Valorant Split Map Guide 2026: Why Vertical Control Wins (Rope Timings, Callouts, Best Agents)

Split’s defenders win 53.0% of rounds against attackers’ 48.1% [5] — one of the larger sided imbalances in Valorant’s current map pool. Most guides treat that number as a footnote next to a callout list. It shouldn’t be one. Split was the first map built around rope ascenders [4], and the teams that win consistently on it are fighting for vertical real estate — Tower angles, Rafters, mid ropes — instead of just pushing harder through the ground lanes the way you would on Bind or Ascent.

This guide covers the full callout list, the actual mechanism behind that defensive win-rate gap, a repeatable rope-timing protocol for moving through mid without dying to it, and the agents genuinely winning on Split as of July 2026 — not whatever a generic tier list recommends. New to the game entirely? Start with our Valorant Beginner’s Guide 2026 for economy and rank basics before diving into map-specific strategy.

Quick Start: How to Win More Rounds on Split

1. Learn the three primary ascenders before anything else. Two connect Sewer to A Lobby, one connects B Hell to B Tower — those three ropes carry most of the map’s rotation and flank traffic [2].
2. Check Tower before clearing ground floor. A Tower and B Tower control the sightlines into both sites [2]; peeking them first stops you from getting picked from above mid-execute.
3. Default to mid control on defense, not a full site stack. Rotating through mid is faster than the long way around, and it lets you catch a late attacker rotation from behind.
4. Never climb a rope at running pace into a contested area. Rope movement generates a clear audio cue [7] — see the timing protocol below before using one blind.
5. Verify the map’s still live before trusting any \”best comp\” you find elsewhere. Split has cycled in and out of the competitive pool before [4] — confirm it’s active on your current patch first.

Split Map Layout, Sites, and Callouts

Split sits in Shinjuku, Tokyo, codenamed \”Bonsai\” internally, and was one of Valorant’s four launch maps in April 2020 — the first to introduce rope ascenders as a core mechanic [2][4]. The map runs two bomb sites separated by a raised, narrow mid, connected by seven total ascender locations: two linking Sewer to A Lobby, one linking B Hell to B Tower, and four more spread through the Vents system [2][4].

A Tower and B Tower are the two structures worth memorizing first. Both dominate the elevated sightlines into their site [2], and losing either one early usually means losing the information war before the spike is even down. Riot’s own map history backs up how much weight those two structures carry: Patch 6.0 specifically flattened A Tower’s back section to make it less punishing to contest, alongside widening A Main and removing A Rafters’ problematic under-over angle — a direct acknowledgment that the ground-vs-vertical balance needed adjusting [4].

Why Vertical Control Wins More Rounds Than Horizontal Pressure

Split’s 4.9-point defensive win-rate gap [5] isn’t a fluke of tight chokepoints — it’s the direct result of who controls the ropes and towers, not who pushes hardest through Main.

Here’s the mechanism. Split’s ground lanes — A Main, B Main, mid — are narrow enough that a horizontal push through any single one funnels straight into a chokepoint the defense already holds crossfire on. Vertical routes bypass those chokepoints entirely, but only the side that owns the rope in that moment benefits, and defenders start every round already sitting in the vertical positions [2]. Attackers have to actively contest space they never had to begin with, on a clock, with utility they’ve already partially spent getting through Main.

Riot’s own balance history confirms this reading rather than contradicting it. Patch 6.0 deliberately widened A Main’s opening engagement, stripped A Rafters’ under-over angle, and flattened A Tower’s back section specifically to make horizontal attacking easier [4] — an admission that the map’s vertical geometry was tilting rounds toward defenders hard enough to need a ground-side buff. Split still sits at a 53.0% defensive win rate after that rework [5], which means the ropes and Tower positions remain the load-bearing part of the map’s balance, not the ground lanes Riot already tuned.

Role data reinforces it further. Sentinels post the highest average win rate of any role on Split at 51.0% [6] — ahead of Duelists (49.5%), Initiators (49.1%), and Controllers (48.6%) [6]. Sentinels are anchor agents by design: they hold a single vertical chokepoint rather than pushing through one. A passive, position-holding role outperforming every aggressive role on this specific map is the clearest evidence available that Split rewards owning a vertical position over winning a horizontal gunfight.

Play it out in a round: an attacking side that spends its full utility clearing A Main and Screens, then walks into A Site, is fighting a defender who never left Tower and has a full kit left over for the retake. An attacking side that sends one player up the Sewer-to-A-Lobby rope while the rest fake B is contesting the vertical position the defense was relying on to hold cheaply — and that one player often decides the round before the other eight ever trade shots. Prioritize the second pattern over the first whenever your team has the utility to support it.

Rope Timing Tactics: The Audio-Masking Protocol

Ropes are loud enough that a defender expecting one can pre-aim it and win the fight before you’ve finished climbing [7]. The fix isn’t avoiding ropes — it’s timing them so the audio cue stops mattering.

1. Never climb into a contested angle without cover noise first. Pop a flash, smoke, or a decoy footstep source in the same one-to-two-second window as your climb. A defender’s attention and audio focus go to the louder, more urgent cue, not the rope.
2. If your team is executing elsewhere, that’s your climb window. Utility, flashes, and gunfire from a site execute mask rope audio across the map. Use that window to rotate or flank rather than climbing during a quiet round.
3. Hang, don’t sprint through. Holding position on a rope and popping up to peek is consistently described as the higher-percentage play over climbing all the way through [7] — it limits your exposure to a fraction of a second instead of the full climb animation.
4. Treat walk-pace climbing as unverified, not silent. Some community guides claim ropes generate less audio at a walking pace, which lines up with how Valorant ties movement sound to speed tier elsewhere in the game. Riot hasn’t published rope-specific audio ranges, so verify this yourself in a custom game before trusting it in a ranked round — don’t take an unconfirmed claim at face value when your own utility can mask the sound for certain instead.

The highest-value use of this isn’t a straight site push — it’s the mid lurk. A player who climbs the B Hell-to-B Tower ascender, or one of the Sewer ropes, while the rest of the team executes the opposite site arrives behind rotating defenders. Defenders on Split are trained to expect frontal pressure at their anchor position, so a same-flank punish from a rope route catches them looking the wrong way far more often than a second frontal push would. If the rope is already contested when you reach it — footsteps on the far side, a held angle you can hear — don’t force it. A contested rope climb into someone already holding the top is close to a free kill for them; fall back and rotate through a slower, safer path instead.

Split Strategy by Player Type

Best Agents for Split in 2026 (Verified July 2026)

Based on a 45,045-match sample, six agents post win rates above 53% on Split [6]:

Two of these need context beyond the raw number. Sova’s 55.2% win rate comes with a 0.2% pick rate [6] — that’s a tiny sample, and a high win rate on almost nobody’s games should be read as \”strong when drafted deliberately,\” not \”guaranteed.\” Neon is the more reliable pick of the two: a 3.5% pick rate on a 56.3% win rate [6] is a large enough sample to trust, and her kit’s straight-line speed is one of the few things that closes a horizontal gap on a map that otherwise punishes ground pushes.

You’ll find different numbers elsewhere — one earlier aggregation put Sova and KAY/O ahead of Neon by several points. Trust the figures above over that: they come from a directly verified, 45,045-match sample rather than a smaller or unspecified aggregation, and Split’s stats move fast enough between patches that citing the larger, dated source is the safer call [6].

The role split matters more than any single agent pick. Sentinels average the highest win rate of any role on this map, 51.0% [6] — Sage and Killjoy both anchor a Tower position rather than contest one, which lines up exactly with the vertical-control mechanism explained above. If you’re building a comp from scratch, weight a Sentinel anchor over a second aggressive Duelist.

Attack vs Defense: Site-by-Site Decision Framework

Defending A: hold Tower or Rafters, not ground-floor Main. Ground positions get cleared by utility before you get a read on the attack; Tower positions force attackers to commit to a rope contest before they even see the site [2].

Defending B: the B Hell-to-B Tower ascender is the rotation attackers use most for late flanks. A Sentinel trip or camera watching that rope buys your anchor the reaction time a ground-only setup doesn’t get.

Attacking with mid open: send a lurker through mid on a rope rather than stacking a third player on the execute. Split’s mid is small enough that one player controlling it disrupts defender rotations more than an extra gun on site would [1].

Attacking with mid contested: don’t force the rope. Fall back to a standard two-site fake and wait for a cleaner window instead of feeding a defender who’s already holding the angle.

Version Note

Every stat and mechanic in this guide is current as of Patch 13.00, live since June 23, 2026 [3]. Split remains part of the active seven-map competitive rotation this patch — Summit was added and Sunset returned in place of Fracture and Pearl [3], neither of which touches Split directly, but confirms the map itself hasn’t been rotated out. The one Split-specific change in 13.00 was a Viper’s Pit smoke-artifact fix, alongside a fix for the Pit incorrectly splitting when it overlapped certain Sage Barrier Orb or Vyse Shear placements [3]; neither affected win rates or callouts. Verify agent win rates against a current source before a ranked session — Valorant’s balance patches land every few weeks, and Split’s numbers move with them.

FAQ

Why does Split favor defenders so heavily?
Because defenders start every round already holding the map’s vertical positions — Tower, Rafters — while attackers have to actively contest ropes and elevation they don’t start with [2][5]. Riot’s Patch 6.0 rework tried to close that gap by easing horizontal attacking routes [4], and defenders still hold a 4.9-point edge afterward [5], which means the vertical geometry, not the ground lanes, is still doing the heavy lifting.

Is climbing a rope always risky on a defense-favored map like this?
Only unmasked. A rope climbed during your own team’s site execute, flash, or utility barrage is functionally silent in practice, because the defender’s attention and audio focus are already elsewhere. A rope climbed into a quiet round is the version that gets you killed — the round state matters as much as the rope itself.

What’s the single best agent for Split right now?
Neon, on the numbers that hold up — a 56.3% win rate on a 3.5% pick rate is both strong and well-sampled [6]. Sova’s raw win rate sits close behind, but at a 0.2% pick rate it’s a much smaller sample than the number suggests, so treat it as promising rather than proven.

Do I need to master every ascender to play Split well?
Not to play it competently. The three primary ascenders — two Sewer-to-A-Lobby, one B Hell-to-B Tower — cover most rotations and flanks [2]. The four additional Vent ascenders matter more for improvised mid-round reads than for a standard execute, so learn them once the primary three are automatic.

Key Takeaways

• Split’s defenders win 53.0% of rounds to attackers’ 48.1% [5], and the gap traces to who controls Tower and the ropes, not who pushes harder through Main.
• Time rope climbs against your own team’s noise — a flash, a smoke, or a simultaneous execute — instead of climbing into a quiet round.
• Neon, Sova, and Viper post the strongest verified win rates on Split in 2026 [6], but weight Neon’s number highest given its far larger sample size.
• Sentinels are the map’s strongest role on average [6] — direct evidence that anchoring a vertical position outperforms contesting one.

For deeper agent-specific breakdowns, see our guides to Viper, Sage, and Miks, or start from our full Valorant Beginner’s Guide 2026 for economy and rank fundamentals.

Sources

• Riot Games (Official) — VALORANT Maps overview
• Riot Games (Official) — Split map wiki
• Riot Games (Official) — Patch Notes 13.00
• Liquipedia — Split map history and layout
• MetaBot.GG — Split win-rate and match stats
• MetaBot.GG — Split best-agent win rates
• Dexerto — Split callouts and strategy guide

Michael R.

I've been playing video games for over 20 years, spanning everything from early PC titles to modern open-world games. I started Switchblade Gaming to publish the kind of accurate, well-researched guides I always wanted to find — built on primary sources, tested in-game, and kept up to date after patches. I currently focus on Minecraft and Pokémon GO.