A Surprising Best?

If you had to guess which team in baseball had played the best overall game so far this season, who would you guess? The Yankees? The Rays? The surprising Padres? How many teams do you guess before the Twins?

The Twins host the Mariners this weekend for a three game set and while previewing that series over at a local Mariner blog, I took notice of just how good the Twins were ranked across the four team categories –hitting, defense, starting pitcher, relief pitching—that I break teams down by. By my rankings, I have the Twins as the second best offense, sixth best defense (rated by UZR) and fourth best in both starting and relief pitching. The WAR rankings here at FanGraphs agree exactly on the position players but differ slightly, to third in starting and 11th in relieving, on the pitchers. Even still, both of our rankings agree, there has been no better team in baseball this season than the Minnesota Twins.

So how come they are only 56-46 and currently out of the playoff picture? Partly they have been unlucky in allowing runs. Despite a very good defense and a home park that, while too early to tell, seems to be leaning toward pitcher friendly, the Twins have an ERA that is 17 points higher than their FIP.

On the hitting side, Minnesota has also been hurt by some unclutch performances at the plate. Their team -1.40 clutch rating is 11th worst in baseball. They have also been unlucky when it comes to turning runs into wins. According to BaseRuns, the Twins should be expected to have about a .586 winning percentage to date worth between three and four extra wins over their current record. That would put them on par with the Rangers and just a few games behind Tampa and New York, which is more in line with where they belong.

The Twins were a preseason favorite for a reason and even with the loss of Joe Nathan they have played up to the level expected of them. All that is left is for the wins and losses to catch up to the individual at bats performances.




Print This Post



Matthew Carruth is a software engineer who has been fascinated with baseball statistics since age five. When not dissecting baseball, he is watching hockey or playing soccer.


26 Responses to “A Surprising Best?”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Bronnt says:

    And the loss of Joe Nathan is a great example of how relief pitchers are overvalued. The job of Joe Nathan, one of the game’s best closers, falls to Jon Rauch, who is slightly less valuable, and the whole group moves up one slot. Finding pitchers to record the last three outs of a ballgame is much less mystical than most teams and baseball announcers make it out to be.

    Vote -1 Vote +1

    • CircleChange11 says:

      Only because 70% of all innings are scoreless.

      Now, in a situation of 1-run lead facing a good opponent’s 2-3-4, and the situation changes. Then, you want someone with outstanding stuff and composure.

      Many save opportunites just don’t require such a pitcher. Ex: Go out and shut down the 6-7-8 hitters, or the 8-9-1, with a 1-3 run lead.

      IMO, elite closers are worth their stuff. But, Ryan Franklin can notch 40 saves. But can he come in and shut down the 2-3-4, with little to no margin for error.

      I wonder what % of saves in this modern era could have still been recorded as saves, even if a solo home run were allowed?

      Just thinking out loud, I know this can be a heavy issue. Don;t want to hijack the thread.

      To the main point of the post, everything you are saying is why the regular season is so dang long. If after 162 games, the “breaks” haven’t balanced out, one has to wonder how many games it would take.

      I remember some sentiment by Whitey Herzog (and I’m sure other managers) commenting on how to hang in there when things ain’t going your way, and it basically came down to “never lose more than 3 straight”. I haven’t followed the Twins closely, but they haven’t needed a 25-5 type run to be right there.

      Vote -1 Vote +1

      • CircleChange11 says:

        Sorry, “half-innings”, not “innings”.

        Vote -1 Vote +1

      • Jason B says:

        CC – “If after 162 games, the “breaks” haven’t balanced out, one has to wonder how many games it would take.”

        “breaks” don’t balance out; if you take a balanced coin and throw 7 heads in a series of ten tosses, you shouldn’t expect to throw 7 tails in the next series of ten so that it comes back to an even 50/50 split. Over long series of tosses, one outcome or the other tends to take an early lead and stay ahead over the vast majority of the series, even though you would expect a 50/50 split over the entire series or any single trial. Same as in baseball – rotten luck early in the season doesn’t portend better luck late in the year to “make up for” the crummy luck in April and May.

        Vote -1 Vote +1

  2. awayish says:

    i’m pretty sure the twins also rank #1 in the category of teams with an astrological name, why didn’t you take that into account when making this utterly brilliant conjecture?

    -5 Vote -1 Vote +1

  3. Bryz says:

    I’m definitely surprised on how well the starting pitching has ranked in terms of WAR this season, but I suppose Francisco Liriano is definitely someone to “blame.”

