Pricing a Derby Format That's Never Been Played

MLB changed the Home Run Derby format for 2026: 20 swings in Round 1, 15 swings in the semifinals and final, and if your last swing is a homer, you keep swinging until you miss. A format that has literally never been played before.

After the PrizePicks post, you know why this stirred my interest: the best edges usually aren't in the odds — they're in the rules that govern how the odds get paid. And when the rules of a marquee betting event change, every line that's anchored to the old rules is potentially stale. Every recent Derby memory the market has is from the timed era (2015–2025), where hitters got minutes on a clock plus bonus time and the totals got cartoonish — Vlad Jr. once hit 40 in a single round. If books and bettors price a 20-swing round off timed-era memory, the unders might be free money.

But that argument only works if I can answer the actual question: how many home runs does a Derby hitter hit in 20 swings?

The forgotten data

Here's the thing the market forgot: we have swing-limited Derby data. Every Derby before 2015 was attempt-limited — you hit until you made 10 outs (7 in 2014; two 5-out innings in the late '80s). An outs-limited round and a swing-limited round are the same family: bounded attempts, no clock. The timed era is the weird one.

So I scraped every outs-era round I could get: 206 first-turn rounds, 1985–2014, from Baseball Almanac and Wikipedia's scoreboards. (1991–94 got dropped — the public tables only give tournament totals, not round-by-round HR and outs, and I'm not going to blend unverifiable rows into the rate math.)

Each round gives an HR-per-swing estimate:

HR rate per swing = HR / (HR + outs)

Project that onto the new format, with a small bonus for the keep-swinging-until-you-miss rule (a geometric tail):

E[HR] = n·p + p²/(1−p)     where n = 20 or 15 swings

206 outs-era rounds. The median Derby participant ran about a .26 HR-per-swing rate. The all-time outlier ceiling is Josh Hamilton's .737.

What 20 swings actually buys you

Player type	20-swing Round 1	15-swing semis/final
Median outs-era hitter	~7 HR	~5 HR
75th percentile	~9 HR	~6–7 HR
90th percentile	~10 HR	~7–8 HR
Elite / hot round	~12–14 HR	~9–11 HR
Historic outlier ceiling	~16–17 HR	~12–13 HR

And the greatest outs-era rounds ever, translated:

Old round	Old HR	Est. HR/swing	Over 20 swings	Over 15 swings
Josh Hamilton, 2008 R1	28	.737	16.8	13.1
Bobby Abreu, 2005 R1	24	.706	15.8	12.3
David Ortiz, 2005 R1	17	.630	13.7	10.5
Yoenis Céspedes, 2013 R1	17	.630	13.7	10.5
David Wright, 2006 R1	16	.615	13.3	10.2
José Bautista, 2014 R1	10 (7 outs)	.588	12.6	9.7

Hamilton's mythical 28-homer round — the greatest Derby round anyone has ever seen — took him 38 swings. Under the 2026 rules it's a 17. That's the whole story of the format change in one number.

I also simulated 100,000 full tournaments, drawing hitters from the 1995–2014 rate distribution: a typical 8-man Round 1 produces about 50 total homers, the advancement cutoff lands around 7, the Round 1 leader around 12, and 16+ is a genuinely big night. If a book hangs a number with timed-era memories baked in, that's where the money is.

The eight guys tonight: a raw power check

Format tells you the shape of the night; the field tells you who wins it. So I pulled every batted ball for all eight contestants off Baseball Savant — max exit velocity by season, plus how many 110+ mph balls they've hit this year:

Player	Bats	2024 max EV	2025 max EV	2026 max EV	110+ mph balls, 2026
Junior Caminero	R	116.3	116.7	118.5	38
Jordan Walker	R	115.5	117.9	116.6	33
Jac Caglianone	L	—	114.6	116.1	29
Kyle Schwarber	L	115.8	117.2	113.2	15
Willson Contreras	R	115.3	114.7	114.4	17
Munetaka Murakami	L	—	—	114.1	8
Bryce Harper	L	113.8	113.8	113.5	9
Ben Rice	L	110.8	113.8	111.4	4

Statcast, through July 12. Caglianone debuted in 2025; Murakami's 2024–25 were in NPB, so no Statcast there.

