Slardar Guide 7.41f: Build, Skill Order and Matchups

What Changed for Slardar Since 7.39

Slardar’s latest change came in 7.41f on September 15, 2026: Slithereen Crush damage 75/150/225/300 to 60/140/220/300, and base health regen down by 0.25. You can read the full text in Valve’s official 7.41f patch notes. The table below lists every Slardar change in Valve’s patch-note feed from 7.39 to today. The feed lists no Slardar changes in 7.39b through 7.39e, 7.40b, 7.41a or 7.41e.

The pattern is easy to read. The 7.40 changes traded Bash damage for a bigger Crush puddle. The 7.41 rework moved his sustain and damage bonus into a per-level formula and gave Sprint a slow-resistance window. Since then every patch has been a cooldown, damage or armor nerf aimed at the early and middle levels. None of the post-7.41 nerfs touches the level 4 Crush (300 damage at 7s cooldown is unchanged from 7.41 itself), which matters for the damage math below.

Is Slardar Good Right Now?

By public win rate he is average overall and weak at the top. Dotabuff’s last-seven-days All Pick data on October 23, 2026 puts Slardar at 49.23%, a quarter of a point below the median hero, but the gap swings from +2.05 in Herald to -3.60 in Immortal. The table below uses Dotabuff’s heroes table with the rank filter applied to each row, so the median is recalculated for each bracket.

Three things stand out. First, the line is nearly monotonic: the gap shrinks from Herald to Archon, crosses zero between Archon and Legend, and keeps falling to Immortal. Second, the pick rate falls with rank too, from 7.61% in Herald to 4.53% in Immortal, so the players who know the hero best are also the ones choosing him least. Third, Immortal is a smaller and noisier sample than the other rows, so treat the exact 45.98% as a direction, not a verdict.

OpenDota’s rank-bracket data, a separate sample, points the same way. Its heroStats feed on October 23 shows 49.3% overall (127,422 wins in 258,750 public picks), and by bracket 49.9% in Herald (5,893 picks), 49.9% in Guardian (19,989), 50.1% in Crusader (37,027), 49.5% in Archon (54,679), 48.8% in Legend (58,745), 48.2% in Ancient (33,123) and 47.0% in Divine (20,166). In OpenDota’s feed the bracket fields 1 to 7 are Herald through Divine, and the Immortal field is empty, so Divine is the top row we can check there. Two independent sources, one slope. In Turbo, OpenDota shows 52.0% over 110,927 picks (57,672 wins), so the hero does better in the faster mode.

Why the Win Rate Falls as the Rank Rises

The public tables cannot tell us why the slope exists, so here is what the numbers allow and what they do not. We are going to label each idea as supported by data or as our read.

• Supported by data: the nerfs landed on the early levels. The post-7.41 changes cut Crush damage and raised Crush and Sprint cooldowns at levels 1 to 3, while level 4 is untouched. A hero who loses his early lane power and keeps his late power should look better in long games. OpenDota’s duration table agrees: in its sample of 20,823 games, the win rate is 42.7% in the 10 to 14 minute bin (225 games), 49.4% at 15 to 19 minutes, 50.3% at 25 to 29, 52.2% at 35 to 39 and 54.1% at 40 to 44. The sample’s overall win rate (50.9%) is higher than the 49.25% headline, so cite the shape, not the level.
• Supported by data: pick rate falls with rank. Slardar is the 55th most-picked hero in Herald and the 82nd in Immortal on Dotabuff. Fewer high-rank players choose him, and those who do are not winning more.
• Our read, not provable from public tables: his kit is almost entirely single-target and close-range. Crush has a 325 radius, his attack range is 150, and his only gap closers are Sprint and Blink. Higher-rank games punish heroes who need to reach a target. We cannot test that claim from a win-rate table, so treat it as a hypothesis.
• Our read, not provable from public tables: Black King Bar timing. Crush does not pierce debuff immunity in Valve’s ability flags, and Slardar’s whole initiation is a stun. Players who buy a Black King Bar on time take that stun out of the game. Again, a hypothesis.

