The Best Team in Baseball?
If we ran a poll, I’d imagine most people would tab the Yankees or the Rays as the best team in baseball. They have identical records at 78-48, and their .619 winning percentages are #1 in the game. However, neither of them lead MLB in WAR – that distinction belongs to the Minnesota Twins.
At +43.2 WAR, the Twins grade out as the top team to date using measures that don’t take contextual performance into account. As you probably know, WAR is mostly built on the foundation of wOBA, UZR, and FIP as the three main variables, representing offense, defense, and pitching, respectively. The Twins grade out well in each area.
Offensively, they have a team wOBA of .342, trailing only Boston and New York in offensive production. That is even more impressive than it sounds because of how pitcher friendly Target Field appears to be. We can’t make a declarative statement about exactly how the park plays with less than a partial season of data to work with, but all of the subjective evidence lends itself to the idea that it favors pitchers. Despite playing in a lower run environment, the Twins have still been able to put up some serious offensive numbers.
Their position players aren’t just lumbering oafs, either. Well, a few of them are, but they’re overshadowed by some high quality defenders, particularly on the infield. The Twins have posted a UZR of +24.3, 8th best in baseball. Their outfield isn’t much to write home about, but the quality around the infield is staggering. The Twins rank 2nd in UZR at first base, second base, and shortstop, and come in all the way down in 4th place at third base, two-tenths of a run away from a tie for second at that position as well.
Their run prevention isn’t just the defense, though – the pitching is legitimately good, too. They rank 6th in baseball in FIP, and all five teams ahead of them are National League clubs who don’t have to contend with a designated hitter. Their biggest strength is their annual refusal to walk anyone, as they have the lowest BB/9 (2.24) of any team in baseball. They are an evenly balanced staff, as well, as the starters FIP (3.83) and relievers FIP (3.80) are nearly identical.
So, if the Twins have performed well in nearly every aspect of the game, why don’t they have the best record in baseball? Well, the one area they haven’t been great at is the main thing that WAR excludes: situational performance.
As a team, the Twins are hitting .279/.342/.439 with no one on base. When they get someone on, then they’re hitting a nearly identical .280/.355/.428, and with runners in scoring position, it’s a similar .283/.365/.407 line. The problem is that most teams perform better with men on base and runners in scoring position than they do in bases empty situations, so the Twins offensive performance in run scoring situations is worse relative to the rest of the league than it is with no one on base.
The good news for Twins fans? That’s not really the kind of thing that is predictive in nature. That the Twins haven’t hit as well in those situations so far says little to nothing about how they’ll do in those situations going forward. If they begin to get the normal bump that teams see with runners on base, their offense could become even more efficient.
The Yankees and Rays are good teams. The Rangers are, too. But don’t sleep on the Twins – you can make a pretty good case that they’re the best team in baseball this year. I certainly wouldn’t want to play them in October.
Any time a question like this comes up it makes me really wish that they would do away with the imbalanced schedule- it would make evaluating teams so much easier, and be a lot fairer to teams fighting for the wild card as well.
I think they should do away with divisions altogether. Keep the AL and NL, put 15 teams in each, stop obsessing about having to do all interleague series at the same time, and send the top four teams in each league to the playoffs.
I’d go even a step further, and let 5 teams make the playoffs. The bottom two play a short series first. And then the team with the best record gets to play the winner of that series.
What’s the fun in making everything easier to evaluate? And besides, the unbalanced schedule should work against the Yankees and Rays and in favor of the Twins, so it’s not really relevant in this argument.
Every once in awhile, what’s good for the Yankees is also good for baseball. It’s rare, but it’s possible.
A playoff format in which the same two teams would go literally every single year is good for baseball?
… ok.
So instead, let’s come up with one that doesn’t send the best teams to the playoffs! Yeah, that makes sense.
The Yankees went to fifteen of eighteen WS between 1947 and 1964. Worked out for baseball just fine.
The biggest difference between this Twins team and those of the past is a pronounced lack of suck. Brendan Harris had a really rough first half and was demoted and Nick Blackburn did the same on the pitching side (although he is now back up starting with Slowey recently on the DL). Aside from these two temporary problems, there just haven’t been the anchors that have kept the team’s top talent down this year. That’s the beauty of a 95 million dollar payroll versus 65.
A healthy Morneau (WAR leader prior to July 7 injury) would certainly help make the lineup deeper for October as well.
Going strictly by those metrics neglects the weak schedule the Twins benefit from by playing in the AL Central, relative to Rays/Yankees.
Dude he mentions that in his post, albeit briefly
Where?
no, he doesn’t.
No mention of the Padres? OPS+ over 100 (even before additions of Tejada/Ludwick), SB% (in high volume) around 74%, ridiculous FIP in the rotation and bullpen, very good fielding metrics. Division has four teams above .500, too.
You seem to indicate that Tejada would improve the Padres’ OPS+ when his sits at 88 for the year.
His OPS+ with SDP is 115, not counting tonight’s HR.
The Twins are a collective 9-15 against the non-Baltimore teams of the AL East. When you play half your games against the likes of KC, Detroit, Cleveland, and Chicago then a good team is going to put up great numbers.
We already knew the Twins were a top tier team, but they’re not better than the Yankees or Rays.
Dave:
Big time Kudos to you to acknowledge WAR is a measure “don’t take contextual performance into account.”
Your site has the best user interface, a wealth of data, & lots of good commentary (although biased somewhat to over reliance on imperfect stats & questionable data source collection: defense).
It takes a big man to admit this even casually within the context of your article coming on the heels of the recent debate involving the Mariner’s extreme focus on defense over offense this yr in constructing their roster.
While your traditional readers might revolt short term, I believe continuing in this direction, especially leveraging more of the WPA/Win Exp data provided on FG, to elicit value added commentary would be a welcome addition.
Who knows, you might even get a scout or two to join a healthy debate :->
Thanks
WAR’s lack of contextual information is a feature, not a bug. wOBA, FIP, and UZR were all designed to intentionally ignore situation performances, which are mostly just noise and not actual talent or ability.
By the way, scouts are always welcome here, even if you don’t agree with what we write. There’s room for both perspectives at the table.
Why do scouts “not agree” with what you write on here? Don’t scouts and Sabremetricians agree on the overall effectiveness/quality of MOST players in baseball? I mean are there scouts out there running around saying that Chase Utley or Albert Pujols suck? Or is it just the tools that are used to come to that conclusion are completely different?
Well, not all scouts have the same opinions. They’re not a group that walks in lockstep, repeating opinions handed down from above. So, I wouldn’t say that scouts don’t agree with what we write here. I have a good relationship with quite a few scouts who really enjoy the site and agree with a lot of what is written.
However, in general, scouts are more likely to be give credence to things like clutch hitting, team chemistry, and other “hard to measure” things that we mostly dismiss.
Its not so much that scouts tend to disagree as they focus on different attributes. A scout would look more at physical attributes like speed or athleticism and what a player might do, as opposed to sabermetrics who focus on hard data and what a player has done. No scout would say that Utley or Pujols suck. However a scout might say that a player like Gary Brown will be a great CF because he’s fast and makes good contact. However sabermetricians would probably focus on things like his super low walk rate and lack of power. This is just a hypothetical and I think you’re right 90% of the time about sabermetrics and scouting agreeing.
“WAR’s lack of contextual information is a feature, not a bug. wOBA, FIP, and UZR were all designed to intentionally ignore situation performances, which are mostly just noise and not actual talent or ability.”
What exactly do you mean by the last part? Do you mean that (for example) a home run is just a home run, whether it comes with the bases loaded or bases empty? Or any hit or out for that matter?
Aaah, okay thanks Dave. And I think you would also agree that scouts are extremely useful when it comes to amateur players and players in the low level minor leagues? An example that comes to mind is Hanley Ramirez – If you just look at his advanced stats in the minors there was not alot there that pointed to him being an elite offensive player in the Major Leagues. But scouts were all over him and thats why the Marlins made him the centerpiece in the Beckett/Lowell trade.
Yep. I’m a big fan of scouts.
What exactly do you mean by the last part? Do you mean that (for example) a home run is just a home run, whether it comes with the bases loaded or bases empty? Or any hit or out for that matter?
When it comes to determining the actual abilities of a player or team, yes. There’s no evidence that major league players significantly differ in their abilities to perform well in different situations. Giving extra credit to a team that has hit well in the clutch is not a great idea if you’re trying to compare teams actual abilities against each other. Past clutch hitting is just not predictive of future clutch hitting.
That’s what I figured but I wasn’t quite sure.
I know this is more of a “food for thought” article, but I really hope nobody takes this to mean that sabermetrics universally show that the Twins are better than either of those two teams. Jon Heyman comes to mind.
Thats a good point. I could just see the Twins getting swept in 3 games by the Rays in the 1st round of the playoffs and then Jon Heyman writing a column titled “Sabremetrics Doesn’t Account for Heart, Determination” or some shizz like that.
Is there any correlation between team WAR and postseason success?
I think they’re going to have problems in the playoffs when other teams will get the chance to rest line up their rotations…. Liriano vs. the other number 1′s is a wash in my opinion, and they don’t have much 2-3/4. They are a regular season team (2-12 in post season since ’03).
This team would be ten times better if they traded Liriano for Jose Lopez.
Is it just me or are the Mariners jokes not funny? Give it a rest already.
Just you
I think it’s just you actually, the rest of us with a sense of humor find them funny.
Seinfeld is funny. A cat falling into an aquarium is funny. The repetitive trollish mariners jokes aren’t funny to me. I have a sense of humor, but I guess these just aren’t my cup of tea.
A cat falling into an aquarium is ironic. Because that is a result that is opposite of that which was intended, almost as if fate is mocking the cat.
It’s also funny (having seen a cat fall into a bathtub…close enough).
that’s not irony. that’s like saying it would be ironic if a person fell on their grill and cooked themselves.
WAR be damned, Sox will catch the Twins ( I can still dream can’t I?)
Is this just an attempt at drawing hits by getting people to angrily respond?
I enjoy and respect the work that FG does and am a regular reader, but this article does more to show the flaws of WAR as a stat than it does anything else.
“But don’t sleep on the Twins – you can make a pretty good case that they’re the best team in baseball this year. I certainly wouldn’t want to play them in October.”
The Twins are *exactly* the team all the other AL playoff teams hope to draw in October. By ignoring the strength of schedule for all the AL playoff teams, the article undermines it’s own conclusions.
I agree that the Twins are the team that the other 3 would want to play. No one wants to draw NYY or TBR in the division series, and your other choice starts Cliff Lee.
That does not, at all, mean that the Twins are terrible. They’re likely a top 6 team in the AL. Whether they’re 4th, 5th, or 6th requires more research than I am interested in doing.
I don’t know that just adding up WAR is necessarily the best way of doing it … especially given how much more production an extra WAR over 4 WAR gives a team if that batter hits in the middle of the lineup …. versus “making ground” in WAR with bottom of the order players.
Is this a desired application of WAR or just an interesting mental exercise?
The twins look very good, especially Sabermetrically, but I still don’t favor them over the Rays or Yankees in a postseason series.
“If they begin to get the normal bump that teams see with runners on base, their offense could become even more efficient.”
Of course, they could revert to a more normal distribution by lowering their results in no one on situations, which would put them more in line with the rest of the league. They would see the expected bump in the other situations, and it might even make them more efficient, but it would probably result in fewer runs.
For all those complaining about the strength of schedule issue, there are four bad teams in the American League – Baltimore, Seattle, Kansas City, and Cleveland. Total games played against those four teams:
Minnesota: 39
New York: 38
Texas: 35
Boston: 33
There’s a reason I didn’t bother adjusting for it – it’s not a big deal. The differences in schedule strength are not nearly as large as you might think.
Isn’t the bigger difference in playing the “good” teams? For Tampa, playing New York and Boston as opposed to Minnesota playing Chicago and Detroit?
I’d even say that the “okay” teams (a.k.a. the middle of the pack team(s) in each division) are much better in the East: Toronto vs. Cleveland.
Mr. Cameron, are you saying that the Twins would be just as successful if they were members of the AL East as opposed to the Central?
No – I’m saying the difference just wouldn’t be that large.
Here’s strength of schedule for every team in MLB, per ESPN. As you can see, the spreads just aren’t that high.
Thanks for your response, although I’m not totally satisfied. I’m not sure how ESPN’s SOS rankings translate into a different win loss for a team in the AL East vs. the Central, but I would assume that if the Blue Jays and Twins traded divisions, we’d be talking about how great the Jays are since they’d be winning the Central while the Twins would be toiling away in 4th place in the East.
I don’t know if there is a way to factor in SOS into the calculations, but as an American League fan, I have a hard time believing that the Twins are anything but the worst AL playoff team (and probably worse than a couple of AL East teams that miss the playoffs . . .). That doesn’t mean they’re not a good team, they are, they’re just probably the 5th best team in the AL.
Anyway, I’d be curious for your expanded thoughts on this Mr. Cameron. Maybe a topic for a post?
Well how many good teams are there in the AL, and how many games have the Rays/Yankees/Red Sox had to play against them? This is a very simplistic response in my opinion. I don’t think that strength of schedule can be nullified by comparing games played against 4 teams…
I think you’re neglecting that the Yankees/Red Sox/Rays each have to play each other 36 times (each). The Twins combine to play them 19 times. There is a difference there. They may all play the crappy teams the same amount of time, but they aren’t playing the top teams equally.
But whatever, the Twins are good and I don’t think anyone can deny that fact, regardless of schedule strength.
According the Massey Ratings, whose college football ratings are (or at least were) a BCS component, the Twins have had the 13rd hardest schedule thus far. Yankees:12th, Rays:2nd, RSox:4th.
FWIW, his rankings have the Twins 6th in MLB behind TB, NYY, BOS, SD, CIN.
What about interleague, or less-than-crappy teams?
I would be interested to see the weighted average WAR of every team’s compitition.
I remember I did this for just the interleague schedule between Texas and Oakland and it was pretty staggering. I can’t remember exactly, but the A’s were playing on average teams that would be ~85 win teams, while the Rangers where playing ~70 win teams. For 16 games, that’s got to add up to a win or two all by itself.
I think Dave is right, it’s not that big a deal. Here’s the average wOBA of teams faced in the season so far:
Twins: .324 (9th hardest in MLB)
Yankees: .322 (17th)
Red Sox: .325 (8th)
Rays: .325 (7th)
And average opponent xFIP:
Twins: 4.38 (28th)
Yankees: 4.38 (27th)
Red Sox: 4.36 (23rd)
Rays: 4.35 (22nd)
There’s not really that much to get excited about there, is there? The Twins played a pretty tough interleague schedule and at this point have already completed season series against the Yankees, Red Sox, and Rays, and have just two more games to go against the Rangers.
