Wednesday, March 25, 2009

The Lineup.

So it’s the time of year when I start getting my draft prep together for Stratomatic tournaments – I plan to be playing at the Chicago tourney on April 18th and 19th, if anyone is up for a challenge. (I only played once last year, as some of you recall.)

Preparing for a Strat tournament reminds me of the many things I’ve learned about baseball from playing the game, and one of those lessons is how to properly build a lineup. With the identity of the nine starting Royals pretty well cemented (with the exception of second base*), I think it’s time we figure out the best way to arrange those nine players in the best possible fashion.

*: You could argue that the catching position isn’t cemented so much as it’s been cast in Play-Doh, given that Miguel Olivo and John Buck will probably share the job almost equally. But as I’ve written many times, you’d be hard-pressed to find two players more similar than Olivo and Buck, at least offensively, so which one happens to be in the lineup on any given day doesn’t impact the lineup order at all.

Here’s a list of unofficial rules I’ve developed on lineup construction, and how they apply to the Royals:

Rule #1: Build your lineup from the middle out, not from the top down.

In other words, don’t figure out who your leadoff hitter is first and then the guys after him, decide on your #3 and #4 hitters first, and then fill in the slots around them. The single most important principle in extracting the most runs from your lineup is to bunch the good hitters together; placing your best hitter (even if he would make a terrific leadoff hitter) directly behind the #8 and #9 hitters is a less-than-optimal solution.

The best way to demonstrate this principle is to look at what the Royals are doing with David DeJesus. In the abstract, if you were to look at each member of the Royals individually and then decide who would make the best leadoff hitter, you’d almost certainly hit on DeJesus, who after all has the highest OBP on the roster (career .360 mark) and doesn’t have a ton of power.

But DeJesus might also be the Royals’ best #3 hitter, because he’s a high-average hitter who drives the ball in the gaps, with a relatively modest platoon split for a left-handed hitter (lifetime OPS of 722 vs. LHP, 809 vs. RHP.)

The Royals are planning to move DeJesus into the 3 hole, which I think is the right decision, though maybe for the wrong reasons: the Royals appear fixated on the fact that DeJesus hit .419 with runners in scoring position last year, the highest figure in all of baseball and the second highest in Royals history (behind George Brett’s sublime 1980, when he hit .469 with RISP). There’s no reason to think that DeJesus’ performance is anything but a fluke – he had never hit above .300 or below .270 with RISP before – but he does have the profile of someone you want up with a man in scoring position: a line-drive swing, doesn’t strike out a whole lot, not ridiculously patient but not overly aggressive either.

DeJesus isn’t the perfect #3 hitter, just as he’s not the perfect leadoff man, and on another team, one with a nice power core, DeJesus would fit in best higher up in the lineup. For the Royals, he’s the guy the Royals should build the lineup around for now.

Rule #2: The #2 hole is the most underrated slot in the entire lineup.

After the 2001 season, I engineered a massive trade in my perpetual Strat-o-matic league to acquire Barry Bonds. Bonds, as you can imagine, had just been issued a Strat card the likes of which had never been seen before. (Adding to the mystery, there was no player’s name on the card, owing to Bonds’ decision to opt out of the MLBPA’s licensing agreements.) The acquisition of a player that could hit 73 homers and draw 177 walks did create the rather pleasant dilemma of figuring out where he ought to bat in the lineup. With all that power, you wanted him up with men on base; with all those walks, you wanted him up with no one out so that there’d be ample opportunity for the guys behind him to drive him in. Obviously, those two desires conflicted with each other.

The obvious spot for your best hitter is to bat him third, where he can drive in the on-base guys batting first or second, and be driven in by your #4 and #5 hitters. The problem with that edict is that while the third spot in the lineup is the perfect spot to bat when the lineup is humming, it’s the worst spot to bat when it isn’t. Consider this: the #3 hole is the most likely spot in the lineup to bat with two men out and no one on. Assuming your leadoff and #2 hitters have an OBP of .350, the odds that both will make outs their first time up is (1-.350) = .650 squared, or 42.25% of the time. So right out of the chute, your #3 hitter is going to bat in the least useful situation almost half the time his first time up. By comparison, it is impossible for the cleanup or #5 hitters to bat in that situation in their first plate appearance.

