Jason Heyward, Strikeouts and the #2 Spot in the Lineup
February 26, 2013 at 12:35 pm by Franklin Rabon under Atlanta Braves
With Martin “prototypical #2 hole hitter” Prado gone, a lot of debate has gone into who should hit 2nd, and relatedly, where to best leverage Jason Heyward’s talents. As of now, the plan has seemingly been Heyward in the #2 spot in the lineup, which we at Capitol Avenue Club have lauded, regardless of whether or not we have Heyward on our fantasy team. Here I’ll outline a couple of the reasons why we like this so much, and also address the primary criticism of it, that Heyward strikes out too much to hit second.
When it comes to lineup optimization, the most exhaustive research into the topic ever compiled in one place is Tom Tango’s The Book
There are also other concerns, because a lot of the #2 hitter’s value comes in the form of being on base in front of other good hitters, you also want him to be a great baserunner. And while stealing bases is okay from that spot, you especially want him to be successful when he does steal bases. Essentially, you want most likely your overall best hitter, who gets on base and is a good baserunner. ie you want Jason Heyward.
However, there has often been one criticism of Heyward in that spot, namely that he strikes out too much. The line of thinking goes that the #2 hitter should be a contact guy, adept at moving runners over. Under this theory, you want a guy who doesn’t strike out much, and slaps the ball behind runners. Under this mode of thinking you want Martin Prado.
Two very good hitters don’t get much different than Martin Prado and Jason Heyward. Prado is a high contact guy who doesn’t strike out much but doesn’t walk much either. Heyward is a high power guy who does strike out a lot, but also walks at a decent rate as well. How can they both be the prototypical #2 hitter? Well, they can’t; Jason fits better.
One of the more overlooked findings in Tango’s book is that out of the #2 spot, of the ways to make an out, strikeouts are actually the LEAST harmful. Huh, wha? Essentially this is because the #2 hitter comes to bat with a runner on first and less than two outs the most of any hitter. The double play is a huge threat from the #2 spot. Think about how a strikeout happens, by definition you had two strikes, you swung at a pitch (or took a tough pitch to hit) and missed. Now imagine making contact on that pitch instead. It will most likely be weak contact, ie contact that is particularly prone to a double play. Contact hitters can be particularly harmful in the #2 spot, as the popular twitter meme “GIDPrado” indicated.
Now, this effect isn’t huge, it is likely only worth a couple of runs over an entire full season. However, the point isn’t that it’s a huge deal, it’s more that striking out (v. grounding out) isn’t a negative out of the #2 hole. If you happen to have a great hitter who does strike out a lot, it’s actually the best place to put him. So, not only is the idea that Heyward strikes out too much for the #2 spot and this will kill our offensive production flawed, if anything it gets it backwards.
Loading ...
I like this explanation, one quick question. By this logic Justin and BJ Upton (maybe even more so) could also be perfect in the #2 hole. What makes Heyward a better option then them? Or is it just that he bats left handed and the braves are committed to Simmons leading off?
This made me think of something else. What would be the ideal lineup spot for each Braves position player? We have spent a lot of time discussing shuffling them around. I’m curious to know the optimal position of each if there were no other constraints. Maybe we have 3 guys best suited for #2 and 3 more best suited for #4 and none for 1, 3, or 5.
Heyward is just a better overall hitter than BJ. BJ’s steals are better leveraged in front of contact guys lower int he order (so 5 is actually near perfect for BJ, as we leverage his power and steals near perfectly). Justin would be great at 2 as well, but the L-R thing works better with Heyward, given that we probably leverage Freeman best at 4. Heyward is a better baserunner than Justin is, and given that they have nearly identical career OBPs, I’d say that Heyward has the slightest of edges due to baserunning.
Another factor with Heyward is that since the #2 spot comes to bat with a man on first, but not on 2nd the most, it further helps leverage the hole created for a left handed pull hitter when a runner is being held on first, more so than a right handed pull hitter.
Hitting Justin 3rd sucks a little bit, but I don’t know how else you’d do it that would work any better.
Makes sense, thanks!
So what you seem to be saying is that–with the exception of the leadoff spot-Fredi has apparently put together the optimal lineup for what he’s working with? That’s bananas.
pretty close. Though getting leadoff right is one of the most important (and most often screwed up) points of lineup construction. If you get 1 and 2 right, everything else is fairly easy.
Why is that crazy? One of the few things Fredi was good at last year was shuffling the lineup.