The Twins actually have tonight’s game, tomorrow, and then a three game series at home in about a week versus the Rangers.
So five total remaining, including tonight.
Well, the suckiest teams probably have the highest rankings in “opponent strength”, since the sucky teams contribute to the better team’s success and the sucky teams don’t get to play against themselves?
When PIT plays an inter-squad game, instead of getting a win and a loss, they get 2 losses. *grin*
Why not look at something that measures actual runs scored and runs given up, like oh, run differential? Scale it to league/division if need be, and then compare. Why look at wOBA and xFIP, when we have a record of the actual runs scored and allowed?
Boston and Toronto (3rd and 4th best in the east) are better than Chicago and Detroit (2nd and 3rd best in Central). In fact compare the divisions top to bottom. Even if you argue that the Twins are better than whoever you put on top of the East the 2nd place team in the east is better than the White Sox, the Red Sox are better than the Tigers, the Jays are better than the Tribe and the bottom feeders are both bad. The AL east is so much stronger than the central it’s silly. Strength of schedule won’t reflect that accurately because of the unbalanced schedule. The twins play 54 games agains chi/det/cle. That’s easier than TbNY/Bos/Tor any way you slice it.
And anybody can get hot but after Liriano I’ll take TB, NY or Tex pitching staff top to bottom.
“I certainly wouldn’t want to play them in October. ”
So you’d rather play the Yanks, Rays, or Rangers?
Seriously?
Dave is saying the Twins are to be reckoned with in the postseason and the naysayers say exactly the opposite. What I know is this: the Twins have a great team defense and offensewise and just like the Rays, Yankees, Rangers and Red Sox for that matter: I wouldn’t want to face them. Especially in the ALDS. Three-game series is just to short to rely on stats. Mauer could go 0 for 16 in an ALDS for all we know.
“I wouldn’t want to face them.”
The implication here (and from Dave) is you’d rather face someone else.
So who? The other likely playoff teams are TB, NY, TX. You’d rather player one of those?
Stupid me. I said three game series. That should obviously have been five game series.
I wouldn’t want to play anyone in a three game series; it’s a three game series, and there’s way too much variance in a three game series for it to be anything remotely meaningful.
You wouldn’t want to play anyone? Uh…
If you “wouldn’t want to play anyone in a three game series” (of course the playoffs are 5 or 7 games . . .), why bother making the point that you don’t want to face the Twins?
I agree with commenter Bubba that the implication is that you’d rather face another playoff team. And because that’s the implication, I also disagree that the Twins are team you’d want to avoid.
My fovorite team is the yankees. But i think the tampa bay rays are the best team in the sport. They are tied with the defending champion. They have grrat starting pitching. I say the best in the league. That staff is lead by Price, Garza, Big Game James, and that Hellikson kid isnt bad. Plus they have a great bullpen. Benoit and Soriano are lethal in the 8th and 9th inning respectivly. Although their offense isnt great they have CC, Longo, Pena, and others. I think righ now the rays are better than the twins. I dont buy this baloney cause im sure this time last year u said the red sox or some other team besides the yankees and phillies.
So the twins play 72 games against the Indians, Royals, Tigers and White Sox. And they play 37 games against the AL East.
The Rays/Yankees play 72 against each other and the Jays, Red Sox and Orioles, And they play 37 games against the AL Central.
The east teams are a combined 87-68 against the central division. That’s a .561 winning percentage.
The central teams have a .439 winning percentage.
Stretch this over a 162 game season and the east teams are playing like 91 win teams against the central, while the central teams are playing like 71 win teams against the east.
So yes this does make a difference.
The Twins are 33-18 against their own division (.647, or 105 win baseball), they are BELOW .500 against the east!!!
.647 on 35 games is 23 wins. .438 (how the have fared against the east) on 35 games is 15 wins. So take the Twins final record and subtract 8 wins and you have an approximate of how they would do in the east.
So yes the Twins are the blue jays right now if they were in the east and they would not be in the playoff picture. They would be 4th in a tough division so I think all three other american league playoff teams would like to play them!
If you think certain players are clutch, how about you go and prove it to us all instead of firing this nonsense about what you think ALL GMs (or scouts or whatever) believe….
ah post fail, this is meant for Talent Scout
“There’s no evidence that major league players significantly differ in their abilities to perform well in different situations… There’s no evidence that major league players significantly differ in their abilities to perform well in different situations…Past clutch hitting is just not predictive of future clutch hitting.”
–
Umm, Dave & Gentlemen: This is not to be construed, as ‘Ben’ suggested a “troll” (whatever that behavior is) response, but your reply excerpts above are just the type of over generalized SABR talk which causes scouts, field mgrs., & front office jaws to drop in disbelief & hinders wider acceptance of stats based screening for both prospects & advance team visit scouting.
Please understand, the attempt is NOT to be inflammatory at all, but merely to clarify and acknowledge the value of stats BUT WITHIN THE PROPER CONTEXT OF GAME PLAY.
To casually dismiss the data context as merely “noise” diminishes the talent & data assembled on this site.