(In major league baseball’s primordial past, teams were not required to announce their lineups before the game started. Cap Anson, who when he wasn’t enforcing Jim Crow laws was one of the 19th century’s greatest players, took advantage of this rule as player/manager: while he normally batted third, if the first two players made out in the first inning, he would send up someone else to bat third instead rather than waste his best hitter – himself – in a low-leverage situation.)

So I could have batted Bonds cleanup, but the lower down you bat in the lineup, the fewer opportunities you’ll get to bat in a game – to the tune of 18 plate appearances a season for every spot you drop down the lineup. With a performance that good, I wanted Bonds up there more often, not less. I toyed with batting him leadoff – call this the Bobby Bragan Theory – but I couldn’t stomach the thought of all those bases-empty homers.

So I batted him second, and kept him in that spot for the next four years. This worked out wonderfully – he was far less likely to bat with two outs, but he also had the opportunity to bat with a man on base in the first inning.

All this might be just an interesting anecdote, but this article over at Beyond the Box Score (which itself quotes from “The Book” by Tom Tango, Mitchel Lichtman, and Andy Dolphin) corroborates this concept: you should put one of your best, if not your best, hitter in the #2 hole.

(I’ve always felt that the best lineup slot for Ichiro Suzuki, who is in his own way as unusual a player as Bonds was, is the #2 hole. There’s nothing wrong with him leading off, given his ability to get on base and run, but his ability to reach base is disproportionately the result of singles rather than walks. Those singles have the advantage of advancing runners, something walks rarely do – so Ichiro’s talents are particularly useful when there are men on base ahead of him. Just ask the Koreans.)

Which is a long-winded way of saying that in a perfect world, DeJesus wouldn’t bat first or third – he’d bat second.

Rule #3: Stolen bases are most useful when they come in front of your worst hitters, not your best ones.

I think these are the two biggest misconceptions about building a lineup:

1) The #2 hitter should be a “bat-control” guy who may not be a great hitter, but can bunt, hit-and-run, make productive outs, and otherwise find ways to disrupt the link between your best on-base guy and your best hitter. (See Rey Sanchez in 2000, batting between Johnny Damon and Mike Sweeney.)

2) Your leadoff and #2 hitters should have the ability to steal bases, because it’s soooo important to move from first to second base when you’ve got a home-run hitter at the plate.

I’ve never understood this concept. When you’ve got a power hitter at the plate, what’s the point of risking an out – and losing a baserunner – in order to move that runner 90 feet closer to home plate. Isn’t the whole point of a power hitter that you can drive in runners from first base? Your traditional middle-of-the-lineup hitter isn’t any better at hitting singles than the average hitter – he’s just better at hitting doubles and home runs. (Albert Pujols, the best player in baseball, hit singles in 20.2% of his at-bats last year. His teammate Adam Kennedy, he of the .280/.321/.372 line, hit singles in 21.2% of his at-bats.)

So why on earth would you risk losing a baserunner (and making an out) with your best hitters at the plate, particularly since your best hitters ought to be quite capable of driving in that run from first base?

(The argument for speed at the top of the lineup has always included the notion that fast runners are more likely to take the extra base, to score from second on a single or from first on a double. That’s an argument (albeit a weak one) for good baserunners, but not necessarily good basestealers. That might seem like a subtle difference, but in the Royals’ case they have two excellent examples of players who are the former but not the latter: DeJesus, who is a lifetime 56% basestealer but a very good runner, and Mark Teahen, who despite a good success rate (77%) has just 34 steals in four seasons, but is one of the smartest baserunners in the game.)

On the other hand, when the 7-8-9 hitters are up, a single is frequently the best outcome you can realistically hope for, in which case the difference between having a baserunner on first and a baserunner on second is significant. Looking at Kennedy again, he hit singles in 21.2% of his at-bats, but had an extra-base hit in just 6.8% of his at-bats. He’s roughly four times as likely to drive a runner home from second than from first (assuming the runner always takes the extra base) – making a stolen base look awfully tempting.

The overall break-even rate for the stolen base is about 70%; in other words, if the odds that you’ll steal second base safely are less than 70%, you’re better off not making that attempt. But the break-even rate is much higher in the leadoff or #2 spots, maybe on the order of 75%, and much lower in the #6 and #7 spots, maybe around 65%.