What this means for you: if you play Herald through Crusader, Slardar is a comfortable above-median pick today. If you play Legend and above, you need a reason to pick him, which usually means a specific draft rather than a habit.

His Kit With Current Numbers

Everything in the table below comes from Valve’s hero data feed, pulled on October 23, 2026. Slardar is a Strength hero, so each point of Strength adds one point of attack damage and 22 maximum health. His level 1 stats are 21 Strength (+3.4 per level), 17 Agility (+2.1), 15 Intelligence (+1.5), 51-59 damage, 4.83 armor, 25% magic resistance, 300 movement speed, 150 attack range, a 1.7 base attack time, 582 health, 255 mana and 2.35 health regen. Facets no longer exist in the current game, so there is nothing to choose before the match.

Aghanim’s Shard (1,400 gold) makes Slithereen Crush apply Corrosive Haze to everyone it hits for a short time. Valve’s data lists the haze duration as 5 seconds. Aghanim’s Scepter (4,200 gold) upgrades Seaborn Sentinel and Crush: the Crush puddle grows to a 400 radius and lasts 5 seconds longer, and the regen, armor and status resistance bonuses in puddles go up as listed above.

Two interactions are worth knowing because they are not obvious from the tooltips. First, Corrosive Haze makes its target leave a water trail behind it for 7 seconds, and Seaborn Sentinel works in trails. Haze a hero who runs away and you get a lane of water to chase him through, with the regen, armor and damage bonus on. Second, the Crush puddle lasts 7 seconds at a 325 radius (the radius was raised from 250 in 7.40), so every Crush you cast also leaves a Seaborn Sentinel zone under the fight for those 7 seconds.

What the Crush Nerfs Cost in Damage per Second

Two patches touched Crush after 7.41: 7.41c raised the cooldown from a flat 7 seconds to 8.5/8/7.5/7, and 7.41f cut the damage from 75/150/225/300 to 60/140/220/300. Here is what the pair does to damage per second if you cast it on cooldown. This is our own arithmetic, damage divided by cooldown, ignoring mana and the number of targets hit.

The nerf is a staircase. At level 1 Crush loses a third of its damage per second, and by level 4 it has lost nothing. Since the standard order puts a skill point into Crush at levels 1, 3, 5 and 7 (see the skill order section), a Slardar who maxes it on schedule is back to full strength at level 7. The first five minutes are where the nerfs hurt. The stun is untouched at 0.8 seconds and the stun is why you take the skill, so the hero still functions, but most of the early lane damage from Crush is gone.

Late in the game the talents change the picture. The level 20 talent adds +125 damage, which takes a maxed Crush to 425, and the level 25 option cuts 3 seconds from the cooldown, which takes it from 7 seconds to 4. Those two are options at different levels (the +125 damage is a level 20 pick, the cooldown is a level 25 pick), so a level 25 Slardar can have both: 425 damage every 4 seconds if mana allows it. At 100 mana a cast, that is the real constraint. Level 25 Slardar has about 51 Intelligence, which is roughly 690 mana before items.

Corrosive Haze: The Armor Math

Corrosive Haze is the skill that earns Slardar a place in a draft. It is a 5-second cooldown, 25-mana, 900-range debuff that lasts 18 seconds, and it reveals invisible heroes while it is on. The armor reduction is -8/-14/-20, and the level 15 talent adds another -3. The question that matters is how much extra physical damage that is, and the answer depends on how much armor the target started with.

Dota’s armor formula gives a damage multiplier of 1 – 0.06a / (1 + 0.06a) for armor a above zero, and 2 – 0.94 to the power of the absolute value of a for armor below zero. The table below applies it to targets with 0, 5, 10, 15 and 20 armor. The percentages are the share of physical damage the target takes. This is our arithmetic from the standard armor formula, not a Dotabuff or Valve figure.

