week 10
this weekSELL-HIGH
last weekSELL-HIGH
————————
Kevin Kelly
data as of June 3, 2026·week 10
Kelly is a sell-high — his ERA is built on BABIP luck.
His ERA sits at 2.67, looking sharp, but the number keeping runs off the board is his BABIP-against: .211, 84 points below the league average of .295. BABIP-against needs around 800 balls in play to be reliable, and he has 73 — that gap is luck, not skill. Strip out the noise and his expected ERA is 3.52, nearly a run above the box score. His strikeout rate is below league average and his hard-hit rate allowed is above league, so there's no dominant stuff to fall back on when the luck normalizes. The ERA is real on the scoreboard; the xERA is real in the data. Sell high.
Sell High
high0.89
Drivers
- babip_againstNOISEBABIP-against 0.211 suppressed and unstable — flattering the ERA
- xeraNOISEluck-free xERA 3.52 is 0.85 above the ERA — regression coming
metric · nowrolling trajectoryvs. leaguestabilized?
Hard-Hit % Allowed42.5%
lg avg 40.0% ▲+2.5%trustworthy at this sample
SIGNAL73 / 50 BBE
Barrel % Allowed8.2%
lg avg 8.0% ▲+0.2%trustworthy at this sample
SIGNAL73 / 60 BBE
xwOBA Allowed.298
lg avg .315 ▼−.017too early to trust
NOISE98 / 200 TBF
xERA3.52
lg avg 4.10 ▼−0.58too early to trust
NOISE98 / 200 TBF
ERA2.67
lg avg 4.10 ▼−1.43too early to trust
NOISE27 / 200 IP
Strikeout %19.4%
lg avg 22.0% ▼−2.6%trustworthy at this sample
SIGNAL98 / 70 TBF
Walk %5.1%
lg avg 8.0% ▼−2.9%too early to trust
NOISE98 / 170 TBF
BABIP Against.211
lg avg .295 ▼−.084too early to trust
NOISE73 / 800 BIP