Would hitting Freeman 3rd and Justin Upton 4th be best? I don’t have the numbers in front of me, but I’m pretty sure FF makes a good bit of contact.
that would then change the handedness of the lineup (2-5)… L/L/R/R instead of L/R/L/R
Optimizing L-R handedness. With Simmons / JHey / JUp / FF / BJ /Mac / Uggs / FatJuan, you get R / L / R / L / R / L / R / L
Random aside: If Francisco is actually slimmer, do we still call him ‘Fat Juan’?
But that really only matters for late innings. Considering batters see starters for many more PAs than relievers, wouldn’t the value of that outweigh the L/R thing? If you’re up 5 runs in the late innings, the matchups won’t matter. Obviously that’s a stretch, as lineup optimizing doesn’t seem to be worth much anyways, but I’d be interested in seeing the value discrepancy between the two. I imagine it would be a pretty tiny gap. None of these guys are people you pinch hit for late in the game, but Heyward is the only one that really becomes a liability against same-handed pitching, with a career 83 wRC+ against LHP.
The difference in runs scored through the first 6 or so innings is probably more substantial than the ones scored from 7-9, even though the later ones, depending on score, may carry a higher WPA.
As with anything baseball strategy related goes, it is smarter to set up for the more likely scenario, and the one that will occur the most often. A SP will face the batters anyways.
it’s fairly complicated, because those late inning situations tend to be higher leverage. Additionally, being able to take LOOGYs out of the equation, or at least limiting them to one batter, can also be really helpful.
My thoughts have been if you’re facing a team without a true LOOGY, then you probably just put together your most optimal lineup and don’t worry about L-R balance. If you are facing a team that has a true LOOGY (or I guess ROOGY), then you’re better off alternating as much as reasonably possible.
Remember that certain pitchers’ L/R platoon advantages are much more extreme than hitters’ are. If you can neutralize an opposing team’s pure specialists, or at least minimize the amount you face them, you’re at a statistical advantage.
Juan is always FatJuan. Always and forever.
Wouldn’t neutralizing what you face the most be more important though? Specialists are exactly that, and nothing more. SP’s have less to neutralize, but will supply the biggest sample to our hitters. So if we can align our hitters to succeed against the bigger sample, wouldn’t that make sense? And if that could result in lower leverage situations late in games due to scoring more runs early, hasn’t the neutralizing already been done? I have a hard time thinking we should have a lineup crafted to neutralize Peter Moylan rather than Clayton Kershaw.
Leverage is entirely separate from lineup optimizing. Regardless of who is hitting, the situation remains the same. Any one pitcher would be guaranteed to face someone of the opposite hand if they face 3 batters. A team could much more easily bring in a LOOGY for two batters then a ROOGY for the third. But how often does that happen? And how many teams have those true match-up pitchers? And how many times will our lineup position fall into those situations (we could go into the inning 5-7 as R/L/R)?
Having the luxury to even consider L/R all the way through is interesting, but I think that garners even less value than an optimized lineup.
THe thing with SP is that they’re going to face all the hitters. Relievers are picked and chosen based on situation. No matter how you arrange the lineup, Kershaw is going to face roughly the same batters. However, later in the game, if you have three consecutive lefties, you can nearly guarantee you’re going to face a LOOGY who is extremely adept at getting out lefties.
Also lineup optimization (excluding L-R balance) isn’t going to make nearly the difference you are acting like it does. It’s like 2-5 runs OVER AN ENTIRE SEASON. So optimizing it isn’t going to lead to noticeably more blowouts, or most likely it won’t lead to any more blow outs. It will MAYBE lead to a couple games where you MAYBE score one more run.
having a couple of same handed batters in a row isn’t terrible, but when you start stringing 3 or more consecutively, it can be very critical late in games in high leverage situations.
Essentially the idea is this: it’s relatively well known and understood that how you order the lineup doesn’t make a whole lot of difference. We also know that lefty (and to a lesser extent righty) specialists have a HUGE, and noticeable advantage over same handed batters.
Unfortunately we pay more attention to the two times Jason Heyward GIDP in high leverage situations than the 30 times he makes pitchers pay from the #2 spot.
I completely understand that both of these things are of extremely minimal value. But baseball is changing to pay more attention to the tiniest changes that we KNOW will give us an advantage over the course of a season. We have the ability to create an optimized lineup. We don’t have the ability to say that Heyward, Freeman, and McCann will be due up in the 9th in a close game, and the opposing team will have a LOOGY, and that LOOGY will be available. If L/R situations are worth 1-2 runs per year, and an optimized lineup is worth 2-3, wouldn’t it make sense to do it that way?
This is all really trivial but I’m more interested in the difference in value between the two than which route the Braves choose.
all studies indicate platoon matchups are much more important than optimized lineups are.
