The International 10 Teams Overview

Lea Spectral
16 min readSep 24, 2021


Yes, it’s this time of the year. There are just about two weeks left until The International 10 finally happens, and it’s time for my Pre-TI teams recap, which became a tradition at this point.

Версия на русском

The rules are simple: for every TI-attending team there will be a small profile, covering their general stats and most important heroes, as well as “meta graphs” and the team’s general playstyle. It’s more like my own interpretation of what I noticed about the teams, so take it with a grain of salt.

Data used is based on the last 5 months of matches, starting with DPC Season 2. I won’t go into much details, but you can check stats in my Pre-TI10 report.

As said before, every team has a small profile (this time I decided to make some cool graphics instead of writing a lot of text about most played heroes and stats of the teams), as well as a “Meta Graph”: it’s a graphical representation of the most picked pairs of heroes of the team. Colors of edges and heroes represent winrate (blue is closer to 100%, orange is closer to 0%, grey is about 50%), while the size of a node or an edge represents the number of games (bigger node or thicker edge means more games).

Since there will be more graphical content this time, I would like to write less numbers (so there won’t be much about hero winrates and numbers of matches, since you will be able to see it yourself), but instead I’ll try to cover general playstyle of the teams.

There are some exciting things to talk about, too.

First of all, there were a lot of changes to the Stats Hub since TI9. Some of the coolest additions to the teams profiles are inclusion of lists of unique heroes played by every team member, as well as the list of team specific heroes aka “what this team picks more than everybody else” (most important bits of these will be used in this article, here I called this bit “special heroes”).

Aside from these features, teams profiles also have a more detailed version of Picks and Bans table, and the profile overview has the list of best and worst heroes of this team, as well as against it. I won’t go through it here, but you can explore it yourself.

Second, a great friend who’s working on Dota2ProTracker prepared his own TI10 related section. It has details about recent ranked performance of every TI attending team. I recommend you checking out his TI10 hub.

Aside from that, I will be mentioning TI10 Monte Carlo Simulations by Noxville. You can check them out here.

Third, I am very excited to announce that I will go to TI10 as a part of the English Broadcast Talent crew. It wasn’t mentioned in the initial announcement, but you could notice me tweeting about it (or saw a retweet about it from other talents).

And the last thing. I finished working on my Complete The International All-Time stats, starting with TI1. I covered some of it in my previous article, which I think you should check out too!

A note from future me. You might notice this “hero diversity” metric that often goes over 100%. Initially it’s supposed to represent the percentage of games with unique heroes on average (meaning how “diverse” picks of this team are). Percentage-based scale doesn’t really work that well here, and this metric in general seems a bit weird. But in general if a team has more than 100% diversity, then it gets more than one unique hero per game on average. Sort of. It’s more complicated than that, but that’s the jist of it. I noticed it a bit too late when writing this text, so I decided to leave everything as is instead of thinking of a better solution.

Now, let’s go through every TI10 attending team!

Traditional winrates of the teams against each other

Vici Gaming

Vici Gaming are taking a lead with the highest number of games played, while also placing 7th in DPC ranking, securing their direct invite to TI10.

While it’s hard to call Vici Gaming a “slow” team (their average game duration is not even that high, compared to other teams on the list), they sure love to play the long game. They have 4th highest average GPM and 3rd highest average XPM among the TI10 attendees.

It can be noticed in their heroes list too: half of poyoyo’s most played heroes are famous for being traditionally “farm heavy”, and the other half can be played this way. Same applies to Ori, and even eLeVeN’s heroes seem to follow this trend, with Timbersaw and Doom taking first two places.

However, it doesn’t mean their playstyle is passive. While the team’s average tower damage per minute is one of the lowest (which might be because of the team’s love to play a longer game), their average timings for every early objective (first blood, first Roshan or tier 1 towers) is up to a minute earlier than average.


PSG.LGD are probably the team people expect to win at the moment. Placed 2nd at TI8 and 3rd at TI9, this team still stands strong, occupying the second place in DPC ranking.

