The Wholly Unfulfilling Explanation for the Twins
A while back, after one of those occasions where someone accused us of writing about the same teams too much, I checked to see which teams had been written about here most and least often. I had to depend on accurate author tagging, which hasn’t always been consistent, but based on the tags, we’ve written the most about the predictable teams — Boston, New York, Oakland, and so on. The other end was interesting, if also predictable. Not as much about the Astros, although more lately. Not as much about the Padres, although more lately. There’s been comparatively little about the Rockies. And, there’s been comparatively little about the Twins. One thing we can say is that, right now, the Rockies suck. But the Twins? The Twins are demanding devoted and focused content.
Not that we thought things were going to be this way. This year’s Twins were projected to be bad, and that was before Ervin Santana got suspended. Seven games into the season, the Twins found themselves six games out of first place, with a run differential almost twice as bad as the next-worst in the American League. But, since then, it’s been a month. Over that month, the Twins have tied for the league’s best record, with twice as many wins as losses. As things stand, the Twins have the eighth-best record in baseball. They have a better record than the Nationals. They’ve closed the gap on the Tigers and Royals.
There’s competitive baseball in Minnesota. It’s the damnedest thing. But now I have to write the kind of post I don’t like to write. The Twins aren’t good. This isn’t keeping up. I want there to be reasons to believe, but the indicators are what the indicators are. I apologize for the statistics. They’re in charge of me, not the other way around.
The other day, Dave wrote about the unclutch A’s. That shouldn’t be misread as “the A’s are unclutch”; rather, the A’s to this point have had mostly unclutch performances. It’s actually kind of amazing. The A’s are seven wins short of what BaseRuns thinks their record should be. No other team is short by more than three. Billy Beane didn’t assemble a great team, but he assembled a team better than this record, and now it probably doesn’t matter. Oh well.
The Twins are the opposite story. They’re not as extreme as Oakland, but they’ve cleared their BaseRuns record by four wins already. Think of it like this: by actual record, the A’s are seven games worse than the Twins. By BaseRuns record, the Twins are four games worse than the A’s. That wouldn’t mean anything to any of the players. And obviously BaseRuns isn’t perfect. But only a fool would try to argue that BaseRuns is broken.
I know, however, that for many, BaseRuns is kind of obscure. Even if it’s become a familiar concept, it’s not something we can simply calculate on our own at home. So for these purposes, we can stay a lot more simple. Look at this graph. On the y-axis, 2015 team winning percentage. Easy! On the x-axis, team batting OPS – team pitching OPS against. So, simple OPS differential.
There are the Twins, in yellow. Their record is better than the best-fit line would suggest, because while the Twins have more wins than losses, they also have a higher OPS against than OPS for. It stands to reason that would be a bad thing. The Twins aren’t the team furthest from the line. But, they are separated. And seasons — full seasons — tend to balance these things out. Here is last year’s plot:
Pretty solid agreement. OPS differential tends to tell the story, as easy as it is. Of some note:
2015 Twins
- OPS Differential: -0.033
- Win Percentage: .559
2014 Twins
- OPS Differential: -0.045
- Win Percentage: .432
By pitching and hitting and fielding, this year’s Twins haven’t been markedly better than last year’s Twins. Last year’s Twins were an also-ran. So where have these early wins been coming from? As with the A’s, you have to consider clutch performance. And if you don’t like the word “clutch”, you can think of it instead as good timing. The Twins have had pretty good timing. Unusually good timing.
Below is a plot of high-leverage performance compared to low-leverage performance. It combines performance by hitters and performance by pitchers. I found the wOBA differential for each team’s hitters, and then I subtracted the wOBA differential for each team’s pitchers, such that the higher the resulting number, the better a team’s overall timing. Dave looked at something like this for Oakland, but he only considered pitchers.