Read the 10-armor row. A hero with 10 armor takes 62.5% of physical damage. Haze level 3 takes him to negative 10 armor and 146.1% damage, which is 2.34 times as much damage from the same attack. At 15 armor the same Haze is a 2.41x multiplier, and at 20 armor it is 2.20x. The relative gain is biggest against mid-to-high armor targets: 2.2x to 2.4x at 10 to 20 armor, against 1.7x on a target with no armor at all. Heavily armored heroes are exactly the ones a team of right-clickers struggles to kill.

The same table explains the 7.41d nerf. Moving level 1 from -10 to -8 sounds small. Against a 10-armor target it moves the multiplier from 100.0% down to 89.3%, which is a 10.7% drop in your team’s physical damage on that target. At level 3 nothing changed, so late-game Slardar kept his reason to exist, while the early picks got worse.

Two practical points. First, Haze is a team-wide buff: it works for every physical attacker on your side, not just Slardar, so the correct time to use it is when your carry is about to hit, not when you are chasing. Second, Haze has a 5-second cooldown and an 18-second duration, so in a long fight you can refresh it on a second target without ever dropping the first. Do not hold it for the perfect moment. Assault Cuirass adds its own -5 enemy armor aura (1,200 radius, Valve item data) on top of Haze, which is one reason it appears in his late-game build.

Talents

Talent names come from Valve’s hero data feed, and the pick rates are from Dotabuff’s Slardar page on October 23, 2026. We cite pick rates only. Dotabuff’s talent win rates sit far above the hero’s overall rate (the talent panel shows figures from 45.6% to 67.4%, against 49.23% overall), so they are not comparable to the headline number and say little about which choice is better.

• Level 10 is a coin flip, and it should be. Half of players take the +30 Bash damage and half take the longer Sprint. The Bash option gives you 30 extra damage on every third hit. The Sprint option lengthens the speed boost from 10 to 12 seconds. With no clear winner in the data, pick the one that fits the lane: more Bash damage if you are trading by hand, more Sprint if your team needs you to chase.
• Level 15: the armor talent wins 64.6% to 35.4%. The -3 Haze armor is a flat addition to a shred that already goes to -20, and the table above shows it moves a 10-armor target from 146.1% to 155.3%, a 9-point gain in damage taken. The +250 health is the safer pick for an offlaner who dies early.
• Level 20: +125 Crush damage at 89.5% is the closest to a consensus on the hero. The Seaborn Sentinel option is the minority pick. It makes sense only if you are fighting in puddles all game, since the bonus applies only in puddles, trails and rivers.
• Level 25: Haze undispellable at 63.1%. The talent exists to stop opponents removing Haze with a dispel, so the default behavior is that a dispel can remove it. The -3s Crush cooldown, at 36.9%, takes a maxed Crush from 7 to 4 seconds. If your team’s damage already depends on Haze staying on, take the dispel immunity.

Skill Order

Dotabuff’s most popular ability build is used in 15.90% of games, and its first seven levels are clear. Slithereen Crush at levels 1, 3, 5 and 7. Bash of the Deep at level 2. Guardian Sprint at level 4. Corrosive Haze at level 6. Dotabuff’s later levels collapse together in the page text, so we do not cite the order beyond level 7.

The logic holds up against the numbers above. Crush gives the stun and the damage, so you want it up first. Bash comes second because a single point (35 bonus damage every third attack, with a 1-second bash) is already a lane tool. Haze at 6 is the standard timing for an ultimate and it is the skill that matters for the first fight.

Sprint at rank 1 has a 33-second cooldown since the 7.41b nerf (29/25/21/17 to 33/28/23/18), so the first point is an engage and escape tool, not a laning one. That is why the standard order spends the early points on Crush and Bash.

Item Build: What the Public Buys

The table below is Dotabuff’s Most Used Items This Week for Slardar on October 23, 2026. Item costs and effects are from Valve’s item data feed. The win rates are in the table because readers ask for them, but they are mostly a measure of game length: players who reach a 5,125-gold item are already winning, and players who still buy a Bracer are often still behind. Compare the match counts and ignore the win rates when you decide what to buy.

The core: Power Treads, Blink Dagger, Shard, Black King Bar

The first four rows are bought in roughly 150,000 to 215,000 matches each, and they are what most Slardar players build. Power Treads give +55 movement speed for a melee hero, +25 attack speed and +10 to the attribute you pick, for 1,400 gold. Blink Dagger costs 2,250, blinks up to 1,200 units on a 15-second cooldown and cannot be used for 3 seconds after you take damage from an enemy hero. That lockout is the reason Sprint stays in the kit: if you were just hit, Sprint is your gap closer for the next 3 seconds. Aghanim’s Shard at 1,400 is bought in 182,415 matches, behind only Blink Dagger and Power Treads, and it gives Crush a free Haze on everyone it hits. Black King Bar costs 4,050, lasts 9 seconds at its first level, has a 95-second cooldown and gives +10 Strength and +24 damage. It is the item that keeps you alive long enough to land the stun.

Early items and where Echo Sabre fits

The Bracer (505 gold, +5 Strength, +2 Agility, +2 Intelligence, +0.75 health regen, +50 health) and Magic Wand (460 gold, 15 health and mana per charge) are the standard early purchases, and they explain the low win rates in the table: they are bought early in games that have not been decided yet. The Echo Sabre (2,700 gold, +20 damage, +15 Strength, +1.75 mana regen) makes melee attacks hit twice and applies a short movement slow on the first strike. Players buy it in 78,555 matches, a little over a third of the Blink Dagger count. For a full breakdown of the item, read our Echo Sabre guide.

Late game: Assault Cuirass, Harpoon, Scepter

Assault Cuirass costs 5,125 and stacks on Haze: it gives the team +5 armor and +30 attack speed within 1,200 range, and takes 5 armor off enemies. Harpoon costs 4,700 and pulls you and your target together by 35% of the distance, always to melee range for a melee hero, which makes it a second gap closer after Blink. Aghanim’s Scepter is 4,200 gold with +10 to all attributes and +175 health and mana, and on Slardar it also lengthens Crush and Seaborn Sentinel puddles. OpenDota’s item sample for the late game, which is small (low hundreds of games), also lists Nullifier (4,350 gold) among the late purchases. That is the item you buy if the enemy has a dispel-dependent defense you need to break.

For a longer discussion of item choice across the roster, our Dota 2 items guide lists every item by category, and our Monkey King Bar guide covers the item that ignores evasion.

Lanes and Positions

Dotabuff’s lane panel shows where Slardar is played. It is overwhelmingly an off-lane hero.

The off lane is the default, with 85.72% of his games, and it is where the rank table above comes from, so read the slope as mostly an off-lane slope. The mid lane has the best win rate and KDA of the three, with 523 gold and 754 experience per minute on Dotabuff’s panel, but it is only 5.19% of his games, so it is a small sample. The lane panel’s win rates do not average to the overall rate (a weighted average comes out above 50%, against 49.23% overall), so read the ordering and the presence, not the exact percentages. The safe lane is the weakest of the three on win rate. We would not play him there by default.

Game Plan: Early, Mid and Late

The level table below is our arithmetic from Valve’s base stats: health is 120 plus 22 per point of Strength (this reproduces his level 1 health of 582), and damage is the base range plus 3.4 per level, since Strength is his primary attribute. It excludes items, talents and Bash.

Early game: levels 1 to 6

At level 1 you have 255 mana and Crush costs 100, so you can cast it twice and not much else. Regen is 0.75 mana per second, which means mana is the constraint on lane harassment, not cooldown. Crush at level 1 now does 60 damage every 8.5 seconds, so you are not winning the lane with damage. You are winning it with the 0.8-second stun that lets your teammates and your own Bash land, and with the 7-second puddle that gives you Seaborn Sentinel (starting values of 1.75 regen, 1.8 armor and 11.4% bonus attack damage while you stand in it). Stand in the puddle when you trade. Haze arrives at level 6 and changes the lane, because from then on each trade can have a -8 armor debuff on it for 18 seconds.

The 7.41f regen nerf (2.35 now, from 2.6) hurts most in these levels, when 0.25 health per second is a larger share of your sustain. Buy a Bracer and a Magic Wand if you are behind, and play to your puddle.

Mid game: Blink and Shard

With Blink and Shard (about 3,650 gold between them), Slardar becomes an initiator: Blink, Crush with Shard Haze on every target hit, then Bash and attacks. Haze is on cooldown for 5 seconds, so cast it on the highest-priority target first and use Crush to spread the effect. Your job is to make sure the physical damage dealers on your team are hitting a shredded target for the next 18 seconds: as the armor table above shows, a level 3 Haze takes a 10-armor target to 146.1% damage taken. Do not use Blink without a Crush ready. A Blink with no stun to follow it is a donation.

Sprint earns its slot in the mid game as a second gap closer. Its cooldown is long (28s at level 2, 18s at level 4), but it passes through units and gives 100% slow resistance for the first 2.5 seconds, so slows cannot stop you during the window in which you are closing the distance.

Late game: Black King Bar and the damage cap

By level 25 you have around 2,400 health from attributes alone and a maxed Crush at 425 damage with the level 20 talent. Your job is the same as in the mid game, with higher stakes. The enemy has Black King Bars of its own, and Crush does not pierce them, but Haze does. That is your reason to lead with Haze in a late fight: it goes through a Black King Bar, so you can still make a 10-armor enemy carry take 2.34 times the physical damage while your stun is blocked. Buy your own Black King Bar so you survive to cast it, and an aura item for the team (Assault Cuirass is the public standard). The duration data above is consistent with a hero who is built for long games: OpenDota’s sample shows a win rate of 54.1% in 40 to 44 minute games.

Matchups

These matchups are from Dotabuff’s Slardar page on October 23, 2026, which lists the best and worst versus heroes for the past week. The percentage next to each hero is Dotabuff’s advantage or disadvantage score, which adjusts for each hero’s overall strength. That is why a hero can sit on the worst list with a raw win rate close to 50%: Terrorblade has the worst score on the list, yet Slardar wins 50.57% of those games. The score is meant to correct for how strong each hero is overall, while the Slardar win rate column is the raw number.

Best matchups

Worst matchups

We will not invent a reason for each line, because the tables do not give one. Two patterns are safe to state. First, seven of the ten worst matchups are physical-damage carries (Terrorblade, Phantom Lancer, Juggernaut, Gyrocopter, Luna, Medusa and Drow Ranger), and four of those seven (Gyrocopter, Luna, Medusa, Drow Ranger) attack from range against a hero with a 150 attack range. Second, the worst-list matches have large samples: Juggernaut (23,148), Drow Ranger (20,054) and Dragon Knight (18,168) have more than 18,000 games each, so these are not small-sample artifacts.

What you can do with the list is draft around it. If the enemy first-picks Juggernaut or Phantom Lancer, Slardar is not where you want to be, and if the enemy has Lifestealer or Anti-Mage the numbers lean your way. Corrosive Haze reveals invisible units, so check our True Sight guide for the full list of ways to see invisible heroes, and our disables explained guide if you want to see how Crush and Bash compare to other stuns.

How to Play Against Slardar

If Slardar is on the other team, these are the points that follow from the numbers above.

• Respect the first five minutes. His early Crush is weaker than it has been all patch (60 damage at 8.5 seconds cooldown), but the 0.8-second stun and the 7-second puddle are intact.
• Do not stand in his puddle. Seaborn Sentinel gives him regen, armor and damage there. The Crush puddle lasts 7 seconds at a 325 radius, and 12 seconds at a 400 radius with the Scepter.
• Armor is a liability against Haze. If your carry is on 15 armor, level 3 Haze takes him to 126.6% damage taken. Know the cooldown (5 seconds) and the duration (18 seconds), and expect the debuff to come the moment the fight starts.
• Black King Bar blocks Crush, not Haze. Per Valve’s ability flags, Crush does not pierce debuff immunity and Haze does. A well-timed Black King Bar removes the stun and the slow, but you still take the armor reduction.
• You cannot hide from him. Haze reveals invisible heroes, so an invisibility plan fails while it is on you.
• Dispel it if you can. Haze is dispellable unless he took the level 25 talent, which 63.1% of Slardar players do.