The ability to steal bases is actually more valuable for a #5 hitter than the leadoff hitter, even though I’d wager that more than 97% of all major league baseball teams have had a more prolific basestealing threat in the leadoff slot.

Rule #4: Put the guys who get on base ahead of the guys who hit for power.

This isn’t really a rule I picked up from playing Strat; I think every baseball fan figures this out on their own by the time they’re 12. I only bring it up here because Trey Hillman seems to think it makes since to bat Mike Jacobs fifth and Alex Gordon seventh, instead of the other way around.

Jacobs and Gordon are both left-handed hitters. They both have power; Jacobs has shown more to this point, but had never hit more than 20 before last season. I’m willing to bet that Gordon hits more homers than Jacobs this year, but let’s concede the point that at this moment, Jacobs is the more established power threat. Gordon is clearly the more established on-base threat. Why on earth would you bat the guy who’s an all-or-nothing hitter ahead of the guy for whom walks are a key part of his arsenal?

The answer is because Jacobs fits the caricature of the #5 hitter – the slow, oafish, free-swinging lout who parks balls in the upper deck and then trots home behind the two guys on base when he bats. The problem with the traditional concept of the #5 hitter is this: guess who’s most likely to lead off the second inning? That’s right – the one guy in the lineup most ill-suited to lead off.

Rule #5: Other things equal, put your worse defensive players higher in the lineup, so that they can get their full allotment of at-bats and then get removed for defensive purposes earlier in the game.

I can tell you from experience: few things are more annoying to your Strat opponent than when you get away with starting a terrible defensive player with a big bat, because you’re able to pull him out of the game for a defensive replacement before a difficult fielding play comes his way. In the tournament I won last season, I started Jack Cust in left field, but by batting him 2nd in the lineup, I was typically able to get three at-bats from him and still get him off the field in favor of a real outfielder by the sixth inning.

This isn’t a very useful rule in 21st-century baseball, when teams don’t have the roster space for defensive replacements. In the Royals’ case, the only instance where I see this rule being useful is in the event that Mark Teahen is playing second base.

Rule #6: Avoiding runs of same-sided hitters trumps all the other rules.

In modern baseball, you simply can not afford to bunch right-handed or left-handed hitters in a row. As recently as 20 years ago, this wasn’t the case; as recently as 10 years ago, you could get away with it against most teams. Today, every team in baseball has a LOOGY or LOOGY wannabe, and most teams have a ROOGY as well.

Lineup balance is important. It doesn’t matter how strong your lineup is, if you’re running seven right-handed hitters out there, as the Tigers were for much of last season (Carlos Guillen and Curtis Granderson were the only left-handed hitters in the lineup many times), you’re going to be vulnerable in the late innings.

That’s an Achilles’ heel which can be exploited on even the best of teams. The Cubs’ lineup was so potent that they led the National League in runs (and wins) during the regular season – but against the Dodgers in the NLCS, they faced a right-handed starter in each game and still started six right-handed hitters (and that’s not counting the pitcher.) In game 2, their 1 through 5 hitters were all right-handed. They scored six runs in three games.

Fortunately, the Royals have a lot more lineup balance than that. Last year the Royals had remarkable success against left-handed pitching – they went 36-24 when a southpaw started, and just 39-63 against right-handers – which is made even more remarkable by the fact that the lineup was actually a little more left-handed than average. The Royals got 54.1% of their plate-appearances from right-handed hitters last year; the AL average was 55.4%.

The reason for the Royals’ success against lefties was that the right-handed hitters they did have put up enormous splits. Against RHP, all the right-handed batters on the team hit .247/.281/.358; against LHP, they hit .302/.359/.491. For that, you can thank guys like Jose Guillen (.305/.361/.575 vs. LHP, .248/.274/.382 vs. RHP), Billy Butler (.340/.395/.590, .244/.290/.308), Mark Grudzielanek (.395/.459/.523, .265/.301/.355), Mike Aviles (.348/.392/.574, .313/.333/.432), and Miguel Olivo (.262/.296/.534, .251/.268/.399).