The Padres have had the best timing. The Mets have almost had the best timing. And the Twins are almost even with the Mets, with the third-highest mark in baseball and the highest mark in the American League. Compare them to the miserable Indians. The Indians have had godawful timing. It’s tried to ruin them. The Twins have just been riding the wave.
But these are the things you just can’t expect to sustain. Not nearly to such extremes. And it’s not like the Twins have, say, an Andrew Miller/Dellin Betances combination that might allow them to slightly over-perform the underlying numbers. Glen Perkins is legitimately good, but let’s not go crazy.
Something funny takes shape if you look at the Twins’ depth chart. They’re projected to win 45% of their remaining games. The lineup is projected for a .309 wOBA. The lineup right now has an actual .308 wOBA. The bullpen is projected for a 3.93 FIP. The bullpen right now has an actual 3.93 FIP. The rotation is projected for a 4.11 FIP. The rotation right now has an actual 4.45 FIP. In this sense, the Twins have actually under-achieved. According to the projections, Twins starters will pitch a little better, but the team will still win fewer games. Getting better while getting worse.
If the Twins carried their current BaseRuns performance through the whole year, they’d finish with 72 expected wins, and a run differential of -87. The Twins’ full-season projection? It’s 73 expected wins, and a run differential of -75. It’s more or less the same. The Twins have been what one would’ve expected the Twins to be. In all ways except for one way, where that one way is the most important way. The Twins have been the predictable Twins, but a little bit of good timing can make a mediocre team look impressive.
I’m sorry to say I don’t see it keeping up. I’m sorry to say I’m not buying the Twins. These are the arguments that nobody likes, but that doesn’t mean they shouldn’t be made. The good news for Twins fans is simply that record matters more than everything else. The good news is wins have been there, more often than losses, and all a fan can ever ask for is enough of a glimmer of hope. The glimmer for Minnesota is still glimmering, while other teams expected to be bad have been out-and-out bad. As long as there’s a chance, it doesn’t really matter what the chances are.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
I can’t argue with one thing you’ve said as a twins fan I can’t see how they keep this up. The only thing I can say is that molitor is maybe better than what anyone has thought or maybe a mngr and a coaching staff can make more of a differnce than anyone thinks. They have gotten their players to take advantage of situations. Thats what their job is, Its to put their players in the best situations to succeed and to coach them in a manner to have the confidence in themselves to do it. If the twins keep winning and make the playoffs Molitor should hands down be the mngr of the yr. He’s turning lemons into lemonade. can he keep it up?
I’d say Molitor is turning the tiny drops of lemon juice from last Tuesday’s guacamole and turning it into six pitchers of quality lemonade.
He is squeezing the juice of half a lemon over the salmon that is the Minnesota Twins as they are cooked on the grill that is the 2015 season, thereby enhancing their flavor and making the dish just a little more palatable.
He is pulverizing the lemons from last year’s harvest, hoping to extract enough juice to compensate for the lack of any entrees on the menu this year. Fortunately, he anticipates receiving some prime cuts of beef in the not-so-distant future…
Having grown up in the state of Iowa, I have been a lifelong Twins (and Cardinals) fan, and the very poignantly interesting facts and observations (in the above article) simply could NOT pass (by me) WITHOUT comment! Back in the 1980’s, especially in the STL-Milwaukee World Series, there was ONE player (more than any other) whose presence on the field raised my blood pressure…his name was Paul DoucheBag. That guy was a Cardinal KILLER! NOW that he is working for some of the GOOD GUYS, I must relent and start calling him by his (supposed) birth name…Molitor. But was I wrong back then? He DID NEARLY (almost single-handedly) steal a WS title from the Cardinals, so I do NOT think an apology is necessarily mandated, but I AM happy to now view him with the respect he (now) deserves (tongue in cheek – of course)! Great job with those Twins, PAUL! AND – YOU were one heck of a great player too!
That was a really weird comment.
