Stats Show Twins' Pitching Was the Opposite of Clutch. Why?

[Tom Froemming]

The Twins pitching staff ranked fifth in FanGraphs’ version of Wins Above Replacement last season. How is that possible when the team was 19th in runs allowed per game and 21st in ERA?

Image courtesy of © Jay Biggerstaff-Imagn Images

While it can be painful to look back upon the disappointing 2024 Twins season, there’s no shortage of interesting trends and developments to reflect upon. Something that’s really sticking with me is how much the team underperformed compared to what things like WAR would have expected. The collective value of the players on this team did not reflect their overall record. 

There’s no doubt the team’s lack of consistency was a big factor in its struggles, but there are two stats I came across that stick with me. Part of why I can’t seem to shake them is because of how problematic they are. The part is I’m not sure how you’d attempt to go about addressing them.

Clutch Situations
There’s actually a metric for clutch that measures performance in high-leverage situations. Its calculation is based around Win Probability Added and the Leverage Index. Here’s a link to the FanGraphs Glossary entry on how clutch is measured if you’re interested in more info.

Part of why the Twins win-loss record isn’t reflective of how the team fared in WAR is because they performed poorly in the clutch. This is a familiar story. The Twins have been among the worst clutch teams in baseball the past few years. 

Something important to keep in mind is the clutch metric is not predictive. It only contextualizes the past. Still, it’s worth noting that this has become a recurring issue for the Twins. Here is where the Twins ranked in clutch the past three seasons:

Pitching Staff
2024: 24th
2023: 24th
2022: 24th
3-year period total: 27th

Again, the stat is not predictive, so it would be foolish to do something like go out and target players who’ve excelled in this clutch metric. While this has certainly been a problem, going about fixing it isn't a straight-forward task.

Pitching With Men On
If you prefer looking at something more traditional and tangible, the Twins pitching stats with runners on base tell a similar story. This team crumbled under pressure.

Since pitching with the bases empty is easier than pitching with men on, instead of simply comparing the two situations to each other, let’s look at how the Twins stacked up to the rest of the league. Here is the triple slash line opposing hitters posted in each situation and where the Twins ranked:

Bases Empty
.215 AVG (2nd)
.278 OBP (2nd)
.360 SLG (4th)

Men On
.281 AVG (29th)
.338 OBP (24th)
.457 SLG (28th)

So the Twins were elite when the bases were clean and atrocious once any runners reached. Why?

From a purely pitching perspective, it’s not like the Twins were horrible with runners on. They ranked fifth in K-BB%, eighth in xFIP and 15th in FIP. What killed the Twins is they gave up a .334 batting average on balls in play when there were runners aboard. That was the worst mark in all of baseball. With the bases empty, the Twins yielded a .265 BABIP, which was the second-lowest.

In case you were wondering, league averages in those situations are a .297 BABIP with runners on (37 points lower than the Twins) and a .283 BABIP with the bases empty (18 points higher than the Twins).

How is it even possible the Twins had such a dramatic split? If they were simply that bad at defense, you’d expect the BABIP with the bases empty would have also been poor. Was defensive positioning with men on base the culprit? Is pitch-calling strategy not complimenting those defensive alignments? Just plain bad luck?

I don’t have the answers, so I’m curious what thoughts you may have.

As you could probably guess, the Twins struggled to leave runners on base and had one of the worst strand rates in the league. Their 70.1 LOB% ranked 26th. The poster boy for the team’s struggles was Jhoan Duran. He posted a 61.5% LOB%, which was the ninth-worst mark among the 351 pitchers to log at least 50 innings last season. League average LOB% was 72.1.

What about the lineup? Things were much more consistent there, as the Twins ranked 11th in OPS with the bases empty (.238/.305/.402, a .707 OPS) and 12th in OPS with men on base (.258/.328/.423, a .751 OPS).

Circling back to WAR
So how were the Twins a top-five pitching staff in fWAR? A key thing to remember is the FanGraphs version of WAR is based on FIP: Fielding Independent Pitching. So fWAR is primarily about quantifying the specific value of pitching performance. You would think that would typically align with general run prevention, but the 2024 Twins are an excellent example of how that’s not always the case.

It’s not just that the Twins struggled to convert balls in play into outs, it’s that their struggles to do so came at the most inopportune times. When it comes to diagnosing the Twins 2024 failures, nothing is going to overshadow the team’s September collapse, but struggles in clutch situations and failure to strand runners on base were major contributors to this team failing to reach its potential.

View full article

[Coach Wheels]

Since I am absolutely not an analytics person, my generation still prefers the "eye test" to measure talent and whether or not a player can play. 

The 2024 Twins pitching staff lacked talent. Once Ryan went down, the hole in the rotation was glaring, in my eyes. 😀 The bullpen, up to Jax and Duran, was less than perfect and I would hesitate to ever take Jax out of the bullpen and make him a starter; my eyes tell me he has stuff that can finish games or be the set up guy for many years. My eyes also tell me Duran needs to become a 2 pitch closer and eliminate the curveball from his repertoire. Fastball and hard sinker are wipe out pitches and the velocity he throws both of them with gets LH and RH hitters out.

Bottom line, talent makes pitching data better; go get or build more talent and all the data will fix itself.

[Whitey333]

I agree the Twins just don't have the talent.  They had the worst record for letting inherited runners score.  I too use the eye test.  Too many analytics in all of baseball right now. Jax was awesome in bullpen.  Duran not so much so.  He regressed.  Your closer or high leverage situational pitcher simply cannot lose 9 games in the bullpen.  Yes he needs to get down to two pitches.  Twins touted the bullpen as one of the best in baseball before the start of the season.  That didn't work out too well.

[Linus]

The answer might be that the WAR calculation isn’t accurate or to break out the numbers by rotation / bullpen. This is from memory but I believe some of the bullpen arms were not good allowing inherented runners to score. 

[LambchoP]

The poor defense sure didn't help our overall pitching numbers, but poor pitching was the issue. I'm the rotation we had to rely on three rookies. Festa and especially Mathews should never have been relied upon in the majors before they were ready, but they were due to lack of major league depth. In the pen, we dealt with lots of injuries and our closer regressed a bit and was horrible with guys on base. A lot of those poor outings seemed to happen when Rocco sent him out there in a non save situation. Twins just need better talent, plain and simple. How we do that without signing free agents or taking on payroll in trades, I'm not sure. Twins have done a good job producing young talent and a strong farm system. Problem is, even top prospects don't pan out. You need to supplement the home grown guys with MLB caliber free agents, which we just won't do.

[Doctor Gast]

I love your article, Tom. You're spot on. People can twist stats around to get them to say anything they want them to say. IMO, many stats are biased, and even if they don't like what they say, they can even manipulate them to accommodate their philosophy. If they are Sano-type hitters, they are very unlucky, but if they are Arraez-type hitters, they are very lucky.  IMO what is important is the stat Wins/ Losses. A team that wins close games isn't necessarily lucky. It means that the team has clutch players who can come through when the game is on the line. Many believe that if they hit a bunch of HRs & that those HRs = wins. Not so. Odds are that those who focus on hitting HRs will do so when it doesn't matter & make an out when it does. You might be strong & have the technique but unless a pitcher gives you a pitch to mash you're out of luck. IMO performing in the clutch is an ability not necessarily luck (things like chemistry might enter in). You have players when the game is on the line they come alive. They may be average or even poor during the season but watch out when the lights come on.

You might show the stats that the Twins weren't so bad. But IMO the Twins lacked the chemistry to generate clutch performances especially when it mattered the most. We had chemistry in '23, we messed with that in '24. "Weird" analytics where defense at 1B was much more important than offense & offense was much more important at catching than defense, odds of Margot finally getting a pinch-hit hit was enormous, prevailed. Analytics has its place but it shouldn't replace the heart & soul of the game.

[stringer bell]

I'd put one finger on the inherited runners stat for sure. In that regard the Twins were awful.

The defense got worse as the year went on in large part because of the extended absence of Kepler, Buxton and especially Correa. Two rookies (Martin and Lee) didn't show out as good gloves Castro was put in places where he was stretched. 

Some of this is the variation from year to year. Sometimes poorly hit balls fall in and rockets are caught. It seemed to me that was the case for the Twins in 2024. 

Another thought is that pitching out of the stretch might have diminished effectiveness for some, although many deliveries nowadays are almost the same whether runners are on base or not. 