I think this is kind of a moot point because the players are naturally placing themselves into just about the proper spots in the batting order. It’s not like the probable opening day lineup has anyone 2-3 spots away from where they should be hitting. Heyward isn’t batting 7th, Justin Upton isn’t batting leadoff. Against lefties, what should Fredi do? Push all the LHB down to the bottom of the order? The lefty regular that has the hardest time with LHP is Heyward, so if any batter should be realigned in that scenario it should be him. Freeman and McCann have a not-as-horrible split.
I think the presence of the Uptons allows Fredi to keep a consistent lineup that is effective versus lefty and righty starters and–as previously stated–also severely limits the effectiveness of late inning specialists.
In debates with other fans about all the Ks expected this year, I have been dropping knowledge from your article a couple weeks ago with data on the non-relationship between strikeouts and overall team production over time. This GIDP point is just as important as a corollary. It’s absolutely brutal to give away outs simply in the name of “contact.”
Doesn’t exactly apply to Franklins overall point, but there is a good chance the question will come up (also just good general knowledge). Quote from Tom Tango in response to an article yesterday…
“The difference between the absolute best and absolute worst batting order is close to five wins… In any case no one would ever put the absolute worst order together. So, the difference between absolute best and “normally crummy” is 1 to 2 wins.”
Not a lot, but there’s a decent chance that 1 to 2 wins is the difference between winning the East and having to play in the stupid wildcard game again.
and normally crummy is something like your standard Dusty Baker lineup, where your worst non pitcher hitter bats leadoff.
Well thats kinda what the Braves are doing with Simmons
I know the kid is young, but I feel confident that Simmons is at least better than Francisco or Johnson or whoever plays third this year.
Considering that Jason is coming off of a 27 HR year while hitting mostly in the 3 hole, should we expect to see his power numbers dip down to 23-25 due to the new stress on “situational hitting.” I want what’s best for the Braves, but I’d hate to see Jason miss out on a potential 30+ HR season for the sake of giving Justin more baserunners to bat in.
I don’t think Heyward will change his approach based on his lineup position.
That’s one of the things I’ve always wondered: Do hitters actually change their approach at the plate based on where they’re batting in the lineup? I’ve always assumed that hitters go to the plate with the same approach regardless of where they are in the batting order and only change their approach because of the ball/strike count they’re in.
I think some players do, almost always being a change for the worse. However, I think we’ve seen Heyward hit 2nd a lot, so I’m not terribly worried he’ll change his approach dramatically.
Players also change their approach in certain game situations. With nobody on, it’s pretty straight-forward. With a man on third and one out in a late game, they might try to be more aggressive.
Certainly they have to have different approaches in their quiver. One of them being the one Mark pointed out. Simmons will likely have a different approach batting first than he would batting 8th. BJ Upton has a career OBP of .336 and would be a good leadoff guy. But last season, the Rays experienced lots of injuries to power bats and Upton was their only real threat in the lineup. I would think that he became more aggressive attempting to hit for power, coupled with being the main focal point in the lineup for opposing pitchers, that resulted in a .298 OBP.
He has had a season of .273/.383/.401 for a 116 wRC+ and a season of .243/.331/.429 for a wRC+ of 113. I’d have to think that his approach due to protection and lineup position resulted in entirely different, but similarly productive seasons at the plate.
Andrelton Simmons in an interview was talking about how he changes his approach based on his spot in the order. He was a 2 hole hitter in the minors I think, and they moved him to 8th. So he was either forgetting to bat like an 8th hitter or trying too hard to make something happen with the pitcher behind him in a lot of games last year when he was playing. At least thats what I took from it. No idea if that’s normal, though.
I lot of this discussion is based on total taken over an entire season, but lineup construction should be done by taking into account what can happen in a individual game.
For example, perfect L/R balance may only lead to a very tiny amount of additional runs, but the ONE game it stops a manager from using a single LOOGY to face both Heyward and Freeman makes it worth it.
The problem with a lot of sabermetrics is that they look at large sample sizes. I know that sounds weird, but MLB baseball games are won and lost becaue of ONE move, or ONE hit, or ONE run. Scratching out a single run in a pitcher’s duel is far more valuable than adding on a useless runin a blowout.
Not allowing a LOOGY to face both Heyward and Freeman could very well win or lose a single game, and if that game is the 163rd game of the season, we all go home upset. Striking out with a runner on 3rd may not be considered bad by statheads, but if it causes Verlander to beat the Braves 2-1 in the 7th game of the WS we all complain about too many strikeouts.
