[SBA]Terrorblade: No Alarms and No Surprises

TerrorbladeYou’ll have to forgive me if my heart isn’t in it today, but it turns out Terrorblade is kinda boring.  Who knew?

Well, probably everyone.   Including me.  When I say “boring,” I specifically mean in the sense that his skill and item info carries no big secret.  You may or may not like his playstyle, but either way you have to admit that there isn’t a lot of nuance out there right now.

Still, I promised this last week, so here it is.  Phoenix will come later this week because he/she/it is more interesting, and so I want a larger sample for that.  Other 6.80 stuff has to wait a bit longer because 25% of my match returns are RAD.

First on the list of total non-surprises is that Terrorblade is another top 20 entry in Farm Dependency so far.  Two interesting trends are that his winrate does improve some in Very High, and he actually might score rather high in Normal bracket farm dependency, but we’ll go into this more once I have a more complete sample for all heroes in 6.80.

Moving on to skills, here’s what we’re looking at:

Q — Reflection: Slows an enemy hero and creates an illusion of them.  Scaling: Significant Cooldown reduction and Illusion Damage.

W — Conjure Image: Creates an illusion of Terrorblade.  Scaling: Illusion Damage.

E — Metamorphosis: Transforms Terrorblade, giving him a ranged attack (with a slower Base Attack Time) and bonus damage.  Scaling: Bonus Damage and Duration.

And is there an emerging dominant build?

TerrorBuildUsageAs you can see, Meta->Conjure(E->W)with one early point in Reflection is by far the most popular build in all three brackets.  The only real significant competitors are Meta->Reflection(E->Q) and Reflection->Meta (Q->E).  And does the build perform well enough to warrant this kind of popularity?


For the most part it does.  There are a few builds that outperform it in certain brackets, but that’s to be expected with how small the samples are for most competing builds.  Only Q->W outperforms it consistently, and in that case we’re looking at a build with less than 1% of the total share.  That being said, the gap isn’t huge between builds that open with Meta and builds that open with Reflection, so you could maybe find some justification to experiment there.  One somewhat interesting feature is that Conjure Image looks like a pretty bad primary, but possibly the best secondary.  It could be the case that the scaling on Conjure Image is actually pretty important for Terrorblade’s midgame farming and general combat presence, but until levels 8 through 12 he doesn’t have the stats and mana to ever justify maxing it over either of his other skills.

In any case, if you’re new to the hero and you’re looking for the safe and extremely common build, it’s 1/2/4/1.  The only significant decision being whether to take Reflection or Metamorphosis at level 1, and for that you need to evaluate your laning situation.

Moving on to items, we again don’t really have any surprises.

TerrorItemsSange and Yasha declining isn’t much of a surprise given its status as official core item.  I’d be more inclined to believe that SnY just delays your Manta Style, but SnY doesn’t perform terribly here.  One thing of note is that there’s not a lot of evidence of SnY disassembles and neither Sange nor Heaven’s Halberd has a significant amount of usage in any bracket.  This list also reminds me that I should maybe look into 6.79 -> 6.80 item comparisons, as Ring of Aquila actually looks reasonably competitive for a cheap item that’s likely to be sold off.  One last item of note, Radiance Terrorblade is a distinct minority.  I can’t say much else beyond that as Radiance winrates are notoriously unreliable.


8 Responses to [SBA]Terrorblade: No Alarms and No Surprises

  1. nine says:

    “Radiance winrates are notoriously unreliable.” oh? why??

    • phantasmal says:

      Item win rates in general have some big issues, such as differentiating between the item pickup as the cause of a victory or the effect of being in a strong position that would have inevitably lead to a victory regardless of your pickup. Radiance can be considered an extreme case of this cause/effect conundrum, as being able to save up 3800 for a Sacred Relic as your initial large purchase is often evidence of having strong laning phase before the Radiance even comes into play.

  2. Suedars says:

    Why not post item winrates as well as their usage stats? It seems like it’d be nice to know whether S&Y’s drop-off is warranted, or if it’s due to a stigma against the item from better players? Is it because item win-rates are very results oriented (You don’t see the guy who lost with just a Basher in Abyssal’s win rate, so Abyssal gets artificially bumped, since it’s rare that losing players will buy an abyssal)?

    • phantasmal says:

      Yes, it’s because win-rates can be misleading in a whole host of ways, and I prefer not to emphasize them unless something unusual is going on. SnY’s win rate is reasonably in line with Manta Style given the latter items greater expense, but in my opinion there is not enough there to draw a conclusion one way or another.

      For the sake of comparison, I brought up Blink Dagger’s win rate on Legion Commander because it was distinctly outperforming more expensive items that it would likely be competing with for initial big purchase on the hero. That kind of pattern between price and win rate is rather unusual in my experience, so it warranted the attention more than the Manta vs SnY comparison here which is significantly less conclusive.

  3. New Meta says:

    What is the MMR equivalent of your skill brackets ?

    • phantasmal says:

      I don’t have enough information to make that comparison. Some people have talked about being in the High bracket previously and having an MMR somewhere around the upper 3000s, but that’s me relying on hearsay and second-hand information.

      Based off http://blog.dota2.com/2013/12/matchmaking/ my estimate is that if you were VH in old Normal matchmaking, your initial ranked MMR would be at least 4000. Normal->High would be ~2800.

  4. ez says:

    in high skill bracket plays use TB primary Q to counter heroes such as AM and PL, and TB has to build W to stack/pull/lane/scout which secure his farm. That is why primary Q shows a decent winrate

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: