Short Starts Are Not a Minnesota Twins Thing

The title of the article is blatantly wrong. I realize what you're implying - that short starts are a league-wide trend - but the plain fact is, short starts are very much a Minnesota Twins thing. 

Again with the flawed interpretation of the third-time-through stats...sigh. Let me try this again. 

Let's use Archer for example. This season he's faced 15 batters three times in a game. Read that closely - that's not "15 starts he's gone through the order three times." In total in 2022, he's had *15 plate appearances* against a batter who's facing him for the third time. This is such a comically small sample size that it's fundamentally insignificant from a data standpoint. Furthermore, each of these 15 plate appearances were against the #1 or #2 hitters in an opponents' lineup - typically where teams' best players hit. 

So again, you're comparing data sets of once through an order and twice through an order - based on facing 9 batters in those samples - against a data set where a pitcher is facing 2 out of the 9 hitters, and 2 of the *best* hitters to boot. Of course the numbers are going to look lopsided when you're comparing two entirely different data sets and assuming a meaningful equivalence between them.

I consider this to be a rudimentary error of data analysis. And yet here we are, watching the Twins completely throw away a chance to win perhaps the worst division in the wild-card era largely because of this demonstrably flawed philosophy. "Total system failure," I'd say. 

The reason Falvey was brought in was pitching.  “It’s hard” isn’t a valid excuse for accomplishing nothing on that front in 3/4 of a decade, whether developed or external.  There are teams with tighter budgets than the Twins out there getting it done.   

There are plenty of free agents out there every year better than Archer every year that require a Cole mega deal.  I’m not buying the implication that they’re just always stuck with guys like Archer.  But, I’d they’re going to cower at the thought of a significant pitching contract, they can’t screw up evaluations like they do.

You can’t trade for multiple guys who are too hurt to pitch in one year.  You can’t pick Archer over Cueto.  You can’t bring in the worst reliever in the MLB over the past 3 years and watch him blow 15 games because you like his spin rate.  Their decision making and roster construction this year has been pretty horrendous at times.

I don’t care when they pull guys.  But, people do because what they’re doing isn’t working.  They’re questioning all philosophical decisions at this point.  Nobody cares if a similar philosophy is working for Tampa or the As (neither will ever win a title).

Nobody gives a rip about the “value” you got from 170 innings of Dylan Bundy at 4.5 ERA while you run they run the division down their leg.

They’re not paid to theorize.  They’re paid to get it right.  They’re getting it wrong, no matter the intentions/theory behind.  It’s bubbling over.  People are sick and tired.

It's not that I "view the concept of analytics as annoying." Hell,  I was on that train when Bill James was still producing his Baseball Abstract every year and most of the folks reading this were wearing diapers.  But I'm reminded of the wise man (can't remember if it was TK or Gardy) who once uttered that analytics are an excellent way of predicting the past.

This article couldn't be further off. Other teams rely on pitches to determine starts, not TTO. The TTO concept being implemented in such a hard set fashion is seemingly unique to the Twins.

The Twins set an all time MLB record for the longest time into the season for a pitcher to hit 100 innings. - That is not a trend or similar to other teams. It's an all time record (which makes it exceptional).

Archer, Gray and Bundy all rank at or near the absolute bottom of MLB starters in terms of their average appearance length.
75% of all starters in MLB average at least 5.1 innings. Twins = 20% (Ryan = 5.1)
50% of all starters in MLB average at least 5.2 innings. Twins = 0%
25% of all starters in MLB average at least 6.0 innings. Twins = 0%

Pitches per game? Same story.
75% of all starters in MLB average at least 85 pitches per start. Twins = 20% (Joe Ryan = 87)
50% of all starters in MLB average at least 90 pitches per start. Twins = 0%
25% of all starters in MLB average at least 94 pitches per start. Twins = 0%

Percentile rank by pitches per start.
Joe Ryan 35% (78/120)
Sonny Gray 6% (113/120)
Dylan Bundy 4% (115/120)
Chris Archer 0% (120/120)

The Twins are not part of a baseball trend. The Twins are unique. Other teams allow starters to keep pitching when they're effective. The Twins do not. The Twins pull pitchers who are pitching well because of the false pretense of a major drop off in performance related to TTO and the organization seems to be refusing to alter it's seemingly failed strategy.

I know what we are doing here. We are treading water, waiting for Randy Dobnak to come back and be the anchor of the staff with his long term contract.

I don't have to say a thing. The previous commentors said what was needed. Great posts all. Bravo.

1 hour ago, dxpavelka said:

It's not that I "view the concept of analytics as annoying." Hell,  I was on that train when Bill James was still producing his Baseball Abstract every year and most of the folks reading this were wearing diapers.  But I'm reminded of the wise man (can't remember if it was TK or Gardy) who once uttered that analytics are an excellent way of predicting the past.

It must have been Tom Kelly if it was a "wise man".

"Four pitchers have made the vast majority of starts for the Twins this season. Among them, Sonny Gray is the only one you would even consider for a top spot or two in a good rotation. Joe Ryan, "

That is the most important and maybe accurate thing written. What year are we in with this FO? and this is what they have to show for a starting rotation?

5 hours ago, h2oface said:

It must have been Tom Kelly if it was a "wise man".

Actually, odds are, Gardy was actually the more "enlightened" of the two when it comes to the "analytics" side of stuff.

Here's a fun fact:  For all the consternation about their length of starts, Al (Ted) Bundy & Archer are second and fourth on the staff in Innings Pitched.

8 hours ago, dxpavelka said:

It's not that I "view the concept of analytics as annoying." Hell,  I was on that train when Bill James was still producing his Baseball Abstract every year and most of the folks reading this were wearing diapers.  But I'm reminded of the wise man (can't remember if it was TK or Gardy) who once uttered that analytics are an excellent way of predicting the past.

TK said it so it must be true.  Apparently, not a single one of those big companies spending millions on data analytics and the systems / infrastructure that drive them don't have a single person that understands this premise.  All of those guys with ivy league MBAs and the analytics experts that work for them understand that data is useless for predicting outcomes or developing strategies. How the rest of the world operates can be telling.

13 minutes ago, dxpavelka said:

Here's a fun fact:  For all the consternation about their length of starts, Al (Ted) Bundy & Archer are second and fourth on the staff in Innings Pitched.

That has little to do with the length of their starts and more to do with the number of starts.  In other words, it has nothing to do with length of starts and much more about durability in this particular season.  

What in the world did MLB do for over 100 years with starting pitchers? Why is it that in 2022 mid September, there have only been 11 complete game shutouts and no one has more than one? Why suddenly can't starters go deep into games, regularly? While the Twins seem to be the poster child for starter failure, the rest of baseball isn't doing very well either, apparently.

The demise of the 'starting pitcher' has made the games a lot less enjoyable to watch, at least for this long time observer. 'Analytics' and all the flood of 'new' stats is clogging the life out of the 'fun'. (IMHO) Is there anything more useless and unnecessary than 'launch angle' of a HR, when watching a game live at the stadium?  It truly is a new world for baseball in the 2000's. For the younger fan, this is all they will know, so its all relative. They soon won't remember the 300 game winners, and the 20+ complete games. They won't remember when an ERA of over 3.00 was NOT considered elite. They will marvel that when Maris and Mantle and Mays were slamming HR's all over the landscape, there was nary one mention of exit velocity or launch angle... only 'wow--did he hit that one a long way'.

The game is different today. Is it better? Or not? Its in the eyes of the beholder. I still like the 'triple crown' stats for hitters as a measure of their skills. I still enjoy W's and L's, ERA's and CG's for pitchers. I am being told those 'stats' are useless today. They have been replaced by lists and lists of letters, large and small, sometimes with little + signs. But soon my generation of fan will be gone, so we will ride it out, into the sunset.

As for the 2022 Twins...they aren't a good ballclub but neither are the other teams in their division. So despite their awful play, they are still in contention in mid-September for a playoff spot.  And it would be fun to see them get in and try to break that ridiculous post season losing streak....wouldn't it?

Great comments all.  I believe the organization has failed miserably in the pitching development and usage.  I'm not going to repeat all the great things the previous posters have posted.  As I see this team in a continued down spiral over the past three months it's a sign that what they are doing isn't working yet they keep doing it.  They should all be shown the door.

8 hours ago, Yawn Gardenhose said:

I consider this to be a rudimentary error of data analysis. And yet here we are, watching the Twins completely throw away a chance to win perhaps the worst division in the wild-card era largely because of this demonstrably flawed philosophy. "Total system failure," I'd say. 

Agree with the flaw in front office thinking. A way to mitigate the damage or reduce exposure with a third time through the order was to use an opener. This would result in facing batters 5-9, weaker hitters. The Twins don't do that. The only guys our starters see three times is the first four in a lineup. The glaring error for me is that there seems to be very little attention to the actual effectiveness of a pitcher on any given day, except for Bundy. This has resulted in short starts and our starters are totally conditioned to go 4-5 innings and 70-85 pitches. I know we don't have a Gerrit Cole or his contract but look at the number of pitches he threw yesterday (more than normal) and you can see that situation plays a role in managing a pitcher. It all comes down to a failure to develop or manage pitching which is greatly exasperated by failing to sign guys other than Happ, Shoemaker, etc. 

The Twins are in a pennant race despite their issues and we should look to attendance as the arbiter of whether change is needed. I'm interested to see how the paying fans react with their dollars in the next few home series. One would think that Target Field should be packed this weekend.

6 hours ago, bean5302 said:

This article couldn't be further off. Other teams rely on pitches to determine starts, not TTO. The TTO concept being implemented in such a hard set fashion is seemingly unique to the Twins.

The Twins set an all time MLB record for the longest time into the season for a pitcher to hit 100 innings. - That is not a trend or similar to other teams. It's an all time record (which makes it exceptional).

Archer, Gray and Bundy all rank at or near the absolute bottom of MLB starters in terms of their average appearance length.
75% of all starters in MLB average at least 5.1 innings. Twins = 20% (Ryan = 5.1)
50% of all starters in MLB average at least 5.2 innings. Twins = 0%
25% of all starters in MLB average at least 6.0 innings. Twins = 0%

Pitches per game? Same story.
75% of all starters in MLB average at least 85 pitches per start. Twins = 20% (Joe Ryan = 87)
50% of all starters in MLB average at least 90 pitches per start. Twins = 0%
25% of all starters in MLB average at least 94 pitches per start. Twins = 0%

Percentile rank by pitches per start.
Joe Ryan 35% (78/120)
Sonny Gray 6% (113/120)
Dylan Bundy 4% (115/120)
Chris Archer 0% (120/120)

The Twins are not part of a baseball trend. The Twins are unique. Other teams allow starters to keep pitching when they're effective. The Twins do not. The Twins pull pitchers who are pitching well because of the false pretense of a major drop off in performance related to TTO and the organization seems to be refusing to alter it's seemingly failed strategy.

EXACTLY.  Baldelli pulls the starter, while we are ahead, say 3-1, 2-1, etc. All because the dreaded third time through the order is coming. Then, the first guy out of the bullpen gives up 1 or more runs and our lead is gone. How is this intelligent baseball? I would like to see the data on the bullpen, from the FIRST time through the batting order.