Why have WPA/WinExp data on FG if you don’t believe in it’s use as a more approximately relevant tool for to confirm the in-game play context that we all observe.
There’s no way any GM is going to agree with you that clutch absolutely doesn’t hold value in prospect evaluation, trade consideration, & contract considerations.
Otherwise, let’s get Albert Pujols to agree to a substantially lower contract extension than Matt Holliday. (there’s an over 100 pt diff in their RISP rates)
I’m not sure I understand your last point – if you are saying that Albert Pujols should get the larger extension because he’s more clutch than Holliday, I don’t think I’d agree with that. Pujols will get the larger contract because he’s pure awesome, regardless of context. He’s consistently among the top couple players in WAR, year in and year out. He’s pretty much been the best player of the decade, and if he’s slowing down at all, it’s only relative to his Ruthian peak.
Kevin S.
Clarification on my part seems to be in order for you better frame your response.
The issue with clutch, for an example, in the e.g. of Pujols vs Holliday is of establishing or validating what’s the relative value & length of a contract extension.
So, while I would agree with you that Pujols is in fact “pure awesome”, from a practical limit of constrained funds, team utility, player aging, etc., there are other factors which most definitely are part of the consideration set to determine what’s optimal and where the potential inflection point in marginal utility falls w/in a contract extension payback curve.
Shareholders, owners, & even field managers however demand a bit more precision than selective use of certain data such as WAR which IMO simply by design does not reflect the closest approximation of actual game play’s impact on actual (not projected or estimated) odds of winning which is used on the field.
At any rate, not to split hairs over arcane or controversial topics here on the site.
I will take Dave Cameron’s suggested offer into further consideration for offline review.
Thanks all for a very, very illuminating peak inside your data world.
Why have WPA/WinExp data on FG if you don’t believe in it’s use as a more approximately relevant tool for to confirm the in-game play context that we all observe.
Different tools for different things. Context neutral stats do a better job of telling us how good a player actually is, while WPA does a better job of telling us what actually happened in a game. WAR is better for analysis of who you should acquire, while WPA is better for who deserves a pie in the face at the end of the night.
They have different uses. We use the context neutral stats more often because the questions that it answers are generally more interesting. If I wanted to know who had the most “big hits” in 2010, I would use WPA. I f I want to know who the best hitter in 2010 was, I would use wOBA. Different question, different tool.
There’s no way any GM is going to agree with you that clutch absolutely doesn’t hold value in prospect evaluation, trade consideration, & contract considerations.
I’d bet you I could get at least five GMs to agree with me, and a lot of AGMs. You’re correct that the majority of the high level executives still put a lot of value in clutch performance, but I think that’s starting to change.
If you’d like, drop me an email at dave.cameron@fangraphs.com, and I’ll be happy to show you the research that’s been done on the subject, and we can talk about any objections you might have to the data or their methodology. If you want to create a fake email address to stay anonymous, that’s fine too. I’m happy to have a discussion about it.
Wait, what did I suggest?
Ok, so I sort of answered my own question from above about weighted average opponent WAR.
Basically here’s what I came up with
BAL 31.19
BOS 27.27
CHW 26.80
CLE 31.74
DET 27.74
KC 29.58
LAA 29.60
MIN 26.12
NYY 26.95
OAK 27.12
SEA 28.80
TB 28.29
TEX 27.33
TOR 28.46
Now I did this by first completely ignoring the interleague. I did this simply because I didn’t want to spend an extra hour or what ever tracking down the number of games with which NL opponents. It also made some of the later calculations a little easier. Unfortunately the interleague schedule is going to introduce much of the variability in Opp WAR since schedules will be so drastically different.
Anyway, so above is only the AL Opp WAR. The average there is 28.36 and the stdev is 1.67. Not surprisingly the bad teams in the AL central and AL east have the hardest time. Also, you’ll notice the twins have the easiest schedule by this method. Next come the Yankees (somewhat surprisingly) and the White Sox.
So I then wanted to adjust each team’s WAR based on their average opponents WAR. So I just multiplied each team’s WAR by the ratio of their opponents WAR to the average AL team’s opponent WAR. So that looked like:
BAL 12.20
BOS 38.47
CHW 32.69
CLE 14.10
DET 30.42
KC 17.94
LAA 21.29
MIN 39.78
NYY 34.88
OAK 25.15
SEA 17.26
TB 37.60
TEX 33.44
TOR 33.62
So the twins get knocked down the most at a factor of ~.92, but they still lead the way. Red Sox comes in 2nd, Rays 3rd, Yankees 4th, Jays 5th, Rangers 6th, White Sox 7th.
Now I’m not gonna say this is the best way to do this, obviously I’d at least need to include the NL, but it should give us some idea of how much the competition really changes. And if its generally +/- 1-2 WAR on your average opponent, then some teams are going to be on average facing 83 win teams while others will face on average 79 win teams, at least when only talking about intRA-league schedules, which is only going to lead to an extra win or maybe two.