Sometimes guys need to perform ONE time NOW, not 30/100 times over the course of a whole month.
So if the optimized lineup is responsible for a run in this particular game, and we go up 4-0 into the ninth in an away game. (This would never happen because we have Craig Kimbrel) The home team puts up 3 runs in the ninth but still loses the game. Just that once, an optimized lineup was responsible for a win. That’s the thing. They both take a perfect storm to show their value. They’re both nearly meaningless. But things should be measured over the course of a season, because that’s the sample we are given. If we lose the division by one game no one here will be pointing to lineup optimizing or L/R organizing.
n optimized lineup isn’t going to cause that. Having Justin bat one spot later is not going to cause a difference at all. However him batting in between them and preventing a LOOGY from striking out two to start the inning makes a big difference late in games.
An optimized lineup COULD cause that. That’s why it’s measured over the course of a season. This may be the game where that run shows up, and it COULD be responsible for a win. And in this specific instance, it doesn’t really matter if the LOOGY strikes them out back to back to start the 9th, as we would already be up 4-0. It’s a stretch, yes. But it is entirely possible. Just as L/R organizing could possibly yield a win.
I know this will never happen, but I think the optimal lineup is
1. Justin Upton
2. Heyward
3. B.J. Upton
4. Freeman
5. Uggla
6. McCann
7. Simmons/Johnson
8. Francisco/Simmons
The 1,2, and 4 spots should generally be the best three guys on the team, especially since we don’t really have a high OBP no power guy for the leadoff spot. Although BJ, McCann, and even Uggla are still good enough to have a monster year, Freeman, Heyward, and Justin should be our three best hitters. They’re better in every major projection system and they’re younger and less likely to decline from aging and more likely to improve. To keep L-R balance, that necessitates putting Justin first, and Heyward and Freeman 2nd and 4th. After them, BJ fits the bill best for the three spot since he is the closest thing to a high power, low on base player (though it would be nice if his OBP rebounds). Then Uggla and McCann 5 and 6 to keep balance, though I think McCann is a slightly better hitter than Uggla. After that, the bottom doesn’t really matter, so I just went with L-R balance. This lineup puts our three best players in the three most important slots and keeps our six ‘good’ hitters in the first six spots, rather than Fredi’s plan that will force Uggla down to 7th. What do you guys think?
They make a thing for this… This is every position player with their wOBA and SLG input, and the numbers that support the best lineups. I used Juan Francisco as the 3B and Medlen as the P.
http://www.baseballmusings.com/cgi-bin/LineupAnalysis.py?Player0=Jason+Heyward&OBA0=.350&Slug0=.447&Player1=Freddie+Freeman&OBA1=.343&Slug1=.449&Player2=Andrelton+Simmons&OBA2=.324&Slug2=.416&Player3=Dan+Uggla&OBA3=.352&Slug3=.469&Player4=Justin+Upton&OBA4=.359&Slug4=.475&Player5=Juan+Francisco&OBA5=.320&Slug5=.440&Player6=BJ+Upton&OBA6=.333&Slug6=.422&Player7=Brian+McCann&OBA7=.354&Slug7=.475&Player8=Kris+Medlen&OBA8=.177&Slug8=.151&Model=0
I don’t think Uggla is automatically a “better” hitter than Simmons. Andrelton still has a lot to prove, but he seems to have a good swing and could easily beat Uggla’s performance from the past two years, which was pretty bad.
We all hope Uggla will improve, but there is no guarantee that he will. Likewise, there is no guarantee Simmons will develop into the player we all hope he will be, but I’m all for putting him at the top of the lineup where he will see a lot of pitches and give him the chance to show what he can do.
5 spot is more valuable than 3 if I recall…
I think that a healthy McCann is a much better hitter than Uggla. Uggla’s swing is just too inconsistent. Especially in important at-bats, which he didn’t see a lot of in Florida. IF McCann is healthy I think his contributions will make the difference in wether the Braves make a deep run or not.
On your podcast you talked about the meaning of 80 on the 20 to 80 scale.
When scouts talk about a prospect or a major leaguer being a five tool player how does that translate in terms of the 20-80 scale?
Would every tool have to be a minimum of 65, for instance.
Can you use the 20-80 scale to then give more accurate readings of 5 tool, 4 tool players?
Usually 5 tool is applied to any player with all 5 tools at a 50 or greater.
It shouldn’t even be questioned that when healthy McCann is a much better hitter than Uggla (also while healthy of course). Uggla simply has more power. Agree, disagree?
Lemke is only doing spring training games, right? I love the guy but he’s so hard to listen to for more than 2 minutes.