Bean5302 is spot on.  The Twins are much different than other teams and refuse to get off this third time through the order bugaboo.  I also believe that by routinely limiting pitch counts in the minors in the interest of avoiding injuries (how has that worked out??), you are essentially preparing these pitchers to pitch a maximum of 5 innings in the majors.  It is a self fulfilling prophecy.  If you are used to throwing only 60-70 pitches, you are going to create a psychological barrier and you will tire out after 60-70 pitches.  I actually think by babying pitchers, we are doing them and the team a disservice.

8 hours ago, Yawn Gardenhose said:

The title of the article is blatantly wrong. I realize what you're implying - that short starts are a league-wide trend - but the plain fact is, short starts are very much a Minnesota Twins thing. 

Again with the flawed interpretation of the third-time-through stats...sigh. Let me try this again. 

Let's use Archer for example. This season he's faced 15 batters three times in a game. Read that closely - that's not "15 starts he's gone through the order three times." In total in 2022, he's had *15 plate appearances* against a batter who's facing him for the third time. This is such a comically small sample size that it's fundamentally insignificant from a data standpoint. Furthermore, each of these 15 plate appearances were against the #1 or #2 hitters in an opponents' lineup - typically where teams' best players hit. 

So again, you're comparing data sets of once through an order and twice through an order - based on facing 9 batters in those samples - against a data set where a pitcher is facing 2 out of the 9 hitters, and 2 of the *best* hitters to boot. Of course the numbers are going to look lopsided when you're comparing two entirely different data sets and assuming a meaningful equivalence between them.

I consider this to be a rudimentary error of data analysis. And yet here we are, watching the Twins completely throw away a chance to win perhaps the worst division in the wild-card era largely because of this demonstrably flawed philosophy. "Total system failure," I'd say. 

So as I am reading this, you agree with the data saying that facing hitter's for a third time is bad but the data sample is small.  Perhaps this is a Catch-22 situation in which a manager pulls his pitcher before he gets too deep into the third iteration which, in turn, guarantees a small sample size? 

With the limited sample size, pulling the pitcher seems to be the most advantageous thing to do.

43 minutes ago, Major League Ready said:

TK said it so it must be true.  Apparently, not a single one of those big companies spending millions on data analytics and the systems / infrastructure that drive them have a single person that understands this premise.  All of those guys with ivy league MBAs and the analytics experts that work for them understand that data is useless for predicting outcomes or developing strategies. How the rest of the world operates can be telling.

I would be willing to bet that all the Fortune 500 companies have a large group of statisticians helping Upper Management with business plans.

6 hours ago, bean5302 said:

This article couldn't be further off. Other teams rely on pitches to determine starts, not TTO. The TTO concept being implemented in such a hard set fashion is seemingly unique to the Twins.

The Twins set an all time MLB record for the longest time into the season for a pitcher to hit 100 innings. - That is not a trend or similar to other teams. It's an all time record (which makes it exceptional).

Archer, Gray and Bundy all rank at or near the absolute bottom of MLB starters in terms of their average appearance length.
75% of all starters in MLB average at least 5.1 innings. Twins = 20% (Ryan = 5.1)
50% of all starters in MLB average at least 5.2 innings. Twins = 0%
25% of all starters in MLB average at least 6.0 innings. Twins = 0%

Pitches per game? Same story.
75% of all starters in MLB average at least 85 pitches per start. Twins = 20% (Joe Ryan = 87)
50% of all starters in MLB average at least 90 pitches per start. Twins = 0%
25% of all starters in MLB average at least 94 pitches per start. Twins = 0%

Percentile rank by pitches per start.
Joe Ryan 35% (78/120)
Sonny Gray 6% (113/120)
Dylan Bundy 4% (115/120)
Chris Archer 0% (120/120)

The Twins are not part of a baseball trend. The Twins are unique. Other teams allow starters to keep pitching when they're effective. The Twins do not. The Twins pull pitchers who are pitching well because of the false pretense of a major drop off in performance related to TTO and the organization seems to be refusing to alter it's seemingly failed strategy.

