The International 2024 Stats and Meta Wrap Up
The International 2024 concluded (and Competitive Season 2024 with it).
It’s time to look back at the meta of this TI, why it clicked, what was ignored (but is likely to be a part of meta soon enough), who performed how and so on.
As always, you can check all the stats in my TI13 reports section and Competitive Tier 1 events of 2024 section.
How it went
Quick note from future me. After re-reading the blog post, it felt way more negative than I wanted it to be. The goal is make some TI highlights and point at some interesting findings in terms of data. While the event didn’t really shine in some regards and had technical issues, the event itself was still a blast and rather enjoyable to watch and experience.
This TI, while there definitely were some exciting games played, there weren’t many new All-Time TI records set this time. While there were some nominations (like 2nd most tower damage dealt by Pure), generally speaking most of them were pretty dull.
Every TI (at least since TI10) was showing off a rather big change, the nature of the changes was always different: be it economy or comeback mechanic changes, map changes or introduction of some other big change, and this had its impact on the economy and stats. Heroes got richer, games were more fast paced, there were more resources on the map and so on.
However, while there weren’t any economy changes, this TI was the first since Facets were introduced into the game. This effectively doubled the number of heroes in the game: while facets are pretty much meant to be different variants of the hero, they more often than not create completely different playstyles, but with an old and recognizable face.
And this alone is a good reason why the games were rather exciting. Yes, we saw Mirana before, and even saw Core Mirana, but we never saw her with a completely different ult (and a completely new playstyle born out of it). Same applies to Sand King, Monkey King, Lone Druid and many others. While we already saw most of these changes in action throughout the year, many people only come back for TI, which made this so important.
Tournament Stats
This TI had the lowest number of total matches (122) since TI1 (and yes, I am counting Group Stage, Seeding Deciders and Playoff as TI). Some of it can be explained by the differences in group stage structure and schedule, some of it can be explained by the fairly low number of participants (the last time we had 16 teams was at TI6). TI12 had a similar format and had 150 matches played, while TI6 also had 16 teams participating and it had 159 matches played.
Not everything went smoothly though. During the Tundra vs IG series 2 game 1 (which also took the longest too complete because of all the pauses) there were three lobby restarts using Replay Takeover feature to continue the game. This was caused by a bug with disappearing abilities (Mirana Leap). The original match (the beginning of the game) is not available in the game client, so it’s impossible to recover a full replay of this game.
In addition, there were also several cases of servers becoming unresponsive shortly after the game ended or losing the match data, resulting in replay availability delays (or match not being accessible using the game client). This happened twice during the group stage, but also became more of an issue during the last two days, affecting 7 matches for about 1.5–3 hours each.
Thankfully, all the technical difficulties were quickly addressed by PGL and Valve crew (thanks to them for their help in resolving the issues when they arised).
Games were generally shorter compared to TI12: median match duration being 37 minutes for TI13 and 41 minutes for TI12. Other parameters are also lower: median XPM being 574 vs 610, GPM being 469 vs 502, average kills per minute being 1.08 vs 1.22. All of these, though, can be explained by the current meta. But some of it might also be the byproduct of the same reason that led to low number of matches.
During this event I was testing my spin on an MVP system, which I wanted implement (first as a “fantasy” system back in 2018). I tried my best to adjust it to give more credit to core players and to be generally more or less fair.
I will write a more detailed blog post about this feature later, when it will be released to the public. But generally speaking I feel pretty good about the results.
Here is the total awards leaderboard. Some players who stand out:
- Ace with the highest total score in general (and also the highest among cores)
- Sneyking with the highest total score among supports
- Insania with the highest total score in his team
I won’t dig deeper into the player and teams records, since there are cool looking images for this purpose.
State of Meta
So remember the part above where I mentioned short matches and low XPM/GPM? Let’s talk about meta now.