During the last couple of events we weren’t able to see stellar performance of NothingToSay, since legendary xiao8 took his place as a temporary stand-in. And while the team has the third highest winrate among TI10 contestants and the highest chance to win in Noxville’s simulations, we are yet to see the team’s performance at TI10, and it might be different from what we expect.

PSG.LDG takes the second place by the total number of matches, having just 4 matches less than Vici. Their playstyle, however, is drastically different. LGD already have the 5th highest tower damage per minute and 4th highest hero damage per minute (with pretty average game duration), but the lists of special and favorite might say even more in this case.

While the list of most played carry heroes explains the average game duration (which one might expect to be lower than average with an active playstyle), but you might also notice a lot of position 5 Enchantress. All the team’s favorite heroes favor fast and action packed games, too. And pretty much every player, judging by their playstyle, favors a very aggressive fighter playstyle.

Team Aster

Team Aster is another strong team from China, who finished DPC season with the 9th place. This number looks like a curse for the team: they finished 9–12th during the Singapore Major, then during AniMajor, and now by Noxville’s simulation Aster is most likely to finish 9–12th.

This team often doesn’t get as much attention as other teams from the region. Their playstyle seem very similar to PSG.LGD, but Aster seem to be a bit more focused around brutal and straight forward manfighting and sustain (like Troll or Huskar).

But the essential part of Aster’s strategies is LaNm: it’s safe to say that whole team plays around his heroes.

Another thing to note: Aster’s picks are a bit more exotic. You don’t usually see a team pick Pudge or Pangolier so much, they get into the team’s favorite pairs.


Elephant got to TI10 through regional qualifiers, and they are generally considered the weakest Chinese team at the event.

And in a way it shows through their drafts. Elephant like to play longer games an pick farm-heavy and scaling heroes, but at the same time they also tend to pick early fighters and like to play with heroes and strategies that are easier to execute (but also can’t be as strong).

It seems like Elephant’s general idea is to get enemy’s safelane tier 1 tower very early and invade enemy’s jungle, while also trying to get other tier 1 towers early enough to open up the map for farming. It’s not a special idea by itself, but the most important part is “very early”: mean time for getting enemy’s safelane tower is almost two minutes lower for Elephant than overall average.

Invictus Gaming

IG finished DPC season with the 5th place. While the team showed stellar performance at the Singapore Major this year and even got to AniMajor, they had a bit of a drop in their level of performance. Although, they’ve got 3rd places at i-League season 2 and OGA DotaPit season 5, and it might mean they are slowly getting back their form in preparation to TI10.

In terms of playstyle IG seem to favor flash farming heroes and early game damage dealers with burst. Considering the team has 3rd shortest average game duration, it’s clear IG favors faster games.

With their priority for flash farming, it’s not surprising to see huge teamfight potential as well. This can also be seen in IG’s tower damage per minute: they have the lowest value of this metric while having short games. This doesn’t mean IG don’t push, but rather shows their priority on winning fights first.


T1 got to TI10 by securing 6th place in the DPC ranking. And while during the first DPC season their performance didn’t stand out as much, they started to shine later on, ending season 2 as the first place in SEA upper division, third place at AniMajor, 1st place at ESL One Summer 2021 and 4th places at BTS Pro Series 7, OGA DotaPit Invitational and ESL One Fall 2021. It’s also worth mentioning, that while they lost in the latter two events, they were able to hold their ground against Spirit and LGD.

T1 love to take it slow. They have the longest average game duration, and their favored heroes tend to be either lategame oriented or generally favor a slower and patient playstyle.

Another thing to note (and probably more important one) is their priority to play reactively, either by scouting opponents (Nyx, Enchantress), resetting fights (Warlock, Wyvern) or counter reacting, using enemy’s attack to their advantage (Kunkka, Lion, Axe, Magnus).


Fnatic got to TI10 through SEA qualifiers. Sadly, they were not able to get to AniMajor, getting 4th place in DPC Season 2, but showed results good performance afterwards (arguably even comparable to T1) during BTS Pro Season 7 (2nd place) and OGA DotaPit Invitational (5–6th place, lost to Spirit).

