Recently, we took a look at wRC+ (Weighted runs created plus) and how it was a nice indicator of offensive value for a team. Using wRC+, we could see that not only do the Minnesota Twins have two elite offensive assets in Josh Willingham and Joe Mauer, but the rest of their potential 2013 roster is filled with above average offensive contributors as well.
Using this metric, I feel comfortable in saying that the 2013 Twins will be a very good offensive ball club. Unfortunately, the other team gets to bat too.
Pitching is the obvious area of concern when it comes to the 2013 Twins. Their rotation was historically bad in 2012, consisting of a rotating cast of throw away arms, journeymen and AAAA level players. The Twins have addressed the issue in 2013 by… picking up more journeymen. Yikes.
By utilizing another advanced metric – FIP (Fielding independent pitching) we can see just how the 2012 Twins pitchers stacked up against the rest of the league. A variation of that metric, xFIP (Expected fielding independent pitching) can then be used alongside FIP to project the 2013 seasons of the Twins’ starters.
Warning: Fun times with math and numbers ahead!
Before I get too far ahead of myself, let’s go over what exactly FIP is and what it’s composed of.
Think of FIP as a more accurate form of ERA. While ERA, obviously, indicates the average earned runs a pitcher allows over each of their starts, it erroneously charges pitchers with runs that they could do nothing to prevent. For example, poor defensive plays, or random chance hits. FIP removes the randomness and only grades the pitcher on controllable outcomes, walks (both BB and HBP variety, IBB are not counted as they are intentional decisions by a manager – not controllable by a pitcher) strikeouts and home runs.
This metric ends up giving a very clear picture of the elite pitchers, while still staying within the ERA ‘realm’ to help people understand what numbers mean what. During the 2012 season, the league average FIP was at or around 4.00. Anything above 4.00 is below average, anything lower than 4.00 is above average.
League average ERA in 2012? 4.01 – yup, this metric works.
Now, onto the Twins.
In 2012, only 1 pitcher from the Twins tabbed enough innings to even be eligible in the standard FIP comparisons: Scott Diamond. Diamond posted a FIP of 3.94, slightly above his ERA of 3.54 in 2012. This is accurate when you consider what type of pitcher Diamond is – he does tend to give up home runs, relies quite a bit on fielding (Ground ball double plays are his friend) and he limits his walks. As such, FIP is showing that Diamond is slightly better than a league average pitcher – which I think is an agreeable assessment for any honest Twins fan.
Here’s how the other Twins pitchers looked in 2012:
(Note: This list is composed of starters who pitched at least 50 innings in 2012 and contains only starters)
(Blogger aparrently jacks up Excel charts when they are copy / pasted over - they're still discernable, so I'll leave them, sorry for making extra work when you read these).
(Blogger aparrently jacks up Excel charts when they are copy / pasted over - they're still discernable, so I'll leave them, sorry for making extra work when you read these).
FIP
ERA
Scott Diamond
3.94
3.54
Francisco Liriano
4.3
5.34
Carl Pavano
4.38
6
Cole DeVries
4.81
3.99
Brian Duensing
4.81
6.92
P.J. Walters
5.48
5.69
Samuel Deduno
5.5
4.44
Liam Hendriks
5.57
5.59
Nick Blackburn
6.09
7.39
A few things jump off the page right away. We’ve already discussed Diamond, and how his metrics are close enough together to show that Diamond was exactly as he seemed last season.
Francisco Liriano and Carl Pavano, however, were not as lucky. Their FIP (4.30 & 4.38 respectively) is significantly lower than their actual ERA. What does this mean? Well, simply put, they were really unlucky. Liriano and Pavano were both charged runs that were not their fault. The likely cause of this is a porous defense that the Twins call their middle infield allowing balls in play to become runners rather than outs. FIP does not charge those to a pitcher, while ERA will charge those runs.
Brian Duensing also makes the list as an unfortunate casualty to the Twins poor middle infield. His FIP, while still below average at 4.81 is still better than his sky-high 6.92 ERA. Essentially, FIP tells us that Duensing is a below average starter, but not as bad a starter as his ERA would indicate.
Alternatively, a few Twins pitchers came away looking better than they really should. Cole De Vries posted a slightly below average 4.81 FIP, but his ERA was only 3.99. Essentially, FIP is saying De Vries was lucky last year. He gives up too many home runs and walks – this reflects in his metric. While his ERA was ‘decent’ at 3.99, his FIP says we shouldn’t expect that year in and year out.
Samuel Deduno is another starter who’s ERA doesn’t tell the full story. Deduno’s 4.40 ERA, while not ideal, doesn’t show just how lucky he was in 2012. His 5.5 FIP reflects his high walk totals and home runs allowed. Like De Vries, the Twins should not expect Deduno to limit runs to the extent he did in 2012.
Finally, players such as Liam Hendriks, PJ Walters and Nick Blackburn were exactly what their ERA said they were – bad. (It should be noted that Hendriks posted a not-great-but-improving 5.02 FIP in the 2nd half of 2012. His first half FIP of 6.92 explains why his metric stayed so high at season’s end).