I love this - Ted continues to be the Twins cheerleader finding ways to justify what they are doing, but when you do the same thing that everyone else is doing and it doesn't work I think it is time to change.  I loved Danny Gladden saying to Falvey last night - shouldn't we be teaching how to pitch to the lineup a third time.

What are the statistics for overworked BPs?  How much can you use the BP pitcher and have them be effective?  How many BP arms can you rely on?  When you go 4 you have more pressure on the BP which is usually a few good arms and then a lot of failed starters.  

I posted on another thread - Archer is 2 - 7 and the team is 8 - 16 when he pitches.  Bring on the BP!

47 minutes ago, insagt1 said:

What in the world did MLB do for over 100 years with starting pitchers? Why is it that in 2022 mid September, there have only been 11 complete game shutouts and no one has more than one? Why suddenly can't starters go deep into games, regularly? While the Twins seem to be the poster child for starter failure, the rest of baseball isn't doing very well either, apparently.

The demise of the 'starting pitcher' has made the games a lot less enjoyable to watch, at least for this long time observer. 'Analytics' and all the flood of 'new' stats is clogging the life out of the 'fun'. (IMHO) Is there anything more useless and unnecessary than 'launch angle' of a HR, when watching a game live at the stadium?  It truly is a new world for baseball in the 2000's. For the younger fan, this is all they will know, so its all relative. They soon won't remember the 300 game winners, and the 20+ complete games. They won't remember when an ERA of over 3.00 was NOT considered elite. They will marvel that when Maris and Mantle and Mays were slamming HR's all over the landscape, there was nary one mention of exit velocity or launch angle... only 'wow--did he hit that one a long way'.

The game is different today. Is it better? Or not? Its in the eyes of the beholder. I still like the 'triple crown' stats for hitters as a measure of their skills. I still enjoy W's and L's, ERA's and CG's for pitchers. I am being told those 'stats' are useless today. They have been replaced by lists and lists of letters, large and small, sometimes with little + signs. But soon my generation of fan will be gone, so we will ride it out, into the sunset.

As for the 2022 Twins...they aren't a good ballclub but neither are the other teams in their division. So despite their awful play, they are still in contention in mid-September for a playoff spot.  And it would be fun to see them get in and try to break that ridiculous post season losing streak....wouldn't it?

I really enjoyed your post, but all I could do was put a like!  Nice reflections.

45 minutes ago, insagt1 said:

What in the world did MLB do for over 100 years with starting pitchers? Why is it that in 2022 mid September, there have only been 11 complete game shutouts and no one has more than one? Why suddenly can't starters go deep into games, regularly? While the Twins seem to be the poster child for starter failure, the rest of baseball isn't doing very well either, apparently.

The demise of the 'starting pitcher' has made the games a lot less enjoyable to watch, at least for this long time observer. 'Analytics' and all the flood of 'new' stats is clogging the life out of the 'fun'. (IMHO) Is there anything more useless and unnecessary than 'launch angle' of a HR, when watching a game live at the stadium?  It truly is a new world for baseball in the 2000's. For the younger fan, this is all they will know, so its all relative. They soon won't remember the 300 game winners, and the 20+ complete games. They won't remember when an ERA of over 3.00 was NOT considered elite. They will marvel that when Maris and Mantle and Mays were slamming HR's all over the landscape, there was nary one mention of exit velocity or launch angle... only 'wow--did he hit that one a long way'.

The game is different today. Is it better? Or not? Its in the eyes of the beholder. I still like the 'triple crown' stats for hitters as a measure of their skills. I still enjoy W's and L's, ERA's and CG's for pitchers. I am being told those 'stats' are useless today. They have been replaced by lists and lists of letters, large and small, sometimes with little + signs. But soon my generation of fan will be gone, so we will ride it out, into the sunset.

As for the 2022 Twins...they aren't a good ballclub but neither are the other teams in their division. So despite their awful play, they are still in contention in mid-September for a playoff spot.  And it would be fun to see them get in and try to break that ridiculous post season losing streak....wouldn't it?

