As we talked about on Monday, Mike Trout has hit Felix Hernandez very well. After his first-inning home run on Opening Day, Trout is now hitting .441/.447/.794 in 38 plate appearances against Hernandez since being called up to the majors for good in April 2012. The question for the day, then, is this: How well should Mike Trout do against Felix Hernandez?

Which inevitably leads to this: How well would Bartolo Colon do in 600 at-bats against Craig Kimbrel? How many bases would Billy Hamilton steal if every pitcher in the league were Adam Wainwright? But we’re getting ahead of the intro.

To answer the Trout/Hernandez question, we turn to our old friend Log5, a Bill James invention that helps us determine how two teams with different winning percentages would fare against each other, and that can be tweaked slightly to tell us what a hitter of a certain skill level should do against a pitcher of a certain skill level. (Or observed performance level, since we're not regressing to determine each player's true talent over a given period.) If a batter who homers 20 percent more than the league average faces a pitcher who allows 20 percent more home runs than the league average, he should hit even more home runs. And so on. Intro. Words. Segue. Bold text:

Felix Hernandez (2012-2014) vs. Mike Trout (2012-2014)
So we’ll imagine that they face each other 600 times, that Felix Hernandez’s entire season is spent facing Mike Trout, over and over, in a league filled with reasonably-priced Mike Trout clones (and that neither one benefits more from seeing the other so often). In that season,

  • Trout would strike out 155 times and walk 64.
  • That leaves 381 struck balls, of which 19 would be home runs, leaving
  • 362 balls in play, of which 144 are hits (.398 BABIP), 28 of them doubles and nine of them triples
  • For an overall slash line, that goes like this: .305/.379/.499

Most comparable hitter: Matt Holliday (.297/.383/.492)

Most comparable pitcher: Randy Wolf (.313/.370/.510)

So Mike Trout turns Felix Hernandez into Randy Wolf. Felix Hernandez turns Mike Trout into Matt Holliday. See how this works? Onward!

Greg Maddux (1994) vs Tony Gwynn (1994)
The most cerebral pitcher against the most cerebral batter (by reputation, at least), facing each other in each player’s best season:

  • Gwynn would strike out 31 times and walk 32.
  • Leaving 537 struck balls, of which three are home runs, leaving
  • 534 balls in play, of which 181 are hits (.338 BABIP), 30 for extra bases
  • For an overall slash line of .324/.360/.394.

Most comparable hitter: At the time, Bret Barberie, whose .301/.356/.406 line was good for a 97 OPS+. Today, the most similar line would probably be Alex Avila’s .236/.338/.381 since 2012, a 95 OPS+ (though one with a much lower batting average).

Most comparable pitcher: Paul Wagner, who went 7-8 with a 94 ERA+ in 1994. Neither player wins this one, really.

Bartolo Colon (career, as hitter) vs. Craig Kimbrel (2012)
Colon is one of just two active players (Matt Garza is the other) who has struck out in at least half his plate appearances (in a minimum of 100 PA). Craig Kimbrel is the only pitcher in history (minimum of 10 innings) to strike out half the batters he faced in an entire season. Thus,

  • In 600 tries, Colon would strike out 494.4 times, and walk 4.4.
  • Leaving 101 struck baseballs, of which none leave the yard,
  • But of which 20 fall for hits (.201 BABIP), none for extra-bases
  • For an overall slash line of .034/.041/.034.

I should note here that “walks” in this piece includes bases on balls and hit-by-pitches, and in the Colon/Kimbrel matchup all of the “walks” are hit by pitches. While a .034/.041/.034 line is fun, it’s actually Kimbrel’s line I’m more interested in. If he faced an all-Colon lineup for nine consecutive innings, he’d essentially average a one-hit shutout with 23 strikeouts. His chances of throwing a perfect game in any 27-batter stretch would be about one in three.

(Garza, incidentally, would hit .115/.121/.133 against Matt Garza.)

Billy Hamilton (projected 2014 season) vs. Adam Wainwright (2013), as measured in stolen bases
After seeing Hamilton strike out all four times, uglily, on Opening Day, you’d think a .000/.000/.000 line might work, which would limit Hamilton’s opportunities to steal. More realistically, we get this:

  • Hamilton would strike out 158 times and walk 23.
  • Leaving 419 struck balls, of which five would clear the wall, leaving
  • 414 balls in play, of which 133 would be hits (.322 BABIP), 25 for extra bases
  • For an overall slash line of .239/.269/.311.

More to the point, though: PECOTA expects Hamilton to steal .59 bases for every time he reaches first. In this scenario, Hamilton would reach first base 131 times against an All-Wainwright league, and thus steal 78 bases.

And, finally,

Barry Bonds (2002) vs. Pedro Martinez (2000)
The highest OPS+ ever against the highest ERA+ ever. In 600 plate appearances,

  • Bonds would strike out 105 times and walk 119.
  • Leaving 376 struck balls, of which 43 would go out, leaving
  • 333 balls in play, of which 89 land fairly (.267 BABIP),
  • For a .274/.418/.584 line.

Do we believe this? Do we believe that, facing Pedro Martinez 600 times, Bonds would put up a line that, even by 2000-2002 standards, is most comparable to Gary Sheffield’s? Certainly the walk rate is inflated, but do we believe that he’d slug .584? That, if the rest of the league got to face Kip Wells, Aaron Sele, Joe Kennedy and the like, and Bonds faced Martinez over and over and over, that he’d still be an MVP candidate? Boy. Boy, I don’t know. But I do know this, and this is officially my new favorite Barry Bonds Fun Fact ever:

From 2001 to 2004, Bonds had 138 plate appearances against pitchers in the same season that those pitchers got Cy Young votes. In those 138 plate appearances, Bonds hit .327/.522/.786.

Now, none of these pitchers, even in a Cy Young-candidacy season, was as good as Pedro in 2000 (though a number were lefties and/or Randy Johnson, giving them a sizeable platoon edge to boot). But still: Against Cy Young candidates! So I’m not ruling anything out.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
"a .274/.418/.584 line"

Holy. Crap.

Nope, not a Hall of Famer, move along.
For those of you wondering (like me) why he used Log5 instead of (Russell Carleton's favorite) Odds Ratio, it's because both methods give you the same expected number. has details, but you can show it yourself with some algebra.
Barry Bonds was amazing. The Cy Young stat is unbelievable.

Any chance you could do this using Bonds' best season pre-roids?
Since Barry Bonds never failed an in-season drug test, how would we know when that was?
Bonds got lucky in a time they were looking the other way and not cracking down or enforcing.
And if Bartolo Colon faced Bartolo Colon? Other than the obvious black-hole related implications certain when so much mass is so proximate.
And out of curiosity, how frequently does PECOTA project a Billy Hamilton SB when he's on second? Third?
It doesn't project those steals I mentioned specifically when he's on first; I just used the ratio for total times reaching first so I'd have something to prorate.
What if Micah Owings faced Micah Owings?
Or what if Rick Ankiel (2000) pitched to Rick Ankiel (2008)?
We're missing the obvious one: Babe Ruth (1918) vs. Babe Ruth (1927).
This was an awesome read. Would like to see some more comparables. Ruth vs Maddux or Pedro would be interesting as well.