Having a little fun with the pitching staff

[IndianaTwin]

A couple weeks ago, a friend and I were discussing a player’s HOF credentials and noted that the quality of a guy’s Strat-O-Matic card is a good snapshot of a guy’s dominance. Anybody remember that Barry Bonds card from 2001? Anyway, that also reminded me of the year that I dominated my league by using starters for only a couple innings so that I could maximize my dominant relievers.

So, I’ve been batting around an idea in my head, knowing that no team will actually use it, but wondering how it could actually work.  It grows from a stat on MLB Network yesterday saying that OPS goes up by 66 points with the pitcher’s third time through the lineup.

So, indulge me, and let’s have a little fun with it on a holiday weekend. How about a return to the four-man rotation, with each starter having the goal to throw to only 18 batters? For a pitcher with a 1.20 WHIP, that gets them through five innings. For a pitcher with a 1.28 WHIP, it gets you through 4.2 innings. So, the plan is that if the starter gets through 18 batters and there are no runners on base, he gets to finish the inning. On occasion, that may even mean finishing the sixth.

Here’s how that shapes up on this year’s schedule. Starter 1 makes 23 starts on three days’ rest and 17 on four days (including his first start), plus seven days of rest over the All-Star game. Starter 2 makes 25 on three days and 14 on four days (plus one on seven days and one on five when they have two off days in the same week). Starter 3 makes 24 on three and 14 on four (plus one on seven and one on five). Starter 4 makes 23 on three and 16 on four (plus one on seven). So, the reality is that about 41 percent of the starts come on what is now seen as “normal” rest. And no one ever goes long into a game, probably about 80 pitches a start, so they should be able to come back on three days’ rest. At 40-41 starts apiece, the average five innings from a 1.20 WHIP gets a guy to 200 innings for the season.

For argument’s sake, let’s put Santana, Hughes, Millone, and Gibson in the rotation. Then, pair them up with Duffey, May, Berrios, and maybe Meyer. Since so much of pitching is throwing off timing, I tried to match them for the biggest contrast. So, after getting two at bats off Millone, hitters get greeted by Berrios the third time through, etc. You folks who have a chance to see them more often than me might have better matches to screw up the timing. Suggestions?
In general, these relievers are all guys who can go as far as four innings. On many occasions, they could be expected to finish up the game in those combinations. If it seems that the starters are getting worn down at times, protect them a bit by flipping the role. Say Hughes is wearing down – just start May for 18 batters and use Hughes in relief for a couple times through the rotation, etc.

There is also some room for specialists in this plan. Let’s assume that the relief quartet is solid enough to go four innings and get the three-run saves. For one and two-run saves and other high leverage situations when a reliever is needed in the eighth and ninth, let’s reserve Jepsen and Perkins for that.

There will also be a few cases when the starter gets hammered in his first two times through the lineup. If the offense has scored enough runs to keep the team in the game, just bring in the reliever in an inning earlier. Hopefully this happens rarely enough that the reliever may be able to go five innings, or long enough to get to Jepsen/Perkins for the closer. And if the team is out of it, bring in the take-one-for-the-team guy: Nolasco. (I’d actually consider Nolasco over Meyer for the fourth spot in the relief quartet, but it’s not a real TwinsDaily post without taking a potshot at Nolasco, so I’m using him here.)

And one more specialist. The relief quartet is all guys who are used to start innings fresh, as are Perkins and Jepsen. So save a spot somebody like Fien or Tonkin, whose role it is to be used in whatever situation to finish off an inning and get back to the relief quartet, Jepsen, and Perkins starting innings. That could mean finishing off the rare third or fourth inning when the starter gets hammered or the fifth (or even sixth) when he doesn’t quite get through it, and the sixth through eighth when the relief quartet guy doesn’t quite have it.

 
It never works out so cleanly, of course, but it seems there is enough room for flexibility to shuffle roles around. For example, if a starter ends up on the DL, there are four relief quartet guys who could step into the rotation with minimal stretching out. If Nolasco pitches himself out of the take-one-for-the-team (in a good way, I mean), he could get promoted to the relief quartet in place of the guy most deserving of relegation (think European soccer). Studs in Rochester deserving of promotion would step into lower-pressure roles of the relief quartet, rather than coming up to a role that is expected to get six or seven innings. 

I think some teams have tried some variation of this approach in their minors. It seems like the Twins current staff is the type of staff that it would best with, since there’s not the dominant starter that is going to get you 7-8 innings every time out.

So, other than the fact that it’s different enough that no manager would try it, could it actually work in practice? What am I missing? Any other wacko ideas to trot out on a holiday weekend with nothing else to do?

[Thrylos]

Take away the faced batters restriction, cut your long relievers in half and you are back to the 80s    It worked then, but there were more off-days that allowed the starters to go with 5 days rest.  There was a swingman who would start the 5th game if necessary.

[drjim]

If I lacked a dominant, Kershaw like guy, I think the optimal staff could be:

 

4 starters for 18 batters or 80 pitches

3 second guys that rotate every third game for 2-3 innings

1 mop up/swing guy for blowouts and low leverage

3 setup guys that can go 1/3-2 innings depending on matchup (with at least one lefty)

1 closer

 

I think we see this type of strategy sooner than we think.

[DocBauer]

Actually, it's a very cool idea and has real, logical and practical merit. Though I disagree that the "relief starter" has less pressure than the game starter. In practical terms, each starts a game, or an inning, and throws roughly the same number of pitches/innings.

 

I think the biggest flaw in the idea is one you mentioned yourself. That is, the "bad day" scenario. You referenced a game starter coming out of the game early and being replaced early by a relief starter. But what if the game starter is cruising along, is pulled after 4/80 as per normal, and the relief starter has the "bad day". So we've replaced the cruising starter, put in an ineffective pitcher, and then have to turn to one of the other 4 "extra" relievers. Hopefully, he/they dont have a bad day. Does this staff makeup allow for a LOOGY as well as a closer? And does a LOOGY-type specialist mess up the normal scheme of things?

 

All I'm saying is, the idea is very cool and makes a lot of sense. But long term, day after day, week after week, if the flow of the plan/operation hits enough snags, I worry about a perpetual effect that brings down the Genga tower. Lol

 

But again, I think the idea is very cool.

[jorgenswest]

I like your ideas. I think that pitching usage will evolve again.

 

I do wonder about the third time through the line up data. The average start for the Twins was was around 24 batters faced. I would guess that one of the factors driving the difference in the OPS is that there were several instances where the pitcher didn't make it to batter 27. The third time through data would be weighted with more plate appearances from the top 1/3 in the line up rather than the bottom 1/3. A pitcher will almost always make it through the first time around and the plate appearances from the top and bottom of the order would be close to the same. The imbalance caused by not making it to batter 27 must be part of that difference in the OPS. The third time through data has a larger ratio of at bats from the better hitters.

[theBOMisthebomb]

The big flaw is the impact it has on guys like me that struggle to get to the ballpark on time when dragging along the wife and children (they just don't always move quickly). I'd be bummed to only see a high quality starter for an inning or two due to my tardiness.

Seriously, I think it is possible for your plan to be implemented. Think of innovations in sports and the resistance they almost always face - it just takes one maverick to try it with success and then here come the copycats. I do think the Twins currently have a staff where this could work.

[IndianaTwin]

 

If I lacked a dominant, Kershaw like guy, I think the optimal staff could be:

 

4 starters for 18 batters or 80 pitches

3 second guys that rotate every third game for 2-3 innings

1 mop up/swing guy for blowouts and low leverage

3 setup guys that can go 1/3-2 innings depending on matchup (with at least one lefty)

1 closer

 

I think we see this type of strategy sooner than we think.

Thanks for the response. Your variation has a quite similar feel. With not having the Kershaw type, it seems like the 2016 Twins could be a good fit for this approach. That I had to suggest Meyer/Nolasco for the fourth in the reliever quartet might make your approach of three second guys more feasible. That also keeps one more guy available for situational stuff. What I liked about the quartet in relief is two-part: 1) it has the expectation of sometimes getting through the game on two pitchers, which saves the end guys; and 2) it gives opportunity for a structured change in pitching style, which I think could be pretty significant. Whether it's lefty to righty or offspeed to fastballer, it could increase the second guy's effectiveness even more. 

[IndianaTwin]

 

 

Actually, it's a very cool idea and has real, logical and practical merit. Though I disagree that the "relief starter" has less pressure than the game starter. In practical terms, each starts a game, or an inning, and throws roughly the same number of pitches/innings.

I think the biggest flaw in the idea is one you mentioned yourself. That is, the "bad day" scenario. You referenced a game starter coming out of the game early and being replaced early by a relief starter. But what if the game starter is cruising along, is pulled after 4/80 as per normal, and the relief starter has the "bad day". So we've replaced the cruising starter, put in an ineffective pitcher, and then have to turn to one of the other 4 "extra" relievers. Hopefully, he/they dont have a bad day. Does this staff makeup allow for a LOOGY as well as a closer? And does a LOOGY-type specialist mess up the normal scheme of things?
 

Thanks for the feedback. My sense on the second guy having less pressure is that he's typically asked to through four innings (or three to set up the closer) rather than five and that he doesn't have the perceived pressure of starting. But you're also right in that the second guy is potentially pitching in higher leverage system.

 

In some ways, the system takes care of the bad day/good day effect. It takes a 1.50 or worse WHIP to get through 18 batters and not also get through more than four full innings. In most games, the starter will be into the fifth by the time he gets to the 18th batter. The approach I'm suggesting is that if he opens the fifth by getting the No. 8 and/or 9 batter out, he gets the chance to finish the inning. If he's really cruising (it works to a WHIP of 1.13 or better, or getting help from a double play) he'll get to begin the sixth, and again, if he gets the 18th batter out to open the sixth, keep him in there. See next comment on bringing in a "bad day" reliever.

[IndianaTwin]

 

 

I like your ideas. I think that pitching usage will evolve again.

I do wonder about the third time through the line up data. The average start for the Twins was was around 24 batters faced. I would guess that one of the factors driving the difference in the OPS is that there were several instances where the pitcher didn't make it to batter 27. The third time through data would be weighted with more plate appearances from the top 1/3 in the line up rather than the bottom 1/3. A pitcher will almost always make it through the first time around and the plate appearances from the top and bottom of the order would be close to the same. The imbalance caused by not making it to batter 27 must be part of that difference in the OPS. The third time through data has a larger ratio of at bats from the better hitters.

Thanks for the feedback. I teach my kids to always look for an alternative explanation for conclusions drawn from statistical data. You found one with the observation that the third-time-through data has a disproportionate number of PAs from the top of the order. To some extent, that will happen in the second time through as well, though likely not as often. I suppose the way to correct for this is to drill down a step and see what the difference is on a lineup spot by spot basis. That is, does a No. 1 hitter's OPS go up the third time they face a starter, etc. Any idea where to find that data?

 

Counterbalancing that, I wondered if the third-time-thru data is also disproportionately affected by having better pitchers included. Whether it's good pitchers that disproportionately get to the third time in the order or so-so pitchers having a good day, it seems like the pitcher facing hitters for the third time is at least perceived to be effective that day. If a guy is getting shelled, he has fewer chances to face batters a third time.

 

At any rate, I did find that the league average last year was for starters to have .723 OPS against the first time through and a .727 the second time through, but to jump to .751 the third time through. What jumped out at me, however, is that the OPS against for a reliever in his first time through was .703. That may be affected by having a disproportionate number of PAs against the bottom of the order, but it applies to all relievers, so it includes second, third, and fourth relievers as well, when the lineup has turned over. If the difference is indeed genuine, why not maximize the number of at bats against a reliever by asking them to go more than an inning?

[IndianaTwin]

 

 

The big flaw is the impact it has on guys like me that struggle to get to the ballpark on time when dragging along the wife and children (they just don't always move quickly). I'd be bummed to only see a high quality starter for an inning or two due to my tardiness.
 

 

Been there, done that! It gets better though -- at some point, it's your kids hounding you to get there by the beginning of BP. :-)

[DocBauer]

Another thought? I don't know if there is any data to support this one way or another, but for a top pitcher, or any pitcher who's stuff is working that day, the 3rd time through the order is sometimes a "set-up" or sorts. To state the obvious, the hitter has been pounded by a particular pitch or sequence in order to be "changed up", (pun somewhat intentional), the 3rd time through. Would a fresh pitcher mess with the sequencing and perhaps provide a better chance for the batter to get a hit?

 

Now, I understand why the data shows higher offensive numbers the 3rd time through. I'm just wondering if there is someway to correlate this provision?

[Thegrin]

As I read this I started to laugh with delight. It is so purely logical, that no Major League team would ever do it.

 

Then I laughed some

"And if the team is out of it, bring in the take-one-for-the-team guy: Nolasco. (I’d actually consider Nolasco over Meyer for the fourth spot in the relief quartet, but it’s not a real TwinsDaily post without taking a potshot at Nolasco, so I’m using him here".

 

I loved it 

[DuluthFan]

Your chances of ever signing an ace pitcher using this method just became zero.

 

This may work if all the pitchers were young, no name journeymen who just want a shot to pitch, but I highly doubt any established major league pitcher or their agent would readily accept this significant change in their workload.

 

This would need to be thoroughly tested. Any team would then need to fully blow up and start over with the pitching roster, not only at the major league level but in the minors as well.

 

Trying to trade for that pitching upgrade at the deadline? Those prospects you are offering will not look so good to the receiving team who is not using the same system.

 

Any significant pitching injuries would expose your rotation to more AAA pitchers making starts. In a 4-man rotation, that inexperienced pitcher would pitch more often than in a 5-man rotation.

 

A 4-man rotation, you only need 4 starters. Nope. Under this proposal you really need 8 pitchers who can pitch multiple innings like a starter. The 4 starters and the relievers who replace them. I don't think any organization is that deep in starter type pitchers. Sure everyone has a long reliever in the bullpen now, but he is not needed often. He is normally a starter that couldn't cut it and should be at AAA, but there is usually that one start a week that he is needed when a starter gets knocked out early.