Fnatic, like T1, have very long games on average (second highest average game duration), but their playstyles are hard to call similar. Fnatic love to take initiative into their own hands, emphasizing map control and taking advantage over enemy’s movements to be the first to attack.

Fnatic love to play with heroes with flashy spells and initiators, and to make it easier for them to play, the team also tries to get all tier 1 towers as early as possible.

Evil Geniuses

EG ended the season with the 1st place in DPC ranking and didn’t get 3rd place (or lower) even once this year. EG in it’s current shape is expected to compete for the first place at TI10.

EG have two major themes going: very strong emphasis on push (with 4th lowest average game duration and 2nd highest tower damage per minute) and mobility. Although there still are plenty of heroes who aren’t exactly the most mobile, but this aspect tends to be compensated by somebody else on the team.

Quincy Crew

Quincy Crew finished DPC season with the 4th place. They ended Season 2 with the first place and played at AniMajor as well as ESL One Summer 2021, but sadly didn’t end up getting any impressive results.

Crew’s playstyle is similar to EG in terms of speed, but two things that stand out to me stats-wise are QC being first by first pick winrate (I’d like to think that it shows the power of their drafts) and 3rd highest hero damage per minute.

Generally Quincy Crew like to play with bulky cores with tons of survivability, as well as healing and any other kind of sustain. This also explains why they tend to ban Terrorblade and burst heavy heroes more than everything else.


And the last team from North America, Undying, secured its spot at TI10 through qualifiers. Aside from that, the team managed to get 2nd place at BTS Pro Series 5 and 1st places during Pro Series 6 and 7.

Despite having the highest winrate compared to other TI10 attending teams during the last 5 months, Undying’s TI slot was a surprise to many. And to this day the team is not generally favored due to lack of any experience against international teams. The only time we saw Undying playing against teams from other regions was OGA DotaPit Invitational, where the team got 7–8th place. However, Undying is often considered to be the same level as Quincy Crew.

Playstyle-wise UND seem to prefer action packed and aggressive Dota. Their healing per minute, hero damage and tower damage per minute values seem to be pretty high, too, but I expect it to be a side effect of them winning most of their matches.

Generally they prefer to play with a mix of sustain and high damage push, with a slightly more focus on sustain, as well as a tendency to prolong fights and either reset or counter initiate.

SG esports

SG esports got to TI10 through South American qualifiers, but sadly didn’t have much notable achievements aside from that.

Playstyle-wise they prefer what I would call a “controllable chaos”: a lot of initiation and forced non-stop action that helps them create their own way to victory.

Another important thing to note: SG is one of the teams that has a midlane playmaker, and their actions reflect that. They really like to take an early mid tier 1 tower generally play around midlane and leave more hero choices for 4dr, while often getting position 1 and 4 heroes during the first draft stage.

Thunder Predator

Thunder Predator, unlike SG esports, got 12th place in DPC ranking, thus securing their place at TI10. However, these two teams are often compared to each other because of their similar performance.

Their playstyles might look similar at the first glance too, but while SG’s game durations are mostly average, TP’s average game and win durations are the shortest compared to every other TI10 team. Adding more to that, Thunder Predator don’t seem to care about any specific objective early on (or care about objectives in general) or map control.

If SG’s playstyle I would call “controllable chaos”, in case of Thunder Predator it’s more like “brawl and madness”. They love to play with bulky brawlers and fight to the last breath.


Beastcoast finished the DPC ride with the 11th place, but they sure got the first place in the hearts of the community.

That said, there isn’t much to point out about BC’s playstyle. They generally push slower than everybody else and are late with almost every objective on average.

But what I found the most interesting is the fact that Beastcoast tend to play “4+1” lineups, picking utility mid and playing around K1. This is also why we tend to see a lot of flashy spells and beautiful fights from the team. It’s part of the reason why their playstyle works so well against teams like Thunder Predator, but it becomes harder to win against teams who know how to play patiently and control the map, like VG or T1.


Alliance finished 10th in DPC ranking, but didn’t really have much success at any major international event.

The team’s playstyle tends to have a strong emphasis on sustain and fight reset. But what I find more important is their midlane hero leaning more into utility direction.

It’s hard to say Alliance truly plays a “4+1” game, but LIMPP and s4 are definitely the playmakers of the team, creating space for Nikobaby to do his job.


Just like with TI8, OG got to the event through qualifiers, while not showing much at any tournaments throughout the year. However, since last TI the team changed two players: Ana and JerAx were replaced with SumaiL and Saksa. It’s hard to expect anything from OG right now since we haven’t really seen this roster perform yet (not counting qualifiers).

With all the changes OG’s overall playstyle, strangely enough, remained generally the same. SumaiL gets the most flashy picks, while Saksa acts as the main playmaker (practically filling JerAx’s role).

The most notable thing about OG would be their picks. The team’s players are experimenting with different heroes in their ranked games, and even during TI qualifiers we have seen some unexpected strategies, like offlane Alchemist, carry Axe (after all the nerfs), offlane Monkey King.

Team Secret

Team Secret ended up with the 8th place in DPC ranking, while also showing worse performance than expected in the second half of the season. Despite that, Secret are still one of the main favorites of TI10, considering their great performance throughout last two years. One might even say they would’ve won TI10 if it happened in 2020.

However, Secret’s ranked stats (as well as the fact they skipped latest events) clearly show how hard they are preparing to TI10.

The main staple of Secret’s playstyle is “greedy” mid: Matumbaman might be the carry, but the true position 1 of the team is still Nisha.

One thing to note about Secret’s drafts: they love to pick gank-heavy lineups, looking for ways to pick a hero off and quickly burst them down.

Team Spirit

Team Spirit got their TI10 slot through qualifiers. And while they didn’t get quite enough points to get a direct invite, the team not only showed great performance throughout the year, but also improved a lot, and even ended up competing against PSG.LDG in finals of OGA DotaPit Invitational.

Despite longer games on average, Spirit loves to play with active drafts that are able to fight early. They also tend to give a lot of attention to Roshan. Aside from that, generally I would describe all of the trends noticeable in Spirit’s drafts with just two words: balanced draft. finished DPC season with the 3rd place in the ranking, while also being one of the favorites at TI10. However, it’s also the team many feel cautious about: this team, just like its players, is very young and lacks patience, experience and guidance.

But it didn’t stop VP from getting second highest winrate, while also having the second shortest average game duration compared to other TI10 teams.

Similar to Spirit, VP love to play active heroes with an early online, but they also have a stronger emphasis on gank potential and pickoffs. And it’s all while always having sources of consistent and fast tower damage.

Instead of taking it slow, VP prefer to take initiative and maintain their momentum, while also taking objectives. This, in turn, allows them to perform well against teams like PSG.LGD and taking a lead against them.

That’s pretty much what I had to say about the teams.

Does this mean the teams will play like this? No, they might adapt to the meta and perform differently. And in case of teams like Secret we barely even know how they will play.

Do I think I’m right? Hell no, my judgement of the teams’ playstyles is likely to be wrong in one way or another. However, you can still take a look at stats or graphics and make your own mind.

One thing is certain: there are a lot of strong Chinese teams this year, so there’s a high chance one of them might take the Aegis. But it doesn’t mean there’s no competition: every team is strong enough to compete at TI10 and win.

It would be great to take a look at the meta of 7.30c and make a prediction about what’s coming, but I don’t really think that’s going to work. Just like every time with TI, meta is going to evolve in unpredictable ways, and the only thing we can do right now is wait.

Some things can be noted in ranked stats of the players. There definitely is going to be a lot of Sven and carry Tiny, since these two heroes are generally considered strong. You can also notice a lot of experiments with Faceless Void, Void Spirit, Gyrocopter, Clinkz, Silencer, Monkey King and Queen of Pain. Some even expect to see a lot of Spectre, but personally I doubt it.

But in any case, that’s about it.

I guess the next time we meet is going to be after The International 10. But I’d suggest you to watch TI10 stream in the meantime ;)

Until then, you can follow me on Twitter or join my Discord. You can also make a donation here if you like what I do and would like to help, I really appreciate that.