[mikelink45]

The bottom line is the fact we did not make the playoffs.  No matter where anyone ranks in various statistics the team record at the end of the year is what counts.  Clutch hitting, clutch pitching.  It comes down to having the talent to match the situation and a manager who knows how to use his talent in the right way.  

I have no blame for the rookies - they actually did quite well, but we needed the veterans to bring the season to a successful conclusion and that includes the BP.  How many pitchers did we use - 34 according to Baseball Reference. 

Ten had eras of 3 and under - that included Wallner (I think he could be a pitcher) and Castro.  It also includes the recently released Winder. And no starters.

If we go to a 4 era there are 18 pitchers - of the additional 8 Ober is a SP at the top of the list (18) and Ryan is 14. 

Pushing our ERA to 4.5 there are 22 so we add four including SWR, (20) and Lopez (19). 

Going to a 5 ERA we get Festa, Paddock and Boushley. 

That means we have 10 pitchers with an ERA over 5. 

Looking at this I cannot see the Twins staff ranking higher than our place in the standings where we had the 17th best record out of 30.   

[PDX Twin]

What about clutch hitters? It seems like they were about as bad as the pitchers in recent years: leaving the bases loaded with no outs, not knocking in runners in scoring position far too many times.

Does the manager bear any responsibility for getting players to perform at their best when the game is on the line? It seems like some players and managers are "scrappy," coming up with whatever is needed at any given moment and stealing wins. Spreadsheets are not good at this.

[USAFChief]

fWAR is a ridiculously poor way to measure pitching. 

 

 

[Penthang]

I’m wondering if the inherited runners scoring was skewed by who was pitching when they scored. If my memory serves, often when runners got on the starter would be pulled for a middle reliever who might be fresher, but fresher and not good is still not good. Also, of course Duran’s struggles hurt the Twins as well.

[Tom Froemming]

5 hours ago, Doctor Gast said:

I love your article, Tom.

...

But IMO the Twins lacked the chemistry to generate clutch performances especially when it mattered the most. We had chemistry in '23, we messed with that in '24. 

Thanks for reading, I appreciate it. It does seem like teams with good chemistry tend to pull through in a lot of high-pressure situations. But does the chemistry create a team that perseveres, or does a team that perseveres naturally build good chemistry? Either way, they do seem to go hand-in-hand.

2 hours ago, PDX Twin said:

What about clutch hitters? It seems like they were about as bad as the pitchers in recent years: leaving the bases loaded with no outs, not knocking in runners in scoring position far too many times.

Does the manager bear any responsibility for getting players to perform at their best when the game is on the line? It seems like some players and managers are "scrappy," coming up with whatever is needed at any given moment and stealing wins. Spreadsheets are not good at this.

Last season, Twins hitters were 17th in clutch. A bit below average, but they'd been much worse in recent prior seasons. From 2021-24, Twins hitters rank dead last in clutch.

11 minutes ago, USAFChief said:

fWAR is a ridiculously poor way to measure pitching. 

As I suggested in the article, it's not a great way to look at run prevention as a whole, but I disagree that it's a ridiculously poor way to measure pitching, specifically. I think it does a good job in terms of zeroing in on the pitcher himself, but it's pretty blind to any additional context beyond that.

[USAFChief]

55 minutes ago, Tom Froemming said:

 

As I suggested in the article, it's not a great way to look at run prevention as a whole, but I disagree that it's a ridiculously poor way to measure pitching, specifically. I think it does a good job in terms of zeroing in on the pitcher himself, but it's pretty blind to any additional context beyond that.

fWAR is intended to be a backward looking measure. 

Using "Ignore what actually happened. This is what we THINK should have happened" seems to me like a poor method of measuring what happened. 

And that's fWAR for pitching. "What we think should have happened."

[DJL44]

1 hour ago, USAFChief said:

fWAR is intended to be a backward looking measure. 

Using "Ignore what actually happened. This is what we THINK should have happened" seems to me like a poor method of measuring what happened. 

And that's fWAR for pitching. "What we think should have happened."

It doesn't ignore what happened. It assigns the defense more credit for balls in play and the pitcher more credit for HR, walks and strikeouts.

The Twins were last in the league turning double plays - only 75 compared to the league leading 164 of the Colorado Rockies. That's huge. 50 double plays fewer than league average means the up-the-middle defense was poor. Julien and Martin were the main culprits there but Castro and Farmer weren't good either. They need a full season out of Correa and a new partner at 2B who can play the position well. Does anyone have confidence that Royce Lewis or Brooks Lee will be great at turning two? FWIW - the Rockies 2B Brendan Rodgers is freely available after being non-tendered and expected to sign for a modest contract. He has led the league in double plays turned in two different seasons.

Pitching out of the stretch vs the windup was mentioned above. That could be a coaching issue and should be analyzed in more detail. They should also look at the effectiveness of their defensive shifts with runners on base.

[bean5302]

1 minute ago, DJL44 said:

It doesn't ignore what happened. It assigns the defense more credit for balls in play and the pitcher more credit for HR, walks and strikeouts.

The Twins were last in the league turning double plays - only 75 compared to the league leading 164 of the Colorado Rockies. That's huge. 50 double plays fewer than league average means the up-the-middle defense was poor. Julien and Martin were the main culprits there but Castro and Farmer weren't good either. They need a full season out of Correa and a new partner at 2B who can play the position well. Does anyone have confidence that Royce Lewis or Brooks Lee will be great at turning two? FWIW - the Rockies 2B Brendan Rodgers is freely available after being non-tendered and expected to sign for a modest contract. He has led the league in double plays turned in two different seasons.

Pitching out of the stretch vs the windup was mentioned above. That could be a coaching issue and should be analyzed in more detail. They should also look at the effectiveness of their defensive shifts with runners on base.

Colorado (43%) also had the 8th highest ground ball rate and the worst WHIP (1.52), much higher than the Twins' 27th ranked (39%) and 7th ranked WHIP (1.23). The Rockies put a ton more people on base than the Twins did, and since the Rockies had the 26th ranked number of BB's issued (563) vs. the Twins' 3rd best (433). Tons of those extra base runners were on 1B where lots of grounders meant more double plays could be made.

It's not that Brendan Rodgers is necessarily a defensive stud so to speak, but he gets a ton of opportunities. Apart from one year, his DRS has been negative, his UZR/150 is just a bit above average and his OAA is negative.

FIP is a pretty old stat. It doesn't care about batted ball data, park factors or anything else, really. HR, BB, Strikeouts. Everything else is neutral. The Twins' front office invests heavily in pitching strategy and batted ball tendencies. I use fWAR a lot because it's Fangraphs' data is much easier for me to sort through than Baseball Reference's stuff. Also, I don't understand Baseball Reference's pitcher WAR calculation. I thought it was based on ERA or ERA+, but it's not.

[DocBauer]

I'm simply not versed enough on many of these measurements to say what is good or bad, or what is accurate vs not accurate. But I'm smart enough to recognize that there are some off juxtapositions involved on the OP where being good is certain categories and then lousy in others in hard to quantify.

It may be difficult to break down on a monthly basis, but, what I saw was terrible inconsistency in 2024. Basically, the first 4 weeks and the last 6-7 weeks were bad overall baseball and a losing record.

But there was a 3 1/2 month stretch where the Twins had one of the best records in MLB, and were 17 games above. 500 before their nosedive.

So I have to wonder if it's possible to compartmentalize 2 very different portions of the season as a means of explaining how some numbers were so good, some so bad, and also an explanation for a lack of "clutch". If the Heckle/Jeckle Twins were broke up in to these 2 very different seasons within a season, would we see vastly different rankings? Would we see a spike in clutch measurements?

The inconsistent offense the past 2 years, despite good end of the season rankings, appears to be the reason for a change in the batting coaches. But one can't ignore long stretches of high win totals and not assume the VALUES being discussed weren't much better during that time frame.

No question lingering issues to the best and good players and tossing in a rookie SP that wasn't ready, as well as a rookie SP having a solid year but wearing down would have to allow for some of the poor numbers to close the season, and ultimately "doom" final numbers wouldn't they?

In short, the team started poorly. All numbers should look bad. Then they were really good and won a ton of games. All numbers should look good, yes? And the a 6-7 week collapse would again skew all measurements back down again, one would assume.

There's where the complexity of measurement values create odd conflicts with one another, IMO. 

