How Skip Schumaker at 2B is a Pivot for the Cardinals’ Offense
FORT LAUDERDALE, Fla. — The best way that St. Louis Cardinals manager Tony La Russa believed he could keep Skip Schumaker’s bat in the lineup was to put his glove to use elsewhere.
The official pronouncements came in the past week from GM John Mozeliak and La Russa that Schumaker will open the regular-season as the Cardinals’ starting second baseman — will, in fact, be now designated on the roster as a capital-I Infielder. Schumaker openly concedes he is hardly a finished-product in the field. How quickly he can improve on-the-job will determine the number of outs the Cardinals give away on double plays and other infield moments a seasoned second baseman might handle differently or better.
What is clear, however, is the number of runs the Cardinals avoided giving away by keeping Schumaker, the team’s only proven leadoff hitter, in the lineup and gaining another outfielder.
That’s the trade-off La Russa was willing, and trying, to make.
How to illustrate that? Over at Baseball Musings there is a “Lineup Generator” that allows you to type in pertinent stats — OBP and SLG, in this case — and generate an “ideal lineup.” The math behind the generator is fond of having Albert Pujols bat first because of his high OBP and the more ABs over the course of a season. But it’s not the order that helped understand the Schumaker Effect yesterday as I punched in numbers. It’s the runs. With each lineup the generator spits out, it also offers the number of runs that lineup would average. Far from perfect, it does give a spectrum that shows a team’s simplified offensive output.
Last year’s lineup scored 779 runs, fourth-most in the National League and an average of 4.81 per game. Throw the regulars and their stats into the generator and out spits the follow average and range:
5.370 average, 4.774-5.581
Clearly the regulars weren’t available for an entire season — injuries, mostly — and many other variables go into a season, as you know. So keep that in mind as we use the same generator to show how the lineup production shifts and drops depending on if Schumaker is in it or not.
For the OBP and SLG used, I consulted to the PECOTA predictions in Baseball Prospectus and at the Baseball Prospectus web site. While PECOTA romances a few players (Joe Thurston, for example) and is oddly pessimistic about others (Brian Barden, for example). It’s a well-rounded projection though and again helps to illustrate how production rises and falls with personnel.
Take the possibly Opening Day lineup, for example: Schumaker, Rick Ankiel CF, Albert Pujols 1B, Ryan Ludwick RF, Chris Duncan LF, Khalil Greene SS, Yadier Molina C, David Freese 3B, Average Pitcher. Here is what the generator offers:
4.867 average, 4.391 to 5.083 range
How those numbers change with a different second baseman in the lineup, and Schumaker in the outfield:
- with Brendan Ryan at 2B … 4.743 average, 4.291 to 4.987 range.
- with Joe Thurston at 2B … 4.802 average, 4.366 to 5.037 range.
- with Brian Barden at 2B … 4.689 average, 4.215 to 4.215 range.
If you sub Troy Glaus into the lineup and rerun the numbers, the Cardinals lineup powers up to a level more comparable to last year’s numbers.
With Glaus, and Schumaker at 2B:
- 1998-2002 model … 5.009 average, 4.610 to 5.294 range.
- 1959-2004 model … 5.099 average, 4.745 to 5.185 range.
With Glaus, and others:
- Ryan at 2B … 4.770 average, 4.390 to 5.085 range.
- Thurston at 2B … 4.839 average, 4.459 to 5.145 range.
- Barden at 2B … 4.635 average, 4.305 to 5.024 range.
While the difference is clear, the decimals get kind of dizzying. So, it’s almost better to take a look at them in terms of total runs. Again, this is far, far from exact because it takes two raw numbers, assumes health, assumes consistent playing time and calculates a total range. Still, it shows how one lineup played thousands of times compares against another lineup a thousand times.
The difference, in total runs over a season, of the current lineup (read: Freese at third base), depending on the second baseman:
- Schumaker … 788
- Thurston … minus-10
- Ryan … minus-20
- Barden … minus-28
Depending on who you throw into the lineup — Colby Rasmus, for example — the numbers stay generally consistent. It’s a give/take of 10 or so runs. Ten runs with Schumaker pegged, conservatively, for a .287/.343/.384 season. Ten runs on offense. Ten runs according to the dozens of lineups that the generator cranks out and calculates. The highest-scoring of which … has the pitcher hitting eighth.
-30-
Derrick Goold said he was going to Mizzou for capital-J journalism, but after growing up in the Time Zone Baseball Forgot he was really drawn to MU sitting between two major-league cities. Goold joined the Post-Dispatch in 2001 after working for The Times-Picayune and Rocky Mountain News, covering sports from LSU to NHL and every level of baseball in between. 
Makes perfect sense to me. Some has to hit singles to get on base in front of all the power hitters. Schu is that guy. I don’t know why he has trouble getting the credit he deserves.
The more interesting number would be what the net gain/loss is using Schumaker over the other (arguably better defensive) options. Schumaker’s defense will cost the teams some runs over one of the other 2B options. Is his bat enough of an improvement to cover the lesser defense?
Top notch work DG! This stuff is very interesting to a math/stats guy such as myself.
One question- How many runs per season do you think Schumacher would give up with his glove compared to those other guys?
As I was reading this, I was asking the same questions as the comments above mine. Is the bat enough to compensate for his lesser defensive play? How many more runs per season would Shumacher give up as the starter, opposed to the other second base options?
Anyhow, Great work. Keep it up!
Cannot really put out a calculator and punch up that number. Going to have to watch the season and come up with an educated sense of how many runs there are given or taken away defensively.
But I still don’t see how this version of the team is significantly BETTER than last years? Runs last year 779 v. runs this year with optimal line-up 788–unless Glaus’ return generates a lot of runs, aren’t we still 3 or 4th best? Will the changes in defense provide the difference with Skip newly in the infield? Will the pitching cover it with Carp and Wainwright iffy? Where is the vast improvement wanted by fans and sort of vaguely promised by management? Can we get to the play-offs with this line-up?
Good call @Jodi and @CardsFanTX, you’re BOTH absolutely right.
This is not a significant improvement over last year offensively and it’s a **downgrade** defensively from last season. And, let’s remember, last season finsihed off with the Cards in 4th place in the Central Division… Meanwhile there is still no closer and the lefties signed have enjoyed spotty success at best. Hey, at least Albie is ending the spring on a 4-for-24 stretch…
Can’t wait to be wrong, sure hate to be right, but this team ain’t gonna catch a whiff of October without making some moves.
I think this is in accurate.
Its not skip numbers above ryan or thurston. Its essentially Duncans numbers above ryan/thurston.
If skip doesnt play 2b, he starts in the outfield instead of duncan.
Skip would be in the lineup either way, its just a matter of putting in a power outfield bat instead of a weak infield bat from our bench guys.
One way to get around the playing time or “regulars” issue for last year’s stats by going to baseball-reference.com and use the actual splits for each position in the batting order.
Doing so, you’ll find that last year’s team underperformed their “expected runs scored” by around 66 runs. This is due to bad luck, GIDPs, sacrifice bunts, etc, all of which reduce scoring and none of which are considered by the lineup generator.
Here’s a link to a VEB thread from this past october where STLfan and I do this analysis for all the teams. STL was the second-most underperforming team when compared to expectations, next to washington (at a staggering -122 runs).
Anyway, the point of this is, even if we have a few players like Luddy regress, we should score as many or more runs next year simply due to not having such extremely bad luck.
The Lineup Generator model likes the pitcher hitting 8th with Yadier Molina in the 9th spot as the second lead off hitter.
Don’t tell Tony.