    Vote -1 Vote +1

    • Erik says:

      Well, when you get a lot of strikeouts, ground balls, and don’t allow home runs – you are going to compile a pretty impressive WAR.

      Vote -1 Vote +1

  4. awayish says:

    seriously. if you are ranking teams by summing up their rankings in individual war categories, WHY NOT SUM UP THE WARS THEMSELVES

    Vote -1 Vote +1

  5. Anthony says:

    This team defies conventional logic in more ways than one. They lead the league in batting average and, unsurprisingly of course, in BABIP as well – yet they’re also on pace to beat the all-time record for most double plays hit into by a team in a single season by a full nine more.

    Not only that, but Joe Mauer, Jason Kubel, Michael Cuddyer, and Denard Span – all line-up mainstays and considered part of the Twins line-up that “knows how to hit” – have wOBAs beneath most projections (Although Span now appears to be right on target and much closer to that). And they still have the third-best wOBA as a team in baseball. If these guys were even playing up to their average expectations offensively, combined with the few playing well above their expectations right now (Particularly Young and Morneau), this becomes the most feared hitting lineup in baseball.

    Vote -1 Vote +1

    • Brad says:

      Leading the league in batting average implies a lot of contact as well as a lot of runners on first base. Hitting into a lot of double plays seems pretty intuitive.

      Vote -1 Vote +1

      • Anthony says:

        This is what I originally thought as well, but noticed that there appears to be virtually little to no correlation at all when you look at every other team utilizing the same stats (Average and double plays).

        Vote -1 Vote +1

      • joser says:

        I suspect you need to factor out XBH. Singles-hitting teams are going to hit into more double-plays than teams with the same AVG who put more into the air into (or over) the outfield.

        Vote -1 Vote +1

  6. CircleChange11 says:

    If batters were playing to their average production, with a few guys “well above” (career years), A LOT of teams might have the most feared lineup in baseball. Certainly it could be said of the top 10 offenses.

    Vote -1 Vote +1

    • Ray says:

      There’s a giant performance gap at the very top. The number one offensive teams from 2002 to 2009 were 120 runs better each year than the #10 teams and 95 runs better than the #5 teams. If you have the fifth best offense in baseball, each lineup slot on average has to over-perform by something close to ten runs apiece to take over the lead. It would take an overwhelming preponderance of things breaking just right to propel oneself from fifth to first, let alone from tenth.

      Vote -1 Vote +1

  7. MikeS says:

    Maybe the 2010 Twins are just an outlier, but if the total of the individual WARs does not correlate with team wins then maybe WAR is not as accurate a stat as is believed. Then again, scientists tend to learn more from the outliers than those in the pack anyway so maybe instead of just saying the Twins are unlucky (which may be true) we should try to determine why the discrepancy exists.

    Stats like WAR are models. Models must be continuously tested, updated and validated if they are to be meaningful. Sabermetrics is a science, that’s the scientific method. Blindly accepting something without testing it makes us no better than the old guard who quote RBI’s and pitching wins as accurate measures of quality.

    Vote -1 Vote +1

  8. Crumpled Stiltskin says:

    WAR has problems. UZR is a good method to compare who is playing the best at individual positions, such as who is the best at 2b and by how much, but not positions across the board (except perhaps OF.) Since it compares players to an average in every zone instead of just counting the total outs they made, and bases and extra based hits they saved or gave away do to errors, the way you would do with hitting. That’s an obvious problem. (Problem with defense is that it involves massive amounts of guesswork.)

    Hence the positional adjustment, which makes some sense except I’m pretty sure it’s not changed on a yearly basis, and small changes in the game are not likely to be picked up, so as if second basement are all suddenly more competent at fielding their position. If every second basemen was suddenly as good as Chase Utley, then they would all appear average due to UZR and get there negligible positional bump when they were actually saving there teams more runs. I’m not saying that is likely to happen but minor changes in you’re position in terms of the general level of competence are, as changes in the levels of competence of hitters and pitchers change on a yearly basis across the board, and for reasons we don’t know. Like, why are the pitchers doing somewhat better this year?

    I think there’s also a big problem in the replacement bump in that the replacement level players very drastically at different positions and the positional bump is a defensive thing, from what was explained to me by one of the guys who honed that system, is due to the average number of plays made/level of competence at each position (could be wrong here, but that’s the way it was explained) and not due to any difference in replacement players. Could be wrong, but it seems somewhat accurate as guys like Juan Castro are replacement players. And not to mention that, but about half the league’s starting catcher’s seem to play at replacement level or below.