Two things jump out. Caminero is the raw power king of this field, and it isn't close — the hardest single ball (118.5, a career best, this season) and more 110+ contact than anyone. And Ben Rice is the inverse: 31 homers on a max EV of 111.4 — he gets there on pull-side loft, not force. For a longest home run prop, you want the 118 guy, not the 31-HR guy.

Rebuilding the park

The Derby is at Citizens Bank Park this year, and everyone keeps saying it favors lefties. Rather than repeat that, I rebuilt the fence profile — distances and wall heights — and threw every contestant's 2026 air balls at it. Toggle through the eight:

The wall itself makes the lefty case. Clearing a fence means beating the distance plus the wall (a fly ball descending at ~50° needs roughly 0.8 extra feet of carry per foot of wall). Do that math around the park: the right-field alley needs about 380 feet. The left-field alley needs about 387. And just left of dead center sits Monty's Angle — a 409-foot pocket behind a 19-foot wall that needs ~425 feet of carry. Right-handed pull power dies there; left-handed pull power gets the cheapest real estate in the park.

Running all 793 tracked air balls through the profile: the five lefties come out about even with their real-world homer totals (+1 net), while the three righties lose homers (−3 net) — Walker and Contreras each give two back to the Angle. The park effect through this lens is real but modest — a homer or two per player, not five. The reason it matters more tonight: in games these guys spray the ball; in a Derby everyone sells out to the pull side, which is exactly where CBP is most asymmetric.

The wind

Weather Twitter was already on this one, but I pulled the data myself rather than take a screenshot's word for it. NWS hourly forecast for the ballpark (gridpoint PHI 50,77), generated this morning: 8 PM, 80°F, wind SW at 5 mph, going S at 5 by 9 PM, clear.

Citizens Bank Park points almost due north — home plate to center field — so a southwest wind blows out toward right-center. The direction is a lefty's friend, the same side of the park the fences already favor. The magnitude is not: ball-flight physics puts a 5 mph tailwind at roughly 10–15 feet of extra carry on a high fly, and the arrow in the chart above is drawn small on purpose. Wind matters at 15–20 mph. At 5, it's a thumb on the scale, not a hand — enough to nudge a 475-foot ball toward 480+, not enough to change who wins.

So where's the bet?

Here's the anticlimax: after all of that, I didn't find one.

The count-prop prices I could actually get weren't hanging timed-era numbers — the books did the obvious adjustment too. What I was left with, after all the scraping and simulating, were two leans:

• Over 480 feet on the longest home run
• Junior Caminero +400 to hit the longest home run

You can see where both come from in the work above. Caminero owns the two numbers that matter for a longest-homer prop — the hardest-hit ball in the field (118.5) and the most 110+ contact — and a longest HR doesn't care that the park's geometry mildly taxes righties; it only takes one swing, and the Angle can't stop a ball that lands in the second deck. The 480 lean is the same logic plus an 80° night and a small breeze out to right-center. But 20 swings instead of unlimited time means fewer max-effort hacks per hitter than the timed era, which cuts against both — and I couldn't turn "cuts against, but by how much?" into a number I trusted.

A lean means: if you forced me to pick a side, I know which one I'd take. An edge means: I can tell you why the other side of the number is wrong, in one sentence, with a number in it. I had the first thing. I never got to the second. So the answer is the one this blog exists to practice: no bet.

The tuition-priced lesson I keep re-learning is that research isn't a ticket you have to redeem. The work is a filter, not a sunk cost — most good research is supposed to end in "no." The dataset is built now, the script reruns in a minute, and next July when someone hangs a Derby total that doesn't respect the format, I'll be standing right there with the spear.

Tonight I'll just watch.