Dota 2 hero tier list september 2013

Competitive Dota 2 Hero Tier List

It’s been a few months since we’ve seen a list that evaluates the frequency of picks and bans in competitive Dota 2. I really enjoyed reading both the lists and the discussions that spawned from them, mainly because they focused on the roles and strategies used, rather than the raw strength of each hero.

So I took the liberty of preparing a list with 4 tiers of hero picks that should describe the majority of the trends we have seen in official tournaments since the beginning of the year.

Here is the current list for reference.

Tier 5 – Still to come

• Timbersaw

Of course, I do not claim to be infallible or to watch every match of every tournament and I’ve already confessed to a Western bias. Some heroes may have been judged unfairly or completely omitted but I’m sure that there are many useful remarks to be made and lots of details to be learnt. If there’s anything you think could improve this effort, please let me know.

Undying, Shadow Demon, Darkseer and maybe even Magnus should fall to “tier 2” as of recent weeks of competitive play. I’m also not sure if Chen is a tier 2 anymore with the last nerf and lack of Chen play thereafter.

Instead of giving heroes “Tiers” as DoTa is a much more complex game than just picking random heroes on how often they are picked and putting them into tiers, maybe you should add more meaningful categories such as “Most likely picked (Carry/Support/Ganker/Mid-Solo) in first 3 picks” as that is usually when a team will either pick the “1st Tier” heroes or their core heroes for their strategy. Even putting a hero like Vengeful Spirit in the same tier as Anti-Mage and Tide is painful to see. Night stalker, Naga Siren as tier 3 seems a little far-fetched.

Where is Venomancer? I’m almost positive that Veno is a more staple pick than Vengeful, Visage, or even Chen.

I think it is fine to speculate but I would like to see either. more data? Perhaps # of games and # of times picked/ and heroes used in conjunction (# of games that the same set of 2 or 3 heroes are played together) etc? At this point this looks more like you as a fan’s opinion based on nothing but your own observations based on no data. Dota 2 hero tier list september 2013

For example, Wisp gets zero play in the Asian scene and heroes like gyro/chen/clinkz/enchantress/enigma are almost never picked in the chinese scene while the SEA scene does use enigma pretty frequently.

Another example is Balanar (night stalker), who sees almost zero action in the western scene, but gets pick really often in the Asian scene.

PS. There is no way in hell Void, TS, Furion, SF and windrunner are in tier 3.

I’m not flaming or hating, just giving some constructive reasonings, plz don’t hurt me.

EDit. LGD. int’s Chen is SSS tier btw fyi

It’s just my opinion, but I think tier 1 should be exclusive to these heroes:

magnus
nyx
luna
batrider
dark seer
rubick
lone druid
shadow demon
beastmaster
phantom lancer

heroes that every team will run and heroes that are prevalent across all regions.

edit: i saw that you mostly watched the western scene, which would make a lot of sense. I watched G-league and I can tell you that Rubick would be an absolute tier 1 and gyro and wisp are never picked.

I don’t know if it’s useful to have lists per role because we would be missing the synergy between the heroes that leads to teams picking them. It’s very interesting to have Ezalor being first pick in a game when he is going to be played in the hard support role, leaving the most popular mid and carry heroes behind. His tier 1 position is not only because he is an objectively strong hero (all heroes are OP in Dota, imho), but because he compliments the strategies that are trending right now excellently. The fact that there are 6 staple trilane heroes in first tier and three (two and a half?) mid heroes (Bat, Mag, DS) is useful information to discuss. I definitely agree that the list would benefit from some pairing analysis, or even some 5 hero common appearances stats.

That or concrete statistical analysis on the frequency of picks/bans of heroes with a large sample size. Dota-academy people probably know enough about statistics to do this. Honestly, graphs representing the frequency of usage of certain heroes over time would be a lot more interesting.

Also, regional differences play a significant role in what gets picked/banned.

So basically your ordering of the heroes is based on Dota-Academy’s self-admittedly flawed “metagame relevance” statistic, and the only analysis you did yourself was just putting the dividing lines between tiers? Not even any corrective thought based on how more or less important games are weighted, etc?