I agree 100%.  It seem to me that all the metrics freaks use them to rationalize their opinion of what is important.  And if none of the current metrics prove their point, they create a version that does.  OPS becomes OPS+.  WAR becomes FWAR.  The list goes on endlessly.

1 hour ago, dxpavelka said:

Here's a fun fact:  For all the consternation about their length of starts, Al (Ted) Bundy & Archer are second and fourth on the staff in Innings Pitched.

That is about health and staying off the IL, not ability.

When we were looking at Ryan and WInder in the rotation at the beginning of the year, the thought was "what happens when they hit 125 or so innings, who takes their place." Of course, thought Sands, Balazovic, Enlow...whatever.

Griffin Jax was the innings workhorse, followed by Sonny Gray, coming into 2022.

So, the giving pitchers short starts, which was workable when you were carrying 14 pitchers at seasons beginning and still having the shuttle with St. Paul, you will get 30 starts from the arms and a push towards 150 innings, something to build on in 2023.

With the pennant push, and a lineup that has trouble giving you runs, the Twins continue to rely too heavy on bullpen arms, which half-the-time aren't that good. Even the best bullpen pitcher is going to have a bad run thru a couyple of players in a game.

Few have been brilliant. The Twins are a .500 team right now, but that is because of injuries, and a lackluster offense at the plate, for the most part.

Yes, I want to see rotation arms go into the sixthy (or even 7th), consistently hit the 100 pitch mark. It is a given that even on their best days, a starting pitcher is going to give up runs...be it two runs or four runs in a start. Just like a relief pitcher is going to allow a inherited runner to score, or a pitcher is going to give up a hit or a walk or both in any given inning.

Just because you don't like it doesn't mean its not the best thing for our team when it comes to limiting the amount of times opposing batters face the same pitcher.  Now this does require having at least an average bullpen, which we didn't have for the first half of the year.  I blame Bert going on about pitching 300 innings a year over the airwaves for 20 years.  However - things improve and facts don't lie.  Will this come full circle and someday pitchers will go complete games more often than not in the future?  I highly doubt it as pitch speeds increase and pitchers arms are pushed further and further for higher speeds.  If anything, I think pitchers will go less and less innings in the future.  Its not Rocco, its that hitters simply do better they more they see a pitcher on the same day today, and they always have.

1 hour ago, Squirrel said:

If short starts are how it’s going to be, and that was/is their plan, then you absolutely need a BP to withstand that. The FO is egregious in their philosophy one way or the other. If you are going to defend the FO for limiting their starters by saying everyone’s doing it, then they need to be blasted for the construction of their BP. They built a BP on the cheap. You cannot do one without the other. If you are going to limit your starters you need a BP filled with pitchers who can not only withstand the workload, but who can shut down your opponents and actually get the job done. That’s been my complaint all year. Many posters are calling for Rocco to be fired because of pitcher usage. That isn’t going to change a thing when the FO has put this system into place and then has given the manager the BP they have. Of course it’s a disaster.

Perfectly said!  Before the trade deadline it was simply unfair to expect Rocco to both limit how long starters go, while at the same time turning games over to an incompetent BP.  Not sure why they refuse to invest in the Pen in the off season and year after year it is a weakness that we need to then trade prospects to get back to adequacy.  I am curious how our pen ranked in salary compared to other AL teams trying before the trade deadline - I'm better pretty low.

I'm more concerned that a reliever doesn't go more than an inning after throwing less than 12 pitches and giving up no runs. Look, you warm up throwing 20-25 pitches then you come into the game. If you get through a relatively easy inning, why take the guy out. It makes no sense to me and burns the pen faster than anything. The guy probably won't pitch the next day anyway so let him go a second inning if he has been good. That's why I like Archer as a middle reliever rather than a starter. He can do just that.