So my thought would be this:
If against all odds the Twins remain relevant into June, do they start the service clock on someone like Sano? And if so, does that change their projections for RoS in any meaningful way? Or as a Twins fan, should I hope the bottom falls out over the next couple of weeks in order to let the kids continue to develop rather that be tossed into the grinder?
Sano is least likely to be called up IMO. I think it would be polanco or buxton probably both. Santana’s ceiling is probably utility. Hicks is prob 4th of. Polanco is mashing and buxton has been hot if they both continue into june and the twins are still in the mix I could see them calling up both. But I really think ryan waits till after the break to make this move.
I like many have been following since the documentary pelotero. I really hope for the rest of the AL Central he has some struggling initial stages. Reminds me a lot of a young Hanley without the speed
While I would love to see Berrios get called up this year, I think anything outside a September call-up is asking for a lot from an organization that traditionally holds prospects down for quite a while.
OMG, it’s 35 games of baseball. Can we stop trying to “explain” everything?! There’s no explaining results in just 35 games. Some players hot, some not. There’s your explanation. Jesus.
why are you on FanGraphs
I don’t mean to pile on, but this site is all about why the things we see happen happen in baseball, whether it’s one at bat, 30 games or a whole season. So if that’s not something you’re interested in, then there’s no need to be exasperated.
As one who is new to this site, it’s all a bit overwhelming for me. But the first thing I’m learning is that I don’t know very much, in spite of a lifetime spent reading TSN, Bill James, STATS, and even (shudder) Rotoworld. If you can’t figure that out for yourself, if it all seems pedantic and overblown and overly analytical, there are a helluva lot more sites that won’t challenge what you already know and make you feel smart. Or in far less words, like Jeff said…why are you on fangraphs?
Tramps, welcome. Your coming here is a lot like me about two years ago. You will learn a lot here, and really enjoy it too!
How do new people discover the site? Genuinely curious, I’ve been a reader for about 3 years now, but I can’t remember what first brought me here. All I know is I’ve never looked back.
Thanks for the welcome. Don’t recall the particulars, just stumbled across it in my internet travels. References to other sites are noticed, and since I’m the curious type…..it just happened. And I, to this day, remember the awe (that’s not too strong of a word) upon opening my first Baseball Abstract (was that 1985 or so? I still have those books on my shelf).
“Some players hot, some not” Look at the batting averages and the ERAs for the Twins players and tell me who the “hot” players are. Most are having normal or subpar years. I think the timeliness factor is an actual thing and I hope it continues for another 135 games.
Well put. I think anyone actually watching the Twins can tell you their recent success hasn’t exactly been inspiring.
Rule number one for new Fangraphs visitors. DON’T look at batting averages and ERAs under any circumstances. We’re here to teach you new and better ways to evaluate baseball players. Carry on.
Thanks for writing this post. As a Twins fan, I’ve been thinking the same thing myself over the last couple of weeks. There doesn’t seem to be the underlying performance across the whole team that would suggest they’ve turned a corner.
Also, here’s an idea for another Twins-related post: you could look at some of the players currently on the big league roster who have chances to be part of the next good team and see how their performance to date might affect their projections. Specifically, I’m thinking about Trevor May, Oswaldo Arcia, and Trevor Plouffe. How many innings does May need to pitch in order for his projections to reflect he’s improved his command? Basically, are there any guys they have in place who are performing in a way that makes you say “Maybe they have something here,” instead of just saying, “Well, they’re getting really good timing.”
Great comment.
I would also note that their statistics are going to look at little weird because they absolutely got annihilated by Detroit in that first series. Now those games and performances obvious count too, but the performance looks a little more consistent if you ignore those first games.
“And seasons — full seasons — tend to balance these things out. ”
Absolutely, but they are sometimes balanced by the dot move right instead of down. I think the Twins will fall back to earth, but if their record stays this good it will almost certainly because their performance has improved.
I liked the record versus OPS differential comparison for 2015 and 2014. I’d have liked to have seen the same thing for wOBA difference to see if the range tightens up over a long season. It should but never take anything for granted.
Timing influences perception quite a bit. If the Twins had played to their projected W/L percentage for a couple of months and then had this 19-16 stretch, it may not have even been noticed much or written about.
Then again, there are always outliers, even for a full season. The Twins may not keep up their *clutchiness* but they may get hot as a team and surpass their projected performance. Or not.
It took almost a whole season for the truth to come up, but last year’s Milwaukee Brewers were NOT a division leader.
Well, they were for most of the year. The fact that their success was unsustainable doesn’t change that they were a division leader.
A wholly fulfilling explanation for their success can be found here:
http://twtwsports.blogspot.com/2015/03/twtws-2015-al-central-preview.html
Thanks for sharing this link. Every single sentence is solid gold.
It always amazes me how you number geeks are quick to show how accurate you are but by the end of the year when all your pseudo explanations are blown apart you are MIA.
As a casual observer of the ongoing struggle over the hearts of minds of people who couldn’t care less about who’s right and who’s wrong when it comes to baseball games and the people who play them it is my wholly uneducated and unsolicited opinion that the numbers geeks are winning precisely because when their pseudo explanations are blown apart they are, as you say, MIA.
At the end of the year, those who have played, coached and marinated in the game of baseball will be standing on soapboxes and rooftops mocking all the projections that missed and puffing out their gloriously masculine chests because they managed to acknowledge that Miguel Cabrera is pretty darn good at hitting a baseball. A few with less gloriously masculine chests (and possibly as a result less inflated egos) will acknowledge they were wrong about their finely honed baseball instinct and make another guess at what will happen based upon that same instinct. They will return next year with the same level of instinct and the cycle will begin anew.
Meanwhile, the numbers geeks will have retreated to their Fortress of Solitude, which may or may not be Mom’s Basement, armed with another full season of data to meticulously analyze, dissect, and run through a whole bunch of mathematical operations with their computer/best friend in an effort to understand the reasons they got things wrong and attempt to do a better job next time. They don’t need to be told they were wrong, they’re geniuses and are perfectly capable of reading standings, stat lines and box scores and comparing it to their previously stated predictions. A few will file after actions reports and go over what was wrong (and right) and attempt to explain why or how they were able to be correct or incorrect. They will return next year with all new spreadsheets and pictures with dots, lines, bars and pies you can’t eat and the cycle will begin anew.
Which do you think has a better chance of being correct more often in the future? As with all conflict, its usually a good bet to side with the guys that have the most information. They won’t win them all, but they’ll be better prepared to understand why they didn’t and make adjustments.
i get that opportunistic outcomes likely revert to the mean. However, the rest leaves a bit to be desired. Granted, the relationship between OPS diff and wins correlates strongly. However, there’s little evidence that past OPS diff correlates with future OPS diff. This article argues in effect what your broker explicitly has been told not to tell you, to wit, that past performance IS indicative of future results.
That’s why there are projections. And the Twins have actually played about right to their projections! Which are bad.
Except, your broker says “past performance is no GUARANTEE of future performance”, not that you can’t use past performance to predict future performance. A nuance perhaps, but a super important one.
The best teams do not always win and the worst teams do not always lose.
On one hand I understand the numbers guys are usually pretty accurate, but on the other they kind of ruin the game. I feel like I need a phd half the time trying to read an article. Honestly I can’t say I give a rip about WAR, OBP+, or any of those measures. What I do care about is a team that is playing very well right now against well everyone not based in Detroit, that seems to be gelling as a team and enjoying the game. What will the record be in October I have no clue, but the fun is seeing it shake out. I do not consider myself unintelligent nor do I consider myself a genius. But watching the Twins perform like they are right now is reminding me of the division championship teams we had in the early 2000’s. Well timed hits with people in scoring position. I will say when I watch a game on TV or in the stands, I am not watching graphs or dot matrixes, lets leave that to the people that use them for putting together the team. As a fan I do not see it being needed.