You can’t scrape raw DotA-Academy statistics like this without further thought and expect the data to be meaningful. You have to provide some form of your own significant analysis, either directy evaluating heroes, or attempting to correct for known flaws/discrepancies in the data set.

The reason your list is Western-biased is because without any sort of weighting, Western picks are favored solely because there are more recorded Western games. When in actuality G-League games should hold far higher weight than Defense 3 groupstage games between 2 no-name teams.

You can use the Liquid Dota 2 wiki to help get all the hero picks down and dotabuff for match information as well.

No, the list is not just simple dividing based on their numbers. I have promoted or relegated heroes based on recent tournament matches and near future potential.

Then second ban phase they ban strategic picks/counterpicks.

You missed my point. I’m not saying Eastern playstyle are superior. Eastern playstyles aren’t superior, but you have to also be fair to them in giving the OVERALL value of Asian play a similar value to Western play. As it stands, you’re weighting Western play 5-6 times more than Chinese play simply due to the fact that there are 5-6 times as many games being played. Raw Dota-Academy stats don’t account for this. They just report the games played, and it’s up to you to normalize for this fact.

From Dreamhack until now, there have been 279 DotA-Academy-tracked games that could be considered “Western”, 43 games that could be considered “Chinese”, and 66 games that could be considered “SEA”. Do you think it’s reasonable to give those Western games more weight because there’s more of them? If we want to be fair and assume that all regions’ playstyles have equal merit, then you would upscale the value of those Chinese games such that they collectively have similar value to the pool of Western games, and likewise with the SEA games.

You also have to account for the value of tournaments and for the specific point at which a tournament is being played. Weaker teams tend to have weaker drafts. Stronger teams playing weaker teams tend to pick things that are either “standard” or “fun” for the purpose of not revealing valuable information at a point in a tournament when they don’t need to. Playoff-stage games need to be given higher weight than group stages and qualifiers, both because the quality of teams is higher, and because teams are more likely to draft in a way that belies the true strengths of the heroes.

Competitive Dota 2 Hero Tier List

It’s been a few months since we’ve seen a list that evaluates the frequency of picks and bans in competitive Dota 2. I really enjoyed reading both the lists and the discussions that spawned from them, mainly because they focused on the roles and strategies used, rather than the raw strength of each hero.

So I took the liberty of preparing a list with 4 tiers of hero picks that should describe the majority of the trends we have seen in official tournaments since the beginning of the year.

Here is the current list for reference.

Tier 5 – Still to come

• Timbersaw

Of course, I do not claim to be infallible or to watch every match of every tournament and I’ve already confessed to a Western bias. Some heroes may have been judged unfairly or completely omitted but I’m sure that there are many useful remarks to be made and lots of details to be learnt. If there’s anything you think could improve this effort, please let me know.

Undying, Shadow Demon, Darkseer and maybe even Magnus should fall to “tier 2” as of recent weeks of competitive play. I’m also not sure if Chen is a tier 2 anymore with the last nerf and lack of Chen play thereafter.

Instead of giving heroes “Tiers” as DoTa is a much more complex game than just picking random heroes on how often they are picked and putting them into tiers, maybe you should add more meaningful categories such as “Most likely picked (Carry/Support/Ganker/Mid-Solo) in first 3 picks” as that is usually when a team will either pick the “1st Tier” heroes or their core heroes for their strategy. Even putting a hero like Vengeful Spirit in the same tier as Anti-Mage and Tide is painful to see. Night stalker, Naga Siren as tier 3 seems a little far-fetched.

Where is Venomancer? I’m almost positive that Veno is a more staple pick than Vengeful, Visage, or even Chen.

I think it is fine to speculate but I would like to see either. more data? Perhaps # of games and # of times picked/ and heroes used in conjunction (# of games that the same set of 2 or 3 heroes are played together) etc? At this point this looks more like you as a fan’s opinion based on nothing but your own observations based on no data. Dota 2 hero tier list september 2013

For example, Wisp gets zero play in the Asian scene and heroes like gyro/chen/clinkz/enchantress/enigma are almost never picked in the chinese scene while the SEA scene does use enigma pretty frequently.