    There is also an implicit problem with any system that subtracts value, that is, that gives a negative value to a player for a position you need to play the game. It just doesn’t logically make sense. You can’t play baseball without a 1st basemen. How can a 1st basemen then logically begin at 1 win below replacement over the course of a season?? In the American league you can play without a dh, but it wouldn’t be an intelligent thing to do. How then can a dh begin at a win and a half below replacement over the course of a full season? Logically, the dh should begin at 0, not in the negative and the other positions should be bumped up from there. That’s an obvious problem.

    But it’s still better in that it’s more accurate and more accessible than most any other metric. It deals with the offensive stuff very well. It deals with defense about as well as anyone has figured out how to. It allows you to give rough estimates of the values of player, and it puts it into the form of a win. I don’t think they are quite as exact as people tend to treat them, but they’re pretty damn good as well.

    As due to this article, isn’t there an obvious problem in just assuming a team has just been unlucky in giving up runs, and not that there is actually a discrepancy between your two models and what is actually happening on the field? And perhaps this team is a good test case to hone the model and improve it?

    Vote -1 Vote +1

  9. MC says:

    As a Twins fan who’s watched most of their games, it seems to me they’ve been streaky and they’re all going well or all going bad at the same time. Like rolling up on the Orioles and Royals and hopefully now the Mariners. I began the year thinking the Twins were in the top 5 teams in MLB this year without a doubt. But they haven’t played (W or L) that way but I still think their talent is top 5.

    Vote -1 Vote +1

    • Lucas says:

      Yeah, the Twins have absolutely padded their stats this week against the worst teams in the AL. Not that there’s anything wrong with that, but they’ve probably put 40 or so runs onto their run differential this week, and if you’d done the same analysis last week, they probably wouldn’t have been tops. I think they might be quietly awesome though, we’ll see.

      Vote -1 Vote +1

  10. B N says:

    “If you had to guess which team in baseball had played the best overall game so far this season, who would you guess? The Yankees? The Rays? The surprising Padres? How many teams do you guess before the Twins?”

    If I had to guess which had played the best, I would think I’d have to factor strength of schedule into account actually. Given that the Twins have only a modest advantage in total WAR from what I’m understanding, I think one has to admit to the possibility that some of their WAR results from facing inferior hitting (cough cough Royals) or pitching (cough cough still Royals, minus Greinke and Soria). Looking at this rather informal approach (http://www.beyondtheboxscore.com/2010/7/28/1593708/mlb-strength-of-schedule-estimates), it seems to state that a club like the Twins has had a schedule with a winning percentage about 10 points less than a team like the Red Sox. (And close to 20 pts lower than a team like the Phillies and 30 pts lower than the poor poor Orioles).

    So I can entirely believe that the WAR from the Twins is the best in baseball so far. I can also entirely believe that this is at least partially the result of getting to face worse teams. To me, this doesn’t change the fact that either two or three of the best teams in the AL (and probably in baseball) all reside in the AL East. I say two or three, because the Red Sox position in such a ranking would depend heavily on injury rebounds, which are never assured. But being realistic, does anyone really think the Twins would have the best WAR in baseball if they traded divisions with the Orioles? Or that the Orioles would have much better individual stats if they got to make that trade?

    Vote -1 Vote +1

    • Lucas says:

      That’s fair, but the AL Central does have a slightly over .500 record. It’s not the AL East, but it’s not a powderpuff division either.

      Vote -1 Vote +1

  11. CircleChange11 says:

    I don’t think a coin is a good mechanism because it is a pure 50/50, and both teams entering a single game don’t have the same probabilty of winning, unless they are equal teams.

    I think a better mechanism would be dice. The Twins are a good team, so they have a weighted chane to win most games. So take a dice with 6 sides, ant put a Twins loge on 4 sides and a royals logo on 2 of them and roll it 16 times. We’d expect it to come up “Twins” more often than “Royals”.

    No, 4 royals in a row doesn’t mean that we’ll expect 8 Twins in a row later, but given more talent (more logos on the die) they’d have greater probability for good streaks.

    I’ve never understood the coin as a mechanism, because teams aren’t 50/50 to win the game, nor do all batters have the same probability of a hit.

    Vote -1 Vote +1

  12. Greg says:

    The UZR suprises me because their outfield defense is pretty bad. There was a good article on Twinkie Town about how the outfield defense is probably a large part of the reason the pitching staff is underperforming its FIP and xFIP. Cuddyer, Kubel, and to a lesser degree Young are very adept at turning outs into doubles.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>