Aaron-I get what you’re saying, but it doesn’t have to be that way. Analysis will never interfere with my love for baseball. Being able to compartmentalize is necessary. So many so-called “knowledgable” baseball people were wringing their hands only last week wondering why Corey Kluber was suddenly a bum. Because of the numbers, I knew he wasn’t. No I didn’t expect one of the highest game scores in history. But I knew he was capable of that and so did virtually every reader on this site. Take what you need. That part is up to you. But NEVER let it get in the way of the joy baseball fills you with.
Fans enjoy the game in different ways. If numbers guys are ruining the game for you, and I doubt you are not alone, I suggest ignoring them. This site is likely not for you.
For others, our love of the game can’t be ruined by anything and we crave as much info as possible. It helps some of us appreciate the game more, not less.
‘and I doubt you are not alone’
should read, ‘and I doubt you are alone’
Fangraphs projects that the Twins will now win 76 games in 2015. Seems reasonable.
However, averaging the first six weeks of data is not the best way to analyze the Twins and their current record – the deviations are just too great at this sample size.
We know the Twins have performed at extreme levels, by both time period and team:
First week: 1-6,-29 run differential
Next 24 games: 17-7,+51
Last 4 games: 1-3,-15
vs. Detroit: 2-7,-33 (23-56 runs scored)
vs. Others: 17-9,+40 (136-96)
What can we say from this data arrangement?
The Twins obviously have a problem with Detroit.
The Twins have dominated the rest of the opposition.
The “rest of the opposition” are not awful teams – Fangraphs projected a winning record for four of the five (OAK, KC, SEA, CLE).
I have watched most of their games this year. By observation, I can summarize the Twins as follows:
The Twins have some promising young players, but no rotation or bullpen depth. The lineup can score runs in bunches (since July 2014, no less). They lack home-run power. Their batters either ambush the opponent’s starter or never figure him out. Like every other team, they feed off middle relievers.
IMHO, I think the Twins will continue to be streaky. The Twins will bring up several key prospects as the season progresses, which will keep performance variability high. So, while 76 wins seems like a reasonable projection, the range of potential wins is great. I wouldn’t be surprised if they win anywhere between 72 and 84 games.
Can I use your tried and true method for the Indians?
We know the Indians have performed at levels, by both time period and team:
First week: 2-4,- 7 run differential
Next 24 games: 9-16,-12
Last 3 games: 1-2,-4
vs. Detroit: 1-5,-3 (35-38 runs scored)
vs. Others: 11-16,-25 (105-125)
What can we say from this data arrangement?
The Indians will have a big win against Detroit, but only once every 6 games.
The Indians have a losing record against the rest of the opposition.
The “rest of the opposition” are not awful teams – They all have positive run differentials other than CHW (MIN,KC,TOR,HOU,STL).
I have watched a handful of highlights this year. By observation, I can summarize the Indians as follows:
The Indians hit line drives as often as many other teams, but have trouble in games where they can’t get guys on base. They like batting against bullpens, and in most of the games they end up winning, they face multiple relievers.
IMHO, I think the Indians will not lose any more games. Their defense is shaky, but high ceiling prospects should come up later in the year and keep performance variability high. I think 80 wins is a bogus prediction and I think it is almost certain that the Indians win total ends up being a different amount.
One thing to always remember about projections and simulations is that they do account for the possibility of the black swan. When fangraphs or other sites project the season thousands of times, almost every underdog wins some of the time. It is interesting mentally to think of what those seasons look like (i.e. do they call up Buxton and he turns into Mike Trout, Jr., does Torii Hunter find the fountain of youth, or do they all play as expected, lose a bunch of routs but win all the close games). The Twins record is better than expected. If we ran this season 100 times, they will regress in most of them… but not all. But we only get to run it once.