All that said, interleague schedules will vary a great deal more, and from what I remember about doing this previously, just interleague could give a team a one win advantage as well. So, I think its reasonable to assume schedule differences can lead to 2 or 3 win differences over the course of a season.
Dave, as much as I want to agree with your assertions, especially since I’m a die-hard Twins fan, I just can’t until they prove they can actually beat the Yanks on the field. I love stats and sabermetrics and rely on them for my fantasy leagues but the World Series is not a fantasy game and all stats aside, the best team is decided on the field. 100 years from now, no one will care if the Twins are the “statistically” best team.
While this is flattering (as a Twins fan), watching the Twins get out-played in every facet of the game versus the Rangers thus far through their 2+ games doesn’t offer much prove of Dave’s opinion that the Twins are the best team in baseball.
Making an evaluation based on 2 games is silly. Plus they swept the Rangers earlier this year, FWIW.
The distinction would be that both teams are playing well now, and Texas has Lee.
But, I do agree about small sample sizes, but even a full season of games played (less than 20) in different months wouldn’t be enough data.
Texas having Lee now is barely worth mentioning considering Lee wasn’t the one pitching the last three games that the Twins lost.
The better distinction to make would be the prominent absence of the Twins’ best player this year in this series and his appearance in the previous series. Watching every game in their entirety has been painful recently as Cuddyer manages to miss even the most moderately difficult picks at first base that Morneau would very likely not have missed.
Hopefully the best team in baseball wins a playoff game this year. That was disappointing. I was hopinh for a full series of the Twins and Yankees, but instead Nathan killed that with an 0-2 cockshot to ARod.
Seriously, the best team in baseball could reasonably get swept by either the Yankees or Rays.
What the Twins should be labelled as, perhaps, is the “most balanced” team in baseball … not necessarily the best team.
I could see them just as easily sweeping the Rays or Yanks. There’s really not very much separation between the four playoff teams in the AL, IMO. Yanks and Rays are a coin flip, IMO, and I don’t think either would be more than 53/47 or something along those lines in a short series against Texas or Minny. AL playoffs are gonna be fun this year.
I thought they were going to be fun last year in the “Year of Super Joe”.
The playoffs are more about “top tier talent” than they are “balance” and depth, or so it seems.
I would be stunned if the Twinkies swept any of the other 3 teams, but I wouldn;t be stunned (disappointed, sure) if the Twins were swept by any of the 3.
I have a very strong appreciation and respect for how the Twins run their ballclub, and develop players. But, they’re not equal to the other 3, even though baseball teams are much closer to each other than say, football and basketball teams.
I just want to see some competitive series.
But, sometimes it seems as if we, at FG, go out of our way to make the Twins look better than they are. I think the comments about # of games against the ALC, and the win % against the ALE are worthwhile.
If someone were making the point that the Twins had no shot in the playoffs, I would be defending the Twins. But, the “don’t be surprised if they sweep ….” is a little much. Possible? Sure. But it would be surprising.
Dave, I have a problem with using FIP in this context (and most contexts), maybe you could clear it up for me. I understand the use of FIP if you want to measure a pitcher’s “true ability,” but why remove the variable of fielding? For instance, a groundball pitcher in front of a great infield defense (like that of the Twins) will have a great ERA, but who knows what his FIP will be! His FIP will unnecessarily remove the pitcher’s greatest talent, which is pitching to his team’s strength.
I think FIP is great if a pitcher has just changed teams. With a new defense behind him, ERA will be a poor predictor of his future success, but FIP will give us a better idea of his true talent level. But when projecting future performance of a pitcher who will be playing in front of the same defense, why remove fielding from the equation?
For instance, Brian Duensing’s ERA (1.92) is way lower than his FIP (3.48), but I don’t think this means he’s just getting lucky. He induces a ton of groundballs pitching in front of a great defense – using FIP doesn’t give him the credit he deserves.
Because you don’t want to double-count the value the defense provides in the pitcher’s value as well. Since UZR counts towards fielding WAR, we’d want to factor it out of pitching WAR.
But I’m not just referring to WAR calculation, I mean FIP as a stat by itself.
A better example: Mariano Rivera has had a lower ERA than his FIP in 8 of the last 9 years. This is not because he’s lucky, it’s because he induces weak contact that is easy to field. But if I used ERA to say how awesome Mo is, a Sabrphile would likely snort and ask what his FIP was. But using Mo’s FIP is removing one of his greatest talents from his game.
I think most sabrphiles would agree that Mariano Rivera is one of the best pitchers ever, and would accept that his ERA being lower than his FIP every year is an outlier. An interesting one which we can probably glean something from, but an outlier nonetheless.
Squeeze,
If I recall correctly, Mariano and knuckleballers are for the most part the only two groups of pitchers to be able to maintain significantly lower than average BABIP. FIP is scaled for average BABIP, so we should probably take of a couple tenth of a point if the pitcher in question has shown a real ability to reduce BABIP. And then Mariano has also been able to do that while getting 54% GB and a ridiculous 6.3% HR/FB. The man is a freak of nature.
