicenter. It definitely showed what a lot of the top teams favor and like to do and what common picks and bans there are.

    Lion and Beastmaster have been amongst the most contested heroes in pro games for a while now and tied as most picked (31) at Epicenter. However, Lion was only rarely banned, unlike Beastmaster who is definitely banned a lot. In fact, he was banned in 16/18 DreamLeague LAN Finals matches. It makes sense to pick Lion as most picked and Beastmaster as most banned. There’s arguments to be made that Bounty Hunter and Doom could contest Beastmaster in regards to bans. That said, Bounty Hunter has fallen off quite a bit in popularity and he was in fact neither picked nor banned a single time during the Epicenter grand finals. Most analysts seem to agree, although Purge and Siractionslacks brought Disruptor to the table–the hero certainly has risen in popularity. Merlini chose to go with Phoenix, another viable hero that has risen in popularity and success–the grand finals of the North-American qualifier for The Summit 5 had Phoenix win every single game.

    Lion, Beastmaster, Doom, Disruptor, Phoenix. All viable heroes for most picked.

    Lion, Beastmaster, Doom, Bounty Hunter. All viable heroes for most banned.

    The hero with the highest winrate is likely a hero whose overall games played number is just above the minimum requirement of five. All analysts agree that there is no way of making a certain prediction here, so going with a gut feeling is not wrong. Nahaz and Noxville both went for Tidehunter and Gyrocopter, Merlini chose Lycan whereas Purge and Slacks opted for the Phoenix. It makes sense to go with these pocket type of heroes that work in a specific strat that is difficult to counter. Lycan is a Team Liquid favorite and if it’s not banned against them, you could bank on them to bring that prediction home for you. That said, there’s nothing wrong with choosing heroes such as Clinkz, Slark, Medusa or even Viper. Viper is usually picked to dominate a lane and never picked early enough for the opponents to really counter him. In fact, at Epicenter he was picked exactly 5 times and won 4 of those games.

    The hero with the highest kill average should yet again be a hero with only the bare minimum of required games. It’s ideally a hero, that gets kills regardless of the state or outcome of the game. Classic pubstompers come to mind such as Slark, Ursa or Queen of Pain. Some analysts brought up other heroes that have proven to be niche picks by certain players and are also capable of completing kills, such as Templar Assassin or Tiny.

    As for highest assist average, heroes with global spells come to mind. Spectre is definitely a viable pick here, considering her natural ability to be part of every engagement. With her, it’s simply a question of how many games she’ll actually be picked in. Nahaz chose Dark Seer and there are cases to be made for supports or gankers to be chosen as well. In Shanghai, Earth Spirit had a 12.42 assist average after 12 games. At Epicenter, Ancient Apparition had an assist average of 18.67 with 6 games played.

    A lot of analysts look towards Morphling to be hero with the lowest death average, which is fair because the hero is innately difficult to kill and he shines as a last pick when there’s a low amount of counters against him. That said, he’s generally not picked that often, so alternatives are certainly to be considered. Lycan is also very elusive and difficult to kill, but is much more popular. Team Liquid alone picked him up 4 times during Epicenter. Because he is such a popular hero of theirs and they just won Epicenter, there’s a good chance this hero could be banned against them often times. Lone Druid, Anti-Mage or Naga Siren are good alternatives, although Naga is also not sure to be picked at least 5 times.

    The hero with the highest last hit average is another category that splits the opinion of multiple people. Naga Siren, Medusa, Alchemist, Sven and Anti-Mage are often named. A black sheep in this race could be Tinker, a hero that has gained increased popularity as of late. His lasthits are generally not as reliable as those of the other heroes mentioned, but it’s worth a shot in the dark if the other options seem to generic.

    As for highest GPM average, there should only be one answer. In a meta where he is picked, Alchemist leaves his mark on GPM records. However, there are other contestants, similar to the last hit average, that can take the crown, such as Naga Siren or Anti-Mage, that make for a good out of the box prediction.
    Most kills in a game and most last hits in a game cover pretty much the same heroes that were already covered in their average counter-parts. You can be a bit more bold in these predictions though, since there is no minimum amount of games played required. If you didn’t believe that Clinkz would make the highest kill avg. list, you can still pick him here. Similarly, if you don’t think Terrorblade will be picked 5 times but he can still outfarm everybody else at least once, go for him.


    Who is going to be the winner of The Manila Major? A good question with so many possible answers. A lot of analysts choose the “obvious” answer, as Team Liquid showed in quite convincing fashion at Epicenter that they are at least currently the best team. It can be argued though that as top favorites, their number will be up. OG recently won DreamLeague Season 5 with ease and placed third at Epicenter, putting up a strong fight against both Newbee and Liquid. Speaking of Newbee, the Chinese powerhouse, despite a few hiccups after Epicenter, demonstrated that they are back to aim for the top and displayed some high level of play in Moscow. If these three favorites are too boring for you, there are always potential upsets from teams such as the wings gaming, who’ve been impressive lately, or even Team Secret, who’ve not won a single official match since Shanghai but could very well perform at this event, much like they did in Shanghai.