[stringer bell]

To tag on to Bean’s comment, I’m pretty sure the Twins’ ground ball percentage was extremely low. Add that to a low number of base runners on base, especially first base, the Twins had far fewer double play opportunities than most, if not all, other teams. That doesn’t mean they were good at turning two, but explains at least part of the disparity compared to other teams.

 

[FargoFanMan]

6 hours ago, PDX Twin said:

What about clutch hitters? It seems like they were about as bad as the pitchers in recent years: leaving the bases loaded with no outs, not knocking in runners in scoring position far too many times.

Does the manager bear any responsibility for getting players to perform at their best when the game is on the line? It seems like some players and managers are "scrappy," coming up with whatever is needed at any given moment and stealing wins. Spreadsheets are not good at this.

Yes! The stats and analytics say that over the course of 162 games X will be better. One thing this team struggles with is just putting the ball in play with runners on. Putting pressure on the defense. It happened numerous times to the Twins this year hence the article. Toughness and grit aren’t measured in stats. As much as analytics are part of the game just being able to fist one up the middle to score a guy from second and that guy on second reading it’s over the infield and getting a good jump are as important as all the stats. Being able to get a ground ball DP by properly placing a pitch rather than striking a guy out. It’s more work to go 5-8 pitches deep in an AB to get a K than it is to go 1-5 pitches deep to set up and induce a DP. I like stats over 162 games.  I like results in small sample sizes better most of the time.

[USAFChief]

1 hour ago, bean5302 said:

Colorado (43%) also had the 8th highest ground ball rate and the worst WHIP (1.52), much higher than the Twins' 27th ranked (39%) and 7th ranked WHIP (1.23). The Rockies put a ton more people on base than the Twins did, and since the Rockies had the 26th ranked number of BB's issued (563) vs. the Twins' 3rd best (433). Tons of those extra base runners were on 1B where lots of grounders meant more double plays could be made.

It's not that Brendan Rodgers is necessarily a defensive stud so to speak, but he gets a ton of opportunities. Apart from one year, his DRS has been negative, his UZR/150 is just a bit above average and his OAA is negative.

FIP is a pretty old stat. It doesn't care about batted ball data, park factors or anything else, really. HR, BB, Strikeouts. Everything else is neutral. The Twins' front office invests heavily in pitching strategy and batted ball tendencies. I use fWAR a lot because it's Fangraphs' data is much easier for me to sort through than Baseball Reference's stuff. Also, I don't understand Baseball Reference's pitcher WAR calculation. I thought it was based on ERA or ERA+, but it's not.

bWAR for pitchers uses RA9.

[USAFChief]

2 hours ago, DJL44 said:

It doesn't ignore what happened. It assigns the defense more credit for balls in play and the pitcher more credit for HR, walks and strikeouts.

The Twins were last in the league turning double plays - only 75 compared to the league leading 164 of the Colorado Rockies. 

False. 

fWAR for pitchers ignores defense. Completely. 

 

That's kinda the point of FIP. 

 

Except, of course....for infield fly balls. Which fWAR treats as strikeouts. 

For some odd reason.

 

Of course...when just calculating regular old FIP, they DON'T treat infield fly balls as K's.

 

Weird, to be kind.

 

Like I said, fWAR for pitchers is...a poor measure. 

 

 

https://library.fangraphs.com/war/calculating-war-pitchers/

 

 

[tony&rodney]

It is 2024. You went to a number of Twins games and watched them play in the field. You also watched their opponents as well. You worked hard to construct a pile of numbers to make the Twins look average. Now you hear your brain tell you that what you witnessed was not compatible with the statistics that were created. 

It's the offseason now and the overwhelming consensus is to roll it back in 2025 with the same team as 2024, minus the free agents Kepler, Santana, Farmer, and Thielbar. The fans believe. The stats you constructed will work better next time. You don't need to watch because you can look over the best analytics to explain everything. You accept the fate of the team. All praise to AI.

 

[Tom Froemming]

35 minutes ago, tony&rodney said:

It is 2024. You went to a number of Twins games and watched them play in the field. You also watched their opponents as well. You worked hard to construct a pile of numbers to make the Twins look average. Now you hear your brain tell you that what you witnessed was not compatible with the statistics that were created. 