So I think you’re absolutely right, we should learn how in the holy dog fart he does this, but you just can’t point to him and say “see it doesn’t work” because he just so completely different from 99.9% of pitchers.
Exactly Wally – guys like Ichiro and Mo don’t break systems, you just need to realize that they exists so far into the margins that the systems don’t apply to them.
One question: what is the approximate “error” for WAR in a single season per player? In other words, if we were to compare two players that had a +3.3 WAR year and a +3.5 WAR year, we would not necessarily say the +3.5 WAR player is better because the difference is small enough in just one year. So, at point does the difference become significant for one year? If we take that error into account and apply it to the entire team, would the team error interval be big enough for the Twins that it would include the 2nd or 3rd place team?
I believe the Twins lead the Yankees in team WAR by about 1 WAR. Couldn’t that difference just be due to variance in, for example, UZR?
That’s not to say the Twins aren’t a very good team. I just think some care should be taken in using team WAR to determining which team is really the best in baseball.
They lead the Yankees by 7 WAR. They’re +3.2 WAR ahead of any other team, and their UZR only gives them +2.4 wins, so unless you think they’re a below average defensive team that UZR is missing on, that’s not happening here.
Mr. Cameron, I believe you misread your own site:
http://www.fangraphs.com/teams.aspx?pos=all&lg=al&stats=bat&type=6&season=2010&month=0
The difference between the Yankees and Twins as of this morning is 0.9 (27.0 to 26.1).
Thats only position players. Add in pitching.
whoops, yeah I only looked at position players.
Ranking by Run Differential … a measure of how many actual runs they team has scored and allowed, not how many they might be expected to score and allow based on isolation metrics.
AL
——-
[1] NYY — +164
[2] TBR — +144
[3] MIN — +96
[4] TEX — +93
[5] BOS — +66
[6] CWS — +61
Would we really consider CWS to be almost equal to BOS? Or would we say CWS run diff is almost as good because they play in a weaker division? Apply the same method to the Twins in comparison to TBR and NYY.
TEX is trailing MIN very slightly, but they did not have Lee for the full season or the RD might be 10-15 runs better (or even 20 or 30). I only bring that up, because they DO have Lee, and that’s a BIG factor when comparing the teams.
NL
—-
SDP — +128
ATL — +112
STL — +105
CIN — +80
SFG — +78
Not sure why we’re using wOBA, UZR, and FIP of the individuals … when we have metrics that show how the TEAM has performed in terms of scoring runs and preventing them.
Because performance with men on base, one of the big drivers of run scoring and prevention, is not predictive.
But, we’re not predicting whether the Twins are the best team in baseball. We’re deciding if they are based on what they’ve already done, correct?
I admit there could be something in this that I am not understanding. I am not compelling you to explain it all for me, particularly if it takes a lot of time, but …
Aren’t the Twins being “rewarded” for not increasing their production with men on base, as compared to the rest of the teams in the league.
I guess I’m not following the basis for “best team in baseball” or how the wOBA, FIP, etc is more predictive than all of the contextual stats accrued for an entire baseball team over the season up to date (surely, that’s a significant number of PA’s, both with and without men on base).
Are the Twins likely to just start mashing with men on base, when they haven’t elevated their performance in those situations over the course of the season? I guess I’m kinda confused on how their production (runs scored and runs allowed) wouldn;t serve as some type of predictive basis for how they might perform for the remainder of the season (and we’re only talking about this season).
If all of the other teams (or at least the top 5 or so teams we’re comparing the Twins to) hit significantly better with men on base, leading to more runs scored … wouldn’t that be an important aspect to consider when discussing who the best team in the league (or in the playoffs) is?
I don’t necessarily see a difnitive distinction between the ideas of “who has had the best season so far” and “who is the best team in baseball”, except maybe in an extreme surprise situation like the ’10 Padres.
FWIW, I am not making a statement in the form of a question, as if I know something I don’t. I am legitimately asking, because I am kinda confused on why would limit ourselves to the combination of the individuals’ metrics of wOBA, FIP, and UZR when trying to decide who the best team in baseball is.
The main thing you’re missing is that teams don’t “raise their production” or “elevate their performance” with runners on base. That’s not something hitters can control; “hitting with runners on base” is not a skill that can be separated from “hitting.” Over time, teams’ and players’ performances will be a bit better with runners on than without; when that isn’t the case, like with the Twins this year, they’ve just had bad luck in the distribution of their hits and such. You’re not “rewarding” them by ignoring it, you’re just refusing to punish them for something that’s out of their control.
Which is why if you don’t use “predictive” stats, “deciding if they are [the best in baseball] based on what they’ve already done” is meaningless. Say we’re playing poker, and you’re a professional who’s just been getting bad cards all night, and I’m a total novice who happened to win one huge hand by staying in with a 2-7 and ending up with 4 of a kind or something. Without bothering to separate factors that are “predictive” from the other stuff, your method would say that I’m the best poker player based on what we’ve already done, and that conclusion would be useless at best.