    A lot if not all of these categories could be answered with some of the top favorites, but there are some categories that have a lot of “upset” potential. Natus Vincere likes drafting aggressive push and ganking line-ups that can lead to clean and early executions, thus a low death count and possibly the shortest win of the tournament.

    compLexity Gaming is a team that will always keep on fighting and has proven to be a true team through and through, which puts them as a viable candidate for the most assists in a game.

    Fnatic is known to be very diverse in their drafts and carry player Mushi has a reputation of being able to play whatever hero is necessary to secure victory.

    As for the team with the fewest hero picks, it is almost always the team that is likely to be eliminated the fastest and “easiest”. A team that doesn’t get to play many games in the first place. Mineski seems to be the team that most analysts agree on will be eliminated early on.


    The players here are, for the most part, core players from teams that are expected to do well in the tournament. Standout performers, that can decide games by themselves and are known to have high impact and high value performances.

    When it comes to highest kill average and most kills in a game, there are some immediate picks to think of. OG’s Miracle has added Timbersaw to his repertoire of picks enemies will shiver in fear from and even outside of the hero, Miracle has a variety of options to dominate games with. Newbee’s Mu and Hao are both very much in shape right now. There are also other stars that shine in regards to kill scores in a single game, such as MVP’s QO or even compLexity’s Chessie and Empire’s Ramzes.

    Hao is a player to look out for the GPM related records. That said, he’s not a Naga Siren player, so if you are banking on Naga Siren to take the single game record, you could look towards other carry players. Alliance’s Loda comes to mind, Secret’s EternalEnvy or even EG’s Aui_2000. For the averages though, Hao and Miracle are safe picks.

    As for assists, it could be argued that carries are just as viable here or gankers, but there are some high impact supports that should be considered here. The majority of the analysts chose a combination of Liquid’s JerAx and Newbee’s kaka, but there are other strong supports as well. OG’s support duo Cr1t and Fly is worth mentioning here, but also compLexity’s Zfreek or MVP’s Dubu.

    The player with the most different heroes played is another category that applies to a player whose teams advances reasonably far. Mushi has snatched the title in previous tournaments and Fnatic looks to contest the top 8. Nahaz chose to go with EG’s Fear, whereas most other analysts chose OG’s Miracle. It is true that he has played 13 different heroes in his last 20 matches, though Mushi falls behind short by only one (12/20).


    The competition seems to be as fierce as ever with a lot of teams capable of competing for the top. As a result, it seems unlikely that the overall amount of games played is going to be low, as teams will likely play out entire best-of-threes. Any amount below 90 seems unreasonable, as it would mean that at least 22 bo3s would end in a 2-0, or 20 bo3s and a 3-0 grand final.

    The meta is quite diverse and there is a lot of potential for pocket picks and strategies. That said, there are some heroes that are so far out the meta that they seem unlikely even as niche picks. And as always, a tournament will develop its own meta, a meta that teams will adapt to and start picking into.

    Similarly, this limits the amount of heroes banned, thus the overall numbers should be below that of heroes picked.

    All analysts agree on the 81-90 heroes picked and the heroes banned are largely agreed upon to be between 71-80. These predictions seem reasonable, although you could potentially go a bit higher on the picks and a bit lower on the bans if you wanted to go with a bolder prediction.

    As for shortest and longest games, the general consensus seems to be for aa 15-19minute short game and an around 80 minute long game. Going lower on the shortest game seems very unreasonable, but going higher could work out. As prestigious as Manila is, you may not want to call gg so early. As for the longest game, anything between 70-90 minutes seems fair, anything above that is rather extreme.

    With support life becoming a bit easier over the years, high death counts should be quite rare, meaning a most deaths by a hero in a game are unlikely to be above 20. `15-17 seems quite reasonable, but if you feel that some player out there is a notorious “feeder”, then 18-20 could be a good gamble too.

    As for most assists in a game, you have to ask yourself what you think is the hero that will achieve most assists in a single game. And if so, what would the game look like for that hero to achieve that? We see split analysis here, some going for a mid 20 value and some for mid 30. You could go for 28-30 to be in the safe middle.

    Highest GPM, what will it be? There are a lot of heroes out there that are viable enough to set new GPM records, but will it really happen? Most analysts chose to go for 1000+ GPM, but its not unreasonable to assume it’ll be just below that.

    Prediction Videos:

    Nahaz & Noxville


    Purge & Slacks

    As seen on Dotabuff

    Latest articles

    Related articles