Reworking Treant Protector Back Into The Meta
Before patch 7.01, Treant Protector only had one major nerf in four years, dating back to patch 6.75. During that period of time, he’s received buffs over 18 patches, each one slightly tweaking his spells and his base stats. Living Armor has had its numbers adjusted from its health regen, mana cost, and cooldown reduction over a dozen times. And yet, Treant has continued to drift in and out of the pub and pro meta.
The Buffs Before The Rework
Bashlord Treant before his patch 7.01 lvl 25 talent rework
In patch 6.88f, 
Treant Protector was picked in 3.5% of games. Historically, Treant Protector was an unattractive hero. No matter Valve’s efforts to buff him to make him more usable and appealing, without being overpowered, he’s been among the least picked heroes in Dota’s pubs (right between Meepo and Bane). He did have a short period in the competitive spotlight during TI4 NA’s qualifiers, but that quickly faded by the main event. He’s more popular for being EG.ppd’s niche pick, who from our record has played 98 matches on Treant, with a 73.5% win rate.
His unpopularity in pubs was a sign that he just wasn’t that fun to play. Relegated to Living Armor babysitter and jungle scouter, his style was too passive for most people. By the time he had Overgrowth, he was still just a one hit wonder. That didn’t change, even as Valve tweaked the numbers from all his skills. Valve could continue the buffs to force him to be relevant and usable, but it would’ve been covering up his inherent design flaws.
Patch 7.00’s gave Treant Protector an active role in both the early and the late game. The bash from Nature's Guise made him a more dangerous laner when coupled with Leech Seed, and the increased move speed not only helps with his ability to rotate around the map, but compensates for how sluggish it felt to play him. Pub players have noticed as well, and his pick rate has since more than doubled to 7.5% in patch 7.01.
Buffing Through Talents
The new talent trees for every hero gave Valve a more concerted way to target buffs and nerfs from now on. For example, patch 7.01 targeted Nature’s Guise, but instead of only tweaking the numbers (rescaling the bash duration), Valve also removed the level 25 talent that buffed the bash (20% chance of +2 second stun on bash) from the skill.
Patch 7.00’s talent trees mostly benefited supports and gave them late game viability with level 25 talents usually augmenting one of their skills. Treant Protector’s level 25 talent increases Living Armor’s efficacy by 71% (from 7 to 12 block instances), while his cooldown reduction cuts Overgrowths cd from 70s to 59.5s. But pubs seem to be picking everything wrong when it comes to choosing talents for Treant Protector.
Treant Protector’s talent usage rates
Pub players are overwhelmingly favoring attack speed, gold/min, and respawn time, when the alternate choices result in higher win rates. +30 attack/speed seems like a natural synergy for Treant Protector’s slow, high damage base attacks, but it’s at the cost of a significant +2 mana regen. Yes, Treant Protector can also get a +65 damage talent later on, but his impact on the game won’t be from dealing more physical damage. Attack speed won’t affect how often he’ll bash out of Nature’s Guise, so it’s best to optimize his spellcasting for his games.
Passive gold/min gain is an alluring choice for many supports, but for most of them it’s not worth the cost. Ogre Magi and Disruptor lose 100 cast range, and for Treant Protector, he would lose +25 move speed, which also is buffed by 34% from Nature’s Guise. Again, more gold can be effective, but being more dependable to initiate in team fights is better.
The last error Treant Protector players make is opting for -50s Respawn Time over +5 Living Armor Block Instances at level 25. The win rate difference is a significant 5.3%. Across the board, talents reducing respawn times result in lower win rates. One explanation could be that players in losing situations feel like they need a shorter respawn time, but it’s a perspective that is self defeating. In Treant Protector’s case, his alternate talent in Living Armor allows himself or his teammates to survive longer in fights. It’s better to die less than to shorten the death timer when you’re dying more.
Is Treant Ready For Captain’s Mode?
One poorly kept secret is that Treant Protector is absolutely dominating in higher skill brackets. In 5k+ MMR games, Treant has the #1 win rate out of all heroes and is picked 17.25% of the time, more than twice as often across all pubs. For the broader, “Very High,” skill bracket, he has the 3rd highest win rate and is picked 14.47% of the time. His success in the upper tier of pubs can be an indicator of his success in pro games. It may be why Valve is still tweaking his skills in patch 7.01, and why he remains out of Captain's Mode. But even if he's not there yet, he's at least still being played in pubs, now more than ever.
first yay boy
1st kappa
His ulti used to do damage without needing aghs. And he was popular then. I guess that one nerf killed him =/
.......
5th fuck my life
let's not forget that a treant with a blink dagger nowadays is something you don't want around you. not because of the OD, but the bash. so hard to jungle with him around.
@dave.nd agreed. especially with the fade time, Treant can blink out of the trees to initiate with bash.
"The last error Treant Protector players make is opting for -50s Respawn Time over +5 Living Armor Block Instances at level 25. The win rate difference is a significant 5.3%."
I think this is an error in judgement. It seems far more likely that the reason living armor +5 has a 5% increase in winrate is because a level 25 tree that is already winning the game is more likely to pick it. A tree that hits level 25 but is losing the game at the moment he has to choose is going to pick the -50 respawn because he is dying more often.
the other talent discrepancies can have similar explanations or your conclusion can certainly be correct. A tree that is actually playing support as he should instead of solo farming jungle to rush aghs probably has a higher win rate, but there can be other factors.