How (if at all) does the analysis change if you restrict the team compositions to reflect likely playoff rosters and player usage? I’m not sure how things actually look this year, but conceivably, a team with three (or two) really good pitchers and two (or three) much weaker pitchers could look weaker in terms of regular-season WAR than its playoff ability (relative to other playoff teams), right?
Maybe with me on the shelf the twins wont get owned by the yanks in the postseason……NAH
I’m a Twins fan but I don’t see how the Twins are the best team in baseball. They’re among the best but (and I know this is contrary to much written here) they just aren’t very good against the best teams. They play tight (and almost all Twins fans agree w/ this)? They’re balanced (yes)? Can they match-up w/ the best yes, will they? I wouldn’t bet a large sum of money they would. However I would have bet a large sum of money they’d be in the running to be in the exact position they’re in today. In the lead, trying to make the playoffs and pushing for a run at the best teams in the playoffs.
For what’s it’s worth, this team managing to put up these numbers with their injury issues is remarkable.
The list of position players that have been put on the DL at least once this season for Minnesota:
Justin Morneau – 5.2 WAR, .446 wOBA
Orlando Hudson (x2) – 3.3 WAR, .342 wOBA
J.J. Hardy (x2) – 1.8 WAR, .306 wOBA
Nick Punto (x3) – 1.8 WAR, .292 wOBA
Alexi Casilla – .8 WAR, .343 wOBA
The first four are the most noteworthy. Not a single member of the starting infield has missed less than a full month of playing time and most of them have missed much more than that now. The back-ups have missed plenty of time, too.
To compare, the Yankees have had three starting position players overall go on the DL the entire season – each just once (Rodriguez – .351 wOBA, Granderson – .335 wOBA, Posada – .366 wOBA). The Rays also only three (Bartlett – .303 wOBA, Kapler – .264 wOBA, Pena – .343 wOBA), one of them, Kapler, twice. The Rangers only two (Kinsler – .359 wOBA and Nelson Cruz – .412 wOBA).
It would be cool if you could sort total team WAR under the Teams section. You can sort batters and pitchers WAR and then add them up and sort on your own, but it would just make it easier and would update on its own.
Somebody accurately pointed out that Scouts are way better at projecting baseball players as amateurs, than stats are. True. Very true. Using Hanley Rameriez as the example.
That said, one huge shortcoming of stats, is that stats never, NEVER, predict greatness out of any player. Yet, when you study most HOF players, they far exceeded their projections and performance from the Minor leagues. Great players in MLB get better in MLB and the stats crowd never predicts this.
I would like to see the stats crowd speak more about the relative probabilities of player X turning out poor or great. Too often the stats crowd speaks only of the 50% probably range and this fails to acknowledge that all players have a range. It is scouts that obsess over the range and often can articulate the probability of greatness or suckyness better than the stats crowd can.
Give me a scout over a stats nerd any day. The scouts can easily learn the stats and incorporate those, while the nerds and computers can’t ever learn the qualitative portion of the analysis and art of projection.
Well said, TedWilson! Modern day scouts use of stats (some very different approaches than SABR) to assist their qualitative evaluations of players would be surprising to the author’s comments made on this thread; some of which put forth as absolutes just do NOT mesh with multi-period,observed behavior & outcomes on the field:
[paraphrase]: Hitting with men on base is not a skill above merely hitting.
I’ve come to the conclusion, based on the dialogue, this site & core crowd may be focused on fantasy baseball teams, thus the decidely obvious bias towards WAR, FIP, & OBA over Win Expectancy.
Aggregated statistics & their rate based indications certainly have their place, but in the real world of actual games played by real human beings & managed on the field based on outcomes & experiental odds, it’s simply not sufficient nor accurate to hold onto SABR doctrine based on more clinically derived, historical data.
Real humans do choke or rise up to the challenge of batting with men on base.
The data collection or analysis methods used in traditional SABR may simply not capture that underlying behavior your either dismiss as noise or is inherently interaction variables outside of your current definitions.
“Real humans do choke or rise up to the challenge of batting with men on base.”
Which ones?
Isn’t a bit of a logical fallacy to say that the team with the best record isn’t the best team? Especially, if you are using advanced metrics which are based on “win probabilities,” related in turn to run scoring?
Wins are the standard against which all else is judged. We have to remember that advanced stats are simply attempts to figure out what causes wins. Numbers which predict wins cannot be superior than actual wins….
Where do the Mariners rank? #6?
I’ve embraced advanced stats for a few decades now, yet I try to keep a proper perspective. At no point this year have the Twins been better than the Yankees. They are not a better team than the Yankees. That’s obvious.
no, the twins are the best team in baseball, the yankees are just cheaters and payed off the umpires.
What superb text! I have no clue how you were able to write this text..it’d take me days. Well worth it though, I’d assume. Have you considered selling banners on your website?
Who’s to say who the best team is? The team that wins the World Series?…..Uh…no! The Cardinals had only the 8th best record in baseball ,not to mention their only a National league team. And don’t tell me what happens during the season doesn’t count! If that is the case,why even play one?