Shelly is a play,
not a turret coupon.
We followed 1,358 Rift Herald deployments for two minutes: Shelly’s own structure damage, recalls, fights, deaths, gold, lane pressure, and what happened to the player carrying the Eye.
- OBSERVED DROPS
- 1,358
- DIRECT TURRET HITS
- 1,380
- DIRECT DAMAGE
- 3.78M
There is no demonstrated premium for holding Shelly.
Quick drops and one-to-three-minute holds were effectively level on structure conversion: 83.3% versus 84.3%, with overlapping 95% confidence intervals. The study does not establish that waiting creates value.
Use the Eye when a clean window is available and take the gold. Waiting is justified by a concrete wave, lane, or numbers advantage—not by the timer itself. Late or automatic drops converted least often and left the holder dead most often.
The average Shelly play moved the game forward.
All aftermath numbers are measured from the deployment team’s perspective.
Net gold
+691Mean gold-differential movement in the two minutes after deployment.
Net deaths
+0.29Opponent deaths minus allied deaths in the same window.
Team structure
83.2%The deploying team destroyed at least one structure within two minutes.
Win probability
+1.04 ppModeled movement from gold and game time, not a causal Shelly effect.
Most plays got one hit. One in five got more.
The categories are mutually exclusive. Damage is Shelly’s own direct structure damage; gold is the full team swing afterward.
Eventual win is the raw cohort rate. Two-minute win state is the modeled change from gold differential and game time after deployment. It is descriptive movement, not a causal estimate of Shelly’s effect.
Mid was the default. Side-lane Shelly was safer for the holder.
Lane is assigned from the first direct target, or the deployment location when there was no hit.
Mid
809Top
321Bot
228371 deployments met the split-push classifier. Their holders died less often than other Shelly players (45.3% versus 53.1%) while direct-hit conversion stayed nearly flat. Split Shelly was a real pressure pattern, not a turret-damage multiplier.
A recall was detectable. Causality was not.
Shelly is deployed
1,358 observed deployments begin the same two-minute aftermath window.
Nearby recall completes
175 plays, or 12.9%, had an opponent complete a recall near the pressured lane.
A teamfight follows
51 of those 175 recall plays reached a six-player, one-death fight within 60 seconds.
Death edge is positive
42.7% of all qualifying post-Shelly fights ended with a positive death differential for Shelly’s team.
Use it now unless waiting buys a specific setup.
Immediate and one-to-three-minute conversion rates are statistically indistinguishable here; late holding is the only clearly weaker pattern.
A drop that starts a losing fight.
A throw requires an immediate six-player fight, the Eye holder participating, and a negative death differential. Shelly may still hit a structure first; damage and gold are measured afterward, not used to define the label.
141 throws
10.4% of deployments immediately became a holder-involved fight that lost the death trade.
−738 gold
The average two-minute aftermath, measured only after the fight-based label was assigned.
85.1% holder death
The Eye holder died in 120 of the 141 losing immediate contests.
116 banked damage
Most throws still hit a structure first. Only 25 were zero-hit throws; 116 dealt direct damage before the fight was lost.
Methods, definitions, limitations, and sources
Sample
1,358 first observed allied Shelly deployments after the first Herald capture in distinct Patch 16.17 matches. The unit of analysis is the deployment, followed for 120 seconds.
Direct damage
Only damage dealt by Shelly’s charge to the struck structure enters the direct total. Player, minion, and later team damage do not.
Pressure recall
A completed enemy recall beginning near Shelly’s lane before 25 seconds after her last impact, or within 35 seconds of a zero-hit drop. “Pressure” is proximity and timing, not proof of motive.
Fight and throw
A teamfight requires at least six participants and one death. An immediate Shelly fight begins within 35 seconds of deployment or 25 seconds of her last hit and includes the Eye holder. A throw is an immediate fight with a negative death differential, whether or not Shelly hit first.
Interpretation
Herald-taking teams were already selected and often ahead. Raw win rate and modeled state movement describe the cohort; neither estimates what the same team would have done without Shelly.
Mechanics sources
Riot’s Patch 14.1 notes, Patch 26.1 notes, and Herald design article. Study corpus: Manalytics Patch 16.17 replay corpus, accessed September 2026.