Regardless tho, I think your dismal of the level 25 respawn talent is short sighted.
some still play Treant very defensively
@Hopeless
He addresses the fact that people choose it when they are behind, and talks about how it's backwards thinking. If you are behind, you need to win a teamfight, and a lower respawn time won't make you stronger in the now. It will just make you die more often. Get the respawn time when you are ahead, because it closes any window the opposing team has from winning a fight. When you are losing, you are food for the other team, but when you are winning, you are a problem for the other team.
And I think using this logic he is right that people are making an error in opting for the respawn reduction. It's not saying the talent is bad, it's saying that it is chosen in the wrong situations, and it is enough to show up in the statistics.
Look at what icefrog did:
Jungle nerf: Now has 70% resources compared to 6.88.
Added 2 new jungle ganker: Monkey and Tree
Farm blade and iron talon nerf
Added bully ancient creeps to make stack farming a nightmare.
High ground advantage for jungle gankers
I have 114 matches with treant and a 64% winrate, majority of the games before patch 7.00.
The issue with treant is, in low skill brackets people don't look at other lanes. His living armor skill can turn early ganks or close 1v1 fights on mid lane conpletely around, if you manage to be aware of what happens on the map. most people in lower skill brackets dont have this awareness.
the longest win streak i had with treant was 7 won matches in ranked, he always was a very good pick, but people can't play heros that need map awareness or micro
Lvl 25 treant doesnt die at all...
Treant got me to 4.3k mmr before I got stacked in 3.9k mmr. 400 mmr in a week. super powerful in pubs. Treant is great against team with high physical damage carry, ( legion, slark, slark, void, juggernaut... etc. ) because of his living armor. Even if the other team counter wards and put sentry it is hard to kill treant if you play it right. ( 71% win rate - 66 games ). You'll justhave hard time with slardar and gondar.
^ got from 4100 to 4500 in a week, treant is love
lol good reaD'
With a rush on Aghs, he's also a great counter to MK... In my low mmr rank, people still think that low ground wards see into treetops.
Why perma stun is considered op? Treant was ok!
Always enjoyed playing treant, his harrass potential is great, he hits like a truck early game so squishy supports are so easy to push out of lane. people also underrate the ability to get fb. Any slow to medium speed hero with no escape can be ganked easily with leech seed and orb of venom.
I saw this before and would like to say it again:
Lower re-spawn abilities do not result in losing a game.
A game being lost - fighting from behind results in picking the lower re-spawn ability. Would be great if an author considered this before saying that pubbies are doing it "all wrong."
@$nu
The author is not saying that it is a worse talent, but he is saying that it is picked more often and has a lower winrate. He then goes on to theorize that it has a lower winrate simply because it is picked too often in situations where it is a poor choice. So yes, the author is saying the pubbies are doing it 'all wrong' but he clearly addresses the exact thing you think he doesn't, and explains why picking respawn in a losing situation is a bad idea.
@$by
The author did consider it.
"One explanation could be that players in losing situations feel like they need a shorter respawn time, but it’s a perspective that is self defeating. In Treant Protector’s case, his alternate talent in Living Armor allows himself or his teammates to survive longer in fights. It’s better to die less than to shorten the death timer when you’re dying more."
treant is love, treant is live <3
now Treant Protector is maniac
авторы такое ощущение совсем свихнулись , не в первой статье говоря о никчемности уменьшения возрождения из таверны .
I am Groot
About lvl 25 Talent:
50 Secs is a lot! I do think it has its purpose, however i think the way to pick it is counterintuitive which is why without having a huge spotlight in the community through pro intervention we will get it wrong.
Most here say its about being behind when the talent is picked and judging from this i guess yeah thats when its picked conversely the lower winrate. However how about it also being wrong in that case. Heres my line of thought:
50Second respawn time is a niche pick for a level 25 talent and provides MORE benefit in a situation where your team is already ahead by a good margin. Here this talent can act as a throw safety net in case your team really messes up. Whereas if you are behind already or its about equal youd need the living armor charges to even give you a shot at winning a fight / defending the base.
just nerf tat stupid monkey dogshit attack range/ QR skill /50% lifesteal
Does anyone still go for the combat talents instead of mana regen and spell cooldown? i'm not sure how useful mana regen is if i have tranquil/arcane boots.
These analyses seem to me to be lacking. You can't just look at the disparity of win-rate and assume that picking the right talent will increase your win rate. For example, as 'Hopeless' points out, it may be that those doing poorly pick the respawn time talent precisely because they're playing poorly. In this case you need to check the winrate of +5 living armor vs -50 respawn time among those games where a team is performing poorly prior to the pick (or alternatively, check among matches where team was doing well). That's just one factor you need to control for.
Another factor is the overall skill of players. Is the win rate for each particular talent pick the same among the highest skill players as it is among the rest? Is it different for the highest skill players that were doing poorly prior to each talent pick vs those who were doing well? Which talent is best to pick may depend in part on what skill bracket you are in.
There's just too many factors that aren't being controlled for here.
@Brock Hall: 'Hopeless' was pointing out that the information we have does NOT show that -50 respawn time turns into a lower win rate pick than +5 living armor instances in situations where a team was doing poorly already. The justifications given in the article are irrelevant -- they *assume* that the win rate of games where a team is doing poorly and picks -50 respawn is lower than picking +5 living armor. But the stats don't show that -- basically, we don't know if that's the case, so trying to give an explanation for why it is the case is one step too early. Dota Buff has a ton of data, the author doesn't need to guess what's the case -- they can check (assuming the data can be mined easily).