So while the Royals had a pretty balanced lineup last season, there’s no question that they need to figure out a way to hit right-handed pitching. (Although amazingly, given how successful the Royals were against southpaws last season, they actually saw more left-handed pitchers than any other AL team.) The addition of Coco Crisp and more playing time for Alberto Callaspo, both switch-hitters, may help with this problem – and looked at in this light, the addition of Jacobs almost makes sense.

I’m getting off track here…the Royals lineup is likely to include a switch-hitter (Crisp), three left-handed hitters (DeJesus, Gordon, Jacobs), four right-handed hitters, (Aviles, Butler, Guillen, Olivo/Buck), and a second baseman to be named later. Given that kind of balance, there’s no reason why the Royals should ever have to bat two left-handed hitters, or two right-handed hitters, in succession.

So them’s the rules. Ordinarily that should be sufficient to design a lineup, but I have to throw one more monkey wrench into this thing: the order of the lineup obviously depends on your expectations of the players’ performances, and the Royals’ expectations for some of their players (say, Jose Guillen) are markedly different than mine. It’s easy for me to argue that Guillen deserves to bat 8th; easy, and useless, as there’s no chance in hell that the Royals would consider such a thing. So here’s the lineup I think the Royals should start the season with, based on their expectations of how each player will play.

S Coco Crisp

R Mike Aviles

L David DeJesus

R Jose Guillen

L Alex Gordon

R Billy Butler

L Mike Jacobs

R The Catcher

? The Second Baseman

Crisp isn’t the ideal leadoff hitter given his OBP issues, but really, aside from DeJesus who’s better qualified? Leading off with him allows the Royals to go R-L-R-L the rest of the way. Aviles fits the traditional role of the #2 hitter – the scrappy overachieving shortstop – with the added advantage that he might be able to hit. The 2008 Guillen shouldn’t be batting this high in the lineup – maybe not anywhere in the lineup – but the 2004, 2005, or 2007 Guillen is a decent cleanup option. Gordon gives you power (if he comes up in the first inning) and on-base ability (if he comes up in the second inning) out of the #5 hole. Jacobs' combination of power and low-OBP – I don’t care what the Royals are saying – is perfect in the #7 spot.

The one potential change I could see to this lineup order is to flip Crisp with Alberto Callaspo in the #9 slot, assuming Callaspo shows that last year was no fluke. If Callaspo is really a .300 hitter, he has the superior on-base ability, and I’d much rather have Crisp’s speed ahead of Callaspo's singles than the other way around.

That lineup is awfully bland, in all honesty – the only significant difference between my lineup and the tentative one that Hillman has hinted at is that I’ve swapped Gordon and Jacobs. The advantage is that it’s conventional enough that the Royals might actually take me up on it.

But let’s say, come mid-May, that Gordon and Butler are raking and that they’re ready for a more important spot in the lineup. In that case, I present:

L David DeJesus

R Mike Aviles

L Alex Gordon

R Billy Butler

L Mike Jacobs

R Jose Guillen

S Coco Crisp

R The Catcher

? The Second Baseman

Gordon’s emergence allows DeJesus to move back to the leadoff role. Butler and Jacobs both move up two spots – Jacobs is not optimal in the #5 hole, but it’s the one concession that has to be made to keep the lineup alternating from right to left.

Crisp moves to seventh, where his ability to steal bases in anticipation of all the singles hit by Wilberto Bloomaspo and DeJesus comes in handy. The most controversial part of this lineup is putting Guillen sixth – or at least, it will be controversial to Jose Guillen – but if the rest of the lineup is hitting on all cylinders, that’s where he belongs.

Finally, here’s the lineup I’d put together in the unlikely event that not only are Gordon and Butler hitting, but Mark Teahen has proven he can handle second base:

B Crisp

L DeJesus

R Butler

L Gordon

R Guillen

L Jacobs

R Aviles

L Teahen

R Catcher

DeJesus finds his way into the #2 hole that really seems like the best use of his skills. The need to realign the RH and LH hitters moves Butler ahead of Gordon, but it also puts Guillen in the #5 spot, a slot he might actually accept without a clubhouse mutiny. Teahen bats low in the order, which violates Rule #5, except that I don’t envision Teahen coming out of the game for defense. What I envision is that Guillen gets pulled in favor of Bloomquist, with Teahen moving to right field.

That’s a lineup I can get behind. Even if there’s no chance of it actually coming to fruition.