So, there you have it – FIP shows us a better picture of who the 2012 Minnesota pitchers were. Only slightly above average starter and a platoon of below average, poor and awful starters.
Except that’s not the whole story – not until we work in xFIP. What’s the difference between xFIP and FIP? (The x)
I’m kidding!
The ‘x’ stands for ‘expected’ and it means that the number of home runs a player SHOULD have given up are also factored into their FIP. This is calculated by taking the pitcher’s fly ball rate and applying it against the league average fly ball rate. Home runs are random and tend to vary from year to year. Therefore, it’s hard to make a reliable metric that will measure home runs on a consistent basis. What you can measure, however, is a player that lets the ball carry to the outfield too often. These fly balls have a greater chance of being homeruns and thus can accurately be counted against a pitcher as potential future runs allowed. Got it? Good. Let’s add xFIP to the equation and see what we get.
Name
|
FIP
xFIP
Scott Diamond
3.94
3.93
Francisco Liriano
4.3
3.93
Carl Pavano
4.38
4.48
Brian Duensing
4.81
4.48
P.J. Walters
5.48
4.54
Cole DeVries
4.81
4.71
Samuel Deduno
5.5
4.73
Liam Hendriks
5.57
4.75
Nick Blackburn
6.09
5.1
Once again, league average is 4.00
Now the picture gets a little more interesting – furthermore, you can see that these metrics really work. How so? Look how much the Twins’ starters xFIP drops when expected home runs (IE Fly ball rates) are factored in. By and large, Twins pitchers are ground ball pitchers. They’re taught to "pitch to contact", put the ball in play and let the defense make stops. Most starters do not allow a lot of fly balls in the Twins system.
You can see that above, when fly ball rates are factored in, the Twins starting staff drops closer to the average range. In fact, Francisco Liriano would have joined Scott Diamond as an above average pitcher. Remember, however, that xFIP is a better predictor of FUTURE performance and less on what actually happened. Looking at it this way, you can see that 2012 was almost a statistical anomaly across the board. Twins starters SHOULDN’T struggle like they did in 2012.
So, by xFIP, the Twins staff is slightly below average, but not horrid, which is likely what the front office thought when they went into 2012. Unfortunately, a majority of their staff underperformed in 2012, allowing more fly balls and home runs, thereby inflating their FIP.
So, how do we use these numbers to forecast future output? xFIP is a nice correlator to future numbers, as previous seasons xFIP usually ends up being a close approximation to next season’s ERA & FIP. Luckily, we don’t have to crunch the numbers and analyze xFIP to generate 2013 FIP and ERA. Bill James does that already. By his calculations, the Twins staff will put up FIP like this:
Name
|
FIP
Scott Diamond
3.49
Liam Hendriks
3.61
Vance Worley
3.8
Mike Pelfrey
4.2
Cole De Vries
4.22
Samuel Deduno
4.22
Kevin Correia
4.29
NOTE: Mike Pelfrey’s FIP is his career average; projections for 2013 were not available.
NOTE: Kyle Gibson was not included as minor league data was not used in 2013 projections.
Ugh! Right? Pelfrey, De Vries, Deduno, Correia are all slightly below average pitchers while only Diamond, Hendriks(!) and Worley project to be a little better than average.
It can be argued that this is a good thing, however. Think about it, THIS:
Name
|
FIP
Scott Diamond
3.94
Francisco Liriano
4.3
Carl Pavano
4.38
Brian Duensing
4.81
P.J. Walters
5.48
Cole DeVries
4.81
Samuel Deduno
5.5
Liam Hendriks
5.57
Nick Blackburn
6.09
Was 2012’s list. By comparison, 2013’s projections aren’t so bad. If Correia, Pelfrey and De Vries or Deduno can be just below average while Hendriks, Worley and Diamond all pitch at or above average, the rotation will actually be vastly improved from 2012.
Of course, 2012 FIP and xFIP indicate that the Twins shouldn’t count on De Vries and Deduno to deliver numbers like that…
And there’s the downfall of projected statistics. There’s last year’s metrics, there’s this years projections – sometimes they simply don’t line up. If predicting next years stats was so easy, the game would be a lot less fun to watch anyway!
Where do all these numbers and projections leave us? I’ll sum it up in a few bullet points:
- We can expect more of the same from Scott Diamond
- Vance Worley is the solid starter the Twins needed
- Liam Hendriks is going to make the jump to reliable Major League starter
- The Pelfrey and Correia signings were as underwhelming as we all thought
- 2013’s staff will be better than 2012 by default, 2012 was as bad as it can get
So, if the offense is on track and the pitching will be improved, but still a little below league average, can we assume the Twins win more games in 2013 than 2012?
Maybe.
First, we have to look at one more metric. Defense. Look for that post in the coming days.
No comments:
Post a Comment