Total balance rank of the event was 57.2 (explained here). It was the lowest balance rank throughout all TIs, the closest event in this regard being TI4 (61.2). For context: TI12 had balance rank of 70.2.
And this balance rank similarity is not the only thing similar between TI13 and TI4. This TI was also extremely focused on tempo playstyles and fights, with strong aura items being the biggest focus. Essentially the games went to building auras and fight oriented items ASAP, then shifting to teamfights and objectives.
And the tempo part comes from the level-1 of this meta: Universal heroes as carries. Or rather “ranged carries who hit like a truck” would be more appropriate description (e.g. Solar facet Mirana, Whirlwind Windranger, Eclipse facet Luna, Lone Druid). It seems to be that looking for an answer to the healing meta led the players to gravitate towards the strong damage (that goes well with the heals and auras, but also works well against tank heroes who are usually the aura carriers to begin with).
And level-2 of the meta focusing on initiation, fight reset, with a little big of burst on top of it. Sand King was highly contested even before TI, as well as Doom and Centaur. Same can be said about Naga Siren, though she didn’t really get too many picks to shine. Kunkka became a more surprising addition to this roster, but he fits well for this kind of role.
This was the trend already even before TI13 started, but the actual pool of heroes used quickly shrinked to the most effective ones for each of the tasks.
There were 103 heroes contested, 99 of them being picked (16 of them were not banned at all) and 4 only present in bans. There were 21 uncontested heroes: Arc Warden, Bloodseeker, Bounty Hunter, Dawnbreaker, Juggernaut, Legion Commander, Lich, Lifestealer, Necrophos, Ogre Magi, Outworld Destroyer, Phantom Lancer, Pudge, Shadow Shaman, Silencer, Techies, Tidehunter, Vengeful Spirit, Venomancer, Warlock, Witch Doctor.
Interestingly, Bounty Hunter and Bristleback looked generally great in this meta and still seem to be strong stats-wise at Immortal level, but they didn’t see much play at TI. It’s especially strange that BH was completely ignored, but the hero will probably get attention soon enough.
Same issues can be seen when looking at the best items at this event. Most of them are either good for grouping up and fighting, or very efficient in terms of stats for ranged cores (especially Universal).
As for the neutral items, there isn’t much to say about them. I just like looking at these pretty icons aligned as a big grid. Seeds of Serenity is still incredibly popular though, and Safety Bubble is still stupid strong.
Competitive Season Wrap-Up
And as an ending it would be nice to include some trivia about Tier 1 competitive events played during this season (since TI12 until the end ot TI13, counting only teams with more than 20 matches)
- Team Falcons had the highest winrate of 68.04% over 219 matches, followed by Cloud9/Entity (65.74%, 108) and Xtreme Gaming (61.93%, 197)
- Falcons also had the highest average number of kills (26.9), XPM (3246), GPM (2574)
- The most diverse teams were Team Secret (78.49% hero diversity over 63 matches), BB Team (74.30%, 167) and G2.IG (73.76%, 151) // you can read about hero diversity metric here
- Teams with the longest median match durations are Virtus.pro (44:12), Team Secret (42:50) and BB Team (42:46)
- There were 48 rampages total
- Shortest match: NaVi vs LGD game 2 at Clavison Snow Ruyi, 13:20 (match 7868764939)
- Longest match: BB Team vs Liquid game 1 at BB Dacha Dubai, 1:20:26 (match 7589398098)
- Match with the most average kills per minute: Aurora vs G2.IG game 2 at PGL Wallachia S1, 3.19 kpm (match 7741499600)
- Biggest networth difference: BB Team vs G2.IG game 1 at DreamLeague S22, 90177 (match 7610924034)
- Most total kills: OG vs Tundra game 2 at ESL One Birmingham, 108 (match 7699942711)
Closing words
And that’s about it!
While this is all I have for TI13, I hope to come back soon enough with some cool findings and features I was playing around with.
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.
See ya around!
~Lea
P.S.:
