Considering Tandem Starters in St. Louis

Is there any way that an idea that makes some sense in A-ball, the tandem starting rotation, could be applied to the St. Louis Cardinals at the major league level?

On more than one occasion recently, I have been asked to declare my view of what might/could/should happen with the St. Louis Cardinals starting rotation in the upcoming weeks.

As many Cardinals fans know, right-hander Joel Pineiro is nearing his return from the disabled list and Mark Mulder is reportedly about a month away. While the current five-man rotation has gotten out of the gates quickly, two of them will soon be facing role changes unless trades occur or injuries strike.

The question is, "which two?"

Of the five, I will consider Adam Wainwright and Kyle Lohse firmly entrenched, with the other three, Braden Looper, Todd Wellemeyer and Brad Thompson, fair game for debate and discussion. I am also including Anthony Reyes in this work, since he is traditionally a starter, as well.

When I first considered this article, I thought I would review factors like minor league options remaining and ability to perform in a relief role. But, the reality is that mound performance in the upcoming days will most likely play heavily into the decision of Tony La Russa, Dave Duncan and John Mozeliak.

Yes, the general manager is listed here for good reason. If one can believe the reports, the primary reason Reyes is in the majors today is that the GM overruled his coaching staff and made the call to keep Reyes, though it did mean a change in role from starting to relieving.

It is worthwhile to note that Thompson and Reyes do have one minor league option year remaining, making them easiest to send down from a business perspective. Yet, two pitchers at the back of the pen, Kelvin Jimenez and early-season darling Kyle McClellan, do too.

And this doesn't even take into account relievers Tyler Johnson and Josh Kinney, each working to come off the disabled list, though the return of neither seems imminent. Both do have minor league options, increasing flexibility.

Bottom line, there is a lot of pitching on the way to St. Louis, perhaps justifying a different approach to best utilize it all. More on that in a bit.

Career pitching stats

None of the above would lead to a satisfyingly clear conclusion as to who should stay and who should go from the rotation. So, I moved to looking at career pitching splits data from the four in question prior to 2008.

I began with the most basic of results, ERAs when starting versus relieving. Only in the case of Reyes, having just 12 innings in relief before this season, is there not at least a fair amount of data from these pitchers in both roles.





Starter 18 94.2 4.75
Reliever 109 146.1 3.51




Starter 13 61.1 3.38
Reliever 124 192.2 5.33




Starter 38 195.2 5.34
Reliever 7 12 7.50




Starter 31 180.2 4.78
Reliever 573 607.2 3.58

It could be argued that Thompson's and Looper's relief results are superior to their returns in a starting role with just the opposite for Wellemeyer and Reyes.

Next is to look at career runs allowed by inning. This is enlightening but imprecise for our purposes as runs allowed in middle innings could have been either in a starting or relief role.

Also, these are all runs scored, whether earned and unearned, which might skew the data a bit. Finally, partial innings are not accounted for, which in aggregate means these "RA" or runs allowed can only be considered an inflated, rough approximation of a real ERA.

Still, there are some interesting trends.





1 19 12 5.68
2 21 10 4.29
3 21 14 6.00
4 22 7 2.86
5 24 16 6.00
6 35 17 4.37
7 52 21 3.63
8 52 15 2.60
9 25 7 2.52
10+ 6 1






1 14 8 5.14
2 19 9 4.26
3 21 13 5.57
4 30 22 6.60
5 41 20 4.39
6 46 22 4.30
7 46 19 3.72
8 40 27 6.08
9 21 4 1.71
10+ 10 12






1 38 35 8.29
2 37 12 2.92
3 37 31 7.54
4 37 13 3.16
5 35 11 2.83
6 22 11 4.50
7 8 2 2.25
8 3 1 3.00
9 2 2 9.00




1 31 19 5.52
2 31 13 3.77
3 31 14 4.06
4 31 19 5.52
5 35 10 2.57
6 50 27 4.86
7 104 49 4.24
8 254 127 4.50
9 282 119 3.80
10+ 51 15


Thompson has typically had two bad innings, the third and fifth, but generally his ERA trended downward throughout the game. Wellemeyer seems strongest in the fifth through seventh frames. Once past the third inning, Reyes seems fine. Looper has no obvious pattern, other than problems in the first, fourth and sixth.

Now we'll look at batting average, on-base and slugging percentage of the opposing hitters during the pitchers' first time through the batting order, then the second and finally, the third.

1st time 0.258 0.327 0.406
2nd time 0.367 0.438 0.593
3rd time 0.275 0.318 0.480
1st time 0.259 0.368 0.402
2nd time 0.238 0.325 0.464
3rd time 0.190 0.306 0.429
1st time 0.274 0.358 0.524
2nd time 0.243 0.318 0.472
3rd time 0.233 0.301 0.393
1st time 0.264 0.327 0.376
2nd time 0.240 0.291 0.343
3rd time 0.319 0.374 0.573

Thompson generally struggles after the first time through the order. Reyes' initial batters faced are his worst and while his averages are in check, his opposing slugging percentages are high. Wellemeyer generally improved each time around though his on-base marks are inflated due to walks while Looper is really only good for two times through the opposing lineup.

The final data view is how the four pitchers have done based on the number of pitches thrown. Hitting results are listed in groupings of 25 pitches. Not surprisingly, results are comparable to the turns through the order listed immediately above.

Pitch cts BA OBP SLG
1-25 0.250 0.320 0.396
26-50 0.351 0.420 0.538
51-75 0.316 0.347 0.516
76-100 0.263 0.344 0.491
Pitch cts BA OBP SLG
1-25 0.262 0.364 0.408
26-50 0.253 0.370 0.447
51-75 0.221 0.302 0.356
76-100 0.176 0.316 0.500
Pitch cts BA OBP SLG
1-25 0.300 0.389 0.638
26-50 0.250 0.312 0.411
51-75 0.227 0.302 0.454
76-100 0.228 0.318 0.386
101+ 0.286 0.375 0.286
Pitch cts BA OBP SLG
1-25 0.266 0.325 0.378
26-50 0.231 0.310 0.328
51-75 0.284 0.348 0.459
76-100 0.333 0.385 0.578
101+ 0.133 0.133 0.200

Thompson looks solid through his first 25 thrown and then heads south. Wellemeyer shows some improvement, and once Reyes gets past the first 25 pitches, he is better. Looper seems to lose effectiveness after firing 50 bullets.

Can they change?

Much has been said and written about the need for all four of these men to pitch more economically, to get their outs more quickly in the count, thereby working deeper into games.

I find it impossible to argue with the logic, but wonder how well it can actually be executed.

After all, let's take Todd Wellemeyer. He has been a professional pitcher since 2000 and first appeared in the major leagues since 2003, having thrown over 750 innings. Don't you think he has wanted to get his pitch counts down for years and issue fewer walks? Wouldn't you think he could have accomplished a transformation by now if he could?

Seems more likely that he is what he is, which isn't necessarily all that bad.

My intent is not to pick on Todd or anyone else, but instead to suggest that the Cardinals might consider accentuating what these men do well rather than trying to get them to become someone they maybe can never be.

The big idea

Now that I am this far into the article, I can pose the idea, as ludicrous and unlikely as it is of ever actually being implemented.

What if the St. Louis Cardinals implemented the tandem starter system at the major league level – not for their entire rotation, but for its back end?

With the possible exception of the closer, aren't any team's best pitchers their starters? Why not maximize their use in the job for which they were groomed? The Cardinals have an excess of starters, so why not continue to use them in that role while keeping the best 12 pitchers on their staff?

I couldn't help but consider this as Wellemeyer was sent out to pitch the seventh inning against the free swingers from Houston on Monday night. With Todd spinning a shutout through six, wouldn't you rather have stepped away from the table with money still in front of you?

Wouldn't it have been good to see another starter, all prepared to enter the game, come in? The alternative on that evening was to witness a tiring Wellemeyer give up a pair of home runs and turn it over to Jimenez, who surrendered one of his own.

On Tuesday, Anthony Reyes spun three shutout innings in relief amid ongoing concerns that his intricate pre-game preparation ritual does not align well with the spontaneity of being called upon to warm up quickly as a reliever.

In addition to a rotation that may not be able go deep into games, the 2008 team features an effective, but aging bullpen. Risk of overwork as the season progresses is relatively high.

Raise your hand if you are looking forward to a reprise of a 2007, a year in which now ex-Cardinals relievers Brian Falkenborg, Dennis Dove, Troy Cate, Andy Cavazos and the like took their turns getting spanked by the opposition only to be returned to Memphis upon detonation. The situation became desperate enough that position players Aaron Miles and Scott Spiezio made a total of three mound appearances.

What is the tandem system?

The tandem (or piggyback) system was implemented at the lower levels of the Cardinals minor league system last season after being deployed in the Texas Rangers and Cincinnati Reds systems earlier in the decade, among other places.

Its primary intent is to allow the organization to prepare as many starters as possible by pairing two pitchers. They take turns starting every fourth day, with the other following in relief. The starter might go five innings or 70-80 pitches, with the second starter scheduled to take over in the sixth.

The closer is still on the roster, as are three other relievers to complete the 12-man staff. This is because in the minors, there are typically only four pairs of starters, going every fourth day.

When implemented in the lower levels of the Cardinals minors in 2007, there was some grumbling about the system, especially from certain baseball traditionalists. Pitchers don't get to stretch out; they don't get the chances to work themselves out of jams; etc., go some of the criticisms.

Yet when all is said and done, the folks in power in the Cardinals player development function must view it to be a success. After all, the approach is being utilized at every level of the minors other than Triple-A and Double-A this season.

How might it be done in the bigs?

On the positive side, the pitchers' days on the mound are set in advance and all eight can prepare just as normally four starters would. That could be one supporting reason for its implementation at the major league level.

Looking at the first change, upon the return of Pineiro, the rotation could look like this:

Tandem pair #1
Tandem pair #2

The numbers, especially ERAs as starters and relievers, might suggest a variation of the tandem where one pitcher is always the first starter and the other in the pair is always the second starter. Wellemeyer and Reyes starting with Thompson and Looper serving as the second starters may accentuate each pitcher's strengths.

Yet when also considering 2008 results, I might actually flip-flop Reyes and Looper, with Looper and Wellemeyer as the starters and Reyes and Thompson always scheduled to follow. If Pineiro falters after his return, there is no reason he couldn't enter this tandem mix later on, too.

Making them fit

The above would require seven roster positions for starters. Filling out the staff with Jason Isringhausen, Ryan Franklin and the two lefties, Randy Flores and Ron Villone, would leave one spot for Russ Springer. In the meantime while Springer is on the DL, McClellan and Jimenez could fight it out. Another option might be to go with just one left-hander, since situational matchups may be less prevalent in this model, at least during tandem days.

One risk would be a light bullpen on the days the non-tandem starters would be going. If the top three traditional starters can't regularly pitch deep into their games, the pen could be stressed. Yet under this scenario with a good supply of relievers potentially parked in Memphis just a phone call away, this shouldn't be an ongoing problem.

Looking ahead

Let's think ahead this summer to the return of Chris Carpenter. Might it be better to ease him back into a full workload by participating in the defined structure of a tandem system this season instead of less planned, less controlled use out of the bullpen?

If Matt Clement makes it back at all, a shorter, less demanding role may also have considerable advantages to both him and the Cardinals.

Lets get real!

Would Tony La Russa and Dave Duncan ever mess with their guys' roles and psyches enough to attempt such a maneuver? Not likely. Not likely at all.

On the other hand, La Russa and Duncan do have a history of innovation. Their deployment and refinement of the ninth-inning relief specialist 30 years ago led to the one-inning closer that is the norm across the game today.

La Russa's return to a past experiment of having the pitcher bat eighth, with the ninth-place man a "second leadoff hitter", is taking root. The approach has been validated by sabermetric experts as having a slight advantage over the traditional pitcher batting ninth and it has been implemented by a second MLB manager, Ned Yost in Milwaukee.

Since La Russa favors a second leadoff hitter, couldn't a second starter be palatable, too?

Maybe, just maybe, given the unique nature of their 2008 pitching staff, La Russa and Duncan should look deeper into the crazy idea of a tandem starter approach for these St. Louis Cardinals. Not forever, but perhaps for awhile.

Brian Walton can be reached via email at

© 2008 All rights reserved. This material may not be published, rewritten or redistributed.

The Cardinal Nation Top Stories