Another example is Balanar (night stalker), who sees almost zero action in the western scene, but gets pick really often in the Asian scene.

PS. There is no way in hell Void, TS, Furion, SF and windrunner are in tier 3.

I’m not flaming or hating, just giving some constructive reasonings, plz don’t hurt me.

EDit. LGD. int’s Chen is SSS tier btw fyi

It’s just my opinion, but I think tier 1 should be exclusive to these heroes:

magnus
nyx
luna
batrider
dark seer
rubick
lone druid
shadow demon
beastmaster
phantom lancer

heroes that every team will run and heroes that are prevalent across all regions.

edit: i saw that you mostly watched the western scene, which would make a lot of sense. I watched G-league and I can tell you that Rubick would be an absolute tier 1 and gyro and wisp are never picked.

I don’t know if it’s useful to have lists per role because we would be missing the synergy between the heroes that leads to teams picking them. It’s very interesting to have Ezalor being first pick in a game when he is going to be played in the hard support role, leaving the most popular mid and carry heroes behind. His tier 1 position is not only because he is an objectively strong hero (all heroes are OP in Dota, imho), but because he compliments the strategies that are trending right now excellently. The fact that there are 6 staple trilane heroes in first tier and three (two and a half?) mid heroes (Bat, Mag, DS) is useful information to discuss. I definitely agree that the list would benefit from some pairing analysis, or even some 5 hero common appearances stats.

That or concrete statistical analysis on the frequency of picks/bans of heroes with a large sample size. Dota-academy people probably know enough about statistics to do this. Honestly, graphs representing the frequency of usage of certain heroes over time would be a lot more interesting.

Also, regional differences play a significant role in what gets picked/banned.

So basically your ordering of the heroes is based on Dota-Academy’s self-admittedly flawed “metagame relevance” statistic, and the only analysis you did yourself was just putting the dividing lines between tiers? Not even any corrective thought based on how more or less important games are weighted, etc?

You can’t scrape raw DotA-Academy statistics like this without further thought and expect the data to be meaningful. You have to provide some form of your own significant analysis, either directy evaluating heroes, or attempting to correct for known flaws/discrepancies in the data set.

The reason your list is Western-biased is because without any sort of weighting, Western picks are favored solely because there are more recorded Western games. When in actuality G-League games should hold far higher weight than Defense 3 groupstage games between 2 no-name teams.

You can use the Liquid Dota 2 wiki to help get all the hero picks down and dotabuff for match information as well.

No, the list is not just simple dividing based on their numbers. I have promoted or relegated heroes based on recent tournament matches and near future potential.

Then second ban phase they ban strategic picks/counterpicks.

You missed my point. I’m not saying Eastern playstyle are superior. Eastern playstyles aren’t superior, but you have to also be fair to them in giving the OVERALL value of Asian play a similar value to Western play. As it stands, you’re weighting Western play 5-6 times more than Chinese play simply due to the fact that there are 5-6 times as many games being played. Raw Dota-Academy stats don’t account for this. They just report the games played, and it’s up to you to normalize for this fact.

From Dreamhack until now, there have been 279 DotA-Academy-tracked games that could be considered “Western”, 43 games that could be considered “Chinese”, and 66 games that could be considered “SEA”. Do you think it’s reasonable to give those Western games more weight because there’s more of them? If we want to be fair and assume that all regions’ playstyles have equal merit, then you would upscale the value of those Chinese games such that they collectively have similar value to the pool of Western games, and likewise with the SEA games.

You also have to account for the value of tournaments and for the specific point at which a tournament is being played. Weaker teams tend to have weaker drafts. Stronger teams playing weaker teams tend to pick things that are either “standard” or “fun” for the purpose of not revealing valuable information at a point in a tournament when they don’t need to. Playoff-stage games need to be given higher weight than group stages and qualifiers, both because the quality of teams is higher, and because teams are more likely to draft in a way that belies the true strengths of the heroes.