What the Pro Data Shows

Slardar is not a frequent professional pick. OpenDota’s heroStats feed lists 7 pro picks, 3 wins and 2 bans for him on October 23, which puts him 60th by picks and 81st by bans (the feed does not state the window, so treat the counts as a rough tally). To get something more detailed we pulled the last 100 league games from OpenDota’s hero-match endpoint, covering August 28 to October 22, 2026.

• Record: 39 wins and 61 losses (39.0%), average length 38.0 minutes, average 5.2 kills, 5.5 deaths and 11.1 assists per game.
• By game length: under 30 minutes, 9 wins in 25 games; 30 to 40 minutes, 14 in 30; 40 to 50 minutes, 10 in 31; 50 minutes or longer, 6 in 14.
• Where the games came from: EPL Masters 2026 (29 games), ExitLag ChampZ (18), DreamLeague Season 30 Qualifiers (14) and BLAST SLAM VIII (9). This is a lower-tier sample, with qualifiers and streamer leagues in it, not a top-tier tournament record.
• Most frequent players: Nicky`Cool (5 games, 1 win), Mirele` (5 games, 2 wins) and Noticed (4 games, 2 wins).

A 39% record in 100 league games is a sample, not a verdict, and it is much lower than the 49.23% public win rate. The game-length split does not show the pattern the public data shows (the 30 to 40 minute bucket is the best, and the long games are worse than the public duration table), so we would not draw a tempo conclusion from the pro sample. What we can say is that Slardar is not a hero the professional scene is reaching for in 7.41f.

When to Pick Him in Ranked

Pick Slardar when the draft gives him a job. The three jobs that fit the numbers are these:

• Your team has physical-damage carries who want armor off. Haze is the best skill in his kit, and it is only worth the pick when someone on your side is hitting the target.
• The enemy has invisible heroes or heroes who rely on Black King Bar to dodge stuns but not armor debuffs. Haze reveals and goes through debuff immunity.
• You are playing Herald through Crusader. Dotabuff has him above the median hero in all three, by 2.05, 0.97 and 1.01 points.

Skip him when the enemy has Juggernaut, Phantom Lancer or Luna on the board, when your team has no one to use the armor shred, or when you play at Divine and above and have no draft-specific reason. If you are climbing and the offlane slot keeps landing on you, our Mars guide covers another Strength initiator with a different profile.

If you are stuck at a rank where drafts and offlane matchups decide your games, our MMR boosting guide explains how rank services work, and our safety guide covers the risks. For a comparison of providers, see our best Dota 2 boosting services review.

The Bottom Line

• Slardar is average overall and weak at the top. 49.23% on Dotabuff, 71st of 127, with a gap that runs from +2.05 in Herald to -3.60 in Immortal.
• Every patch since 7.41 has been a nerf. 7.41b (Sprint cooldown), 7.41c (Crush cooldown), 7.41d (Haze armor) and 7.41f (Crush damage and regen).
• The nerfs are a staircase. Level 1 Crush loses 34.1% of its damage per second, level 4 loses nothing.
• Corrosive Haze is the reason to pick him. A level 3 Haze turns a 10-armor target from 62.5% to 146.1% damage taken, and it pierces debuff immunity.
• The standard build is Power Treads, Blink Dagger, Shard and Black King Bar. Buy Assault Cuirass if the team has carries.
• Avoid him into Juggernaut and Phantom Lancer. They have the largest samples on his worst list.

All numbers in this guide come from sources we name inline: Valve’s patch-note, hero and item data feeds, Dotabuff’s 7-day All Pick tables and Slardar page, and OpenDota’s heroStats, durations, item and hero-match endpoints, all pulled on October 23, 2026. Where a sample was small or a figure was a proxy, we said so in the section it appears in. For what else changed in the patch, see our Dota 2 hero tier list and the 7.41f patch notes guide.