Nice to see you acknowledge that you neglect some teams. Good article about the Twins. Now I’m hoping for articles about the Rockies, Diamondbacks, Blue Jays, Reds White Sox and Pirates.
“But only a fool would try to argue that BaseRuns is broken.”
Not a smart way to put it since in a few decades we would have better statistics almost certainly.
Think “2005 White Sox”
I understand the underlying concepts here, clustering and expected runs scored/allowed and the effect of context on a team’s win-loss record. But I’m having trouble understanding how the numbers in the final chart above were arrived at. I don’t mean to be dense, but I read the paragraph several times and still am having trouble. Anyone willing to help explain it, perhaps in formula form? Much appreciated.
Hmmm….here we are almost 10 days later and the TWINS just keep on Winning! OK….so at what point do you say “Maybe they have something in the water up in Minnesota??” Or maybe the Twins really can win games?
Or….maybe they are better than people gave them credit for? THe difference between Twins of 2014 and Twins of 2015: Torii? Adding Torii didn’t make that much difference so maybe these guys were just underperforming all those sucky years with Gardy?
Now here we are 14 days later and the Twins continue to excel. Projections work better in some situations than they do in others, they are less valuable where you have young players and new coaching staffs, which in turn creates a new culture, which of course is difficult to quantify in terms of overall wins contributed. All I know is I’ve now watched this team through a quarter of the season and these things are clear: 1) The roster construction is vastly better than it was last year, including having legitimate CFs on the roster. 2) Molitor’s bullpen roles are slightly more flexible; more four-out saves for Perkins, for example. 3) The baserunning is better. 4) This team has experienced success, so the players compete as though success will eventually happen. Hard to quantify, obviously, but anybody who saw the Twins the last four seasons saw a team that came to the park expecting to lose, and those prophecies tend to be fulfilled.
Maybe its the optimist in me, but I’d love to identify an explanation for our run-scoring that’s more sustainable than “clutch” hitting. A few ideas/hypothesis:
1) At some point I read that the twins were emphasizing “quality at bats”. I never saw a definition of this, but if we were to assume that it had something to do with having longer at bats and/or forcing pitchers to throw better pitches (e.g. not chasing pitches that are far out of the zone) than (a) do we know if Twins are actually having more quality at bats than other teams and (b) is it possible that the team is getting something out of these at bats even if they still lead to outs (and therefore don’t influence OBP or OPS), like a better understanding of the pitcher that can be leveraged in the right situation?
2) One thing that’s been emphasized a couple of times is that we don’t have a regular hitting above 300 — is it possible that a team with a more consistent line-up (ie less deviation amongst the avg/obp/ops of different hitters) has a higher run-scoring potential than a team with the same overall metrics, but less consistency? Or a slightly different version of this question, is it possible that where the hits are coming from in the lineup has an influence in addition to just the total number of hits?
3) Is it possible that the increase in productivity with runners on is driven not by luck, or “clutch”, but rather by an advantage the Twins have figured out how to exploit? The first thing that comes to mind is stealing signs from the catcher? Or exploiting the change in infield positioning that results from holding baserunners on?
Would be very interested in hearing if anyone has any thoughts/analysis on any of this ideas! In the meantime, lets enjoy first place while it lasts. Go Twins!!
I would like to thank the 2015 Twins for reminding us that not everything can be measured or broken down into numbers, much as we might wish it were so.
This article deserves a follow-up. Twins have the second best record at the All-Star Break. Why? Are they still lucky? Do we still expect them to normalize and perform much worse?
So. Late August and the Twins are still in contention. Still going with the “lucky” thesis?
Yes.
Anyone…anyone…Bueller…Bueller…
Welp, the fine folks at Fangraphs seem to employ a strategy of “if our models and ‘advanced’ statistics can’t explain it, we’ll just ignore it and hope it goes away.
Which, y’know, it hasn’t. It behind mid-September and the Twins are only a game out of the playoffs and all.