It's the offseason now and the overwhelming consensus is to roll it back in 2025 with the same team as 2024, minus the free agents Kepler, Santana, Farmer, and Thielbar. The fans believe. The stats you constructed will work better next time. You don't need to watch because you can look over the best analytics to explain everything. You accept the fate of the team. All praise to AI.

 

I have two questions: 1) Who is "you?" and 2) Can I have some of whatever it is you had this evening? :)

[DJL44]

2 hours ago, stringer bell said:

To tag on to Bean’s comment, I’m pretty sure the Twins’ ground ball percentage was extremely low. Add that to a low number of base runners on base, especially first base, the Twins had far fewer double play opportunities than most, if not all, other teams. That doesn’t mean they were good at turning two, but explains at least part of the disparity compared to other teams.

Their ground-ball percentage was low, but it's still 40% of the batted balls. The Twins were in last place with 16 fewer double plays than the 29th place team.

The only team with worse Total Zone fielding stats was the White Sox. They had a bad season on defense.

[DJL44]

2 hours ago, USAFChief said:

fWAR for pitchers ignores defense. Completely. 

Because batted balls are assigned to the defense, not the pitcher

[USAFChief]

50 minutes ago, DJL44 said:

Because batted balls are assigned to the defense, not the pitcher

Right. But they aren't reflected in FIP.

Because fWAR ignores defense. (Except for infield pop-ups, for some reason .)

So, while I agree the Twins were a poor defensive team in 2024, that's irrelevant to anything fWAR related for pitchers. FWAR doesn't care if Colorado turned 1 million double plays and the Twins zero.

And just another reason fWAR is a really poor way to measure pitchers.

 

 

[John Buckeye1]

The new anti-shift rules have made it less of a problem, but I do not love when we shift on righties. It doesn't seem to actually lead to positive results, which might be because we're literally shrinking the amount of area one of our best defenders (Correa) is trying to cover when we do that

[Doctor Gast]

The poster child Duran, was great in '22 & '23 with LOB.  In '24, this could be explained by early injury & normal regression. But it could also be an overall drop in chemistry & some discouragement on Duran's part for not being considered for SP duties. Duran has been used strictly as an unglamorous yet very important fireman role. IMO we should stop looking at players as #s on a spreadsheet. IMO we should elevate Duran solely as a more glamourous closer. It makes sense both statically low LOB, & esteem wise for Duran to regain more of that fire & little skip that he once had especially with the talk that Jax is considered for SP. This down click has added fire to trading Duran during this Twins proclaimed Fire Sale. IMO for our & Duran's sake we don't do this. If handled right Duran can at least regain most of his former glory.

[h2oface]

The pitchers are all being groomed in the style of clipboard stats through the whole organization. These are Falvey pitchers now, whether he drafted them, traded for them, inherited them (some 7 years ago?) or signed them as free agents and let the pitching coaches get their mits on them. He will tell you the solution is in the clubhouse over and over and over. This is what it is and will be as long as this is the system pitchers are forced to pitch in. They can't hold runners, they can't perform in the clutch, they usually aren't allowed to pitch deep even if they are killing it, and they don't seem to improve, in general (a very limited amount of exceptions), but get worse the longer they are in the system. I don't expect it to change.

[DJL44]

On 12/1/2024 at 6:45 AM, Tom Froemming said:

Bases Empty
.215 AVG (2nd)
.278 OBP (2nd)
.360 SLG (4th)

Men On
.281 AVG (29th)
.338 OBP (24th)
.457 SLG (28th)

11 hours ago, USAFChief said:

that's irrelevant to anything fWAR related for pitchers

None of those numbers are fWAR and defense matters for AVG/OBP/SLG

4 hours ago, h2oface said:

They can't hold runners

This is relevant to run scoring (and they are bad at it) but not AVG/OBP/SLG. 

[USAFChief]

15 minutes ago, DJL44 said:

None of those numbers are fWAR and defense matters for AVG/OBP/SLG

AVG/OBP/SLG are irrelevant to FWAR. 

A pitcher who gives up 10 runs on 12 hits in an inning, with no home runs, but ends up with 3 Ks and no walks, is considered to have pitched a perfect inning by FIP.

Regardless if the defense is good, bad, or indifferent, by the way.