Sunday, May 25, 2014

Sam Demel: Starting pitcher


Sam Demel, who began the year as the Isotopes closer, has been starting recently.

This is somewhat interesting.
Stephen Fife is on the DL
Magill has been moved to the bullpen
These were the two guys most likely to be called up to the majors to make starts.

Red Patterson got an emergency start earlier in the year.
Lee has looked like a work in progress.
Is Demel next in line for major league starts should the need arise?

He's pitched 63 major league innings, as a reliever, with the Diamondbacks in 2010-2012.  He was a starter all through is minor league career.  His two starts with Albuquerque over the last week were his first two professional starts.

As a reliever this season, Demel faced 89 hitters.  He walked 7.9% of them while striking out 27.0%

In his two starts he's faced just 34 hitters.  He walked 11.8% of those hitters and struck out 14.7%.

No particularly impressive, but it'll be something to keep an eye on.


http://www.milb.com/milb/stats/stats.jsp?pos=P&sid=t342&t=p_pbp&pid=451585
http://minorleaguecentral.com/player?pid=451585&type=pitcher




Monday, December 23, 2013

What is an Ace? 2013

After the 2011 season I asked, and attempted to answer, the question, "what is an ace"?

It's time to do that again.




Ok.  While Kershaw is the aciest of aces right now, that's not really the answer that we're looking for.

What I did was to take the average of every starter's fWAR and RA-9 WAR.  Then I used that number to group pitchers into groups of (roughly) 30 - 30 aces, 30 number 2's, etc.  Then, I looked at the average performance of the pitchers in each group.

Here's what I found:
*click to embiggen

There's a couple of interesting things to note.

One is that the best 30 pitchers in baseball are, far and away, the best group.  They strike out the most hitters, they walk the least hitters, they give up the least home runs, they have the lowest BABIP, they're the best.  That's not surprising when guys like the above-picture Kershaw, Cliff Lee, Max Scherzer, Justin Verlander, Matt Harvey and Yu Darvish are in the ranks.

The second interesting thing is how similar the #3, #4 and #5 groups are in terms of performance.  Look,
#3 18.2% K, 7.2% BB, 3.85 ERA, 4.06 FIP, 4.04 xFIP, 4.13 SIERA
#4 18.7% K, 8.2% BB, 3.89 ERA, 3.86 FIP, 3.96 xFIP, 4.09 SIERA
#5 17.4% K, 6.9% BB, 4.26 ERA, 4.09 FIP, 4.02 xFIP, 4.12 SIERA

In many ways, every way other than walks really, #4 starters outperformed #3 starters.  Well, in every way except for number of starts and innings.  Number 3 starters made about 7 more starts and pitched almost 50 more innings than #4 starters.  Similarly, #5 starters were a little worse than both #3 and #4 starters but what really limited them from producing value was that they made 12 less starts and pitched half as many innings as #3 starters.

The third point is similar to the above.  Starters not in the top five accounted for more starts and more innings than the best pitchers in baseball.  That makes sense when you stop to think about it, there are more bad pitchers than elite ones, but we don't think about just how important it is for the other starters to make their starts so these guys don't have to.


As I mentioned when I first did this little exercise after the 2010 season,

Next time your team signs a pitcher with a 10 - 8 record and 3.99 ERA in 160 innings realize just what you are getting.  One of the top 100 pitchers in the league.  

The numbers are a little different now - now the average #3 is 10 - 9 with a 3.85 ERA in 158 innings - but the point remains the same; the average baseball fan vastly underrates pitcher performance.







Contract Analysis: Chris Perez (&Jamey Wright)

The Dodgers have agreed to a deal with Chris Perez.

Perez is the former Cleveland Indians closer who was also busted for ordering shipments of marijuana to come to his house under his dog's name.

When Perez was first released by the Indians I hoped the Dodgers would give him a look.  Obviously, they did.

Unfortunately, I took another look as well.

Over the last 3 seasons Perez:
K'd 20.7% of hitters, walked 8.6%, gave up line drives 21% of the time, 3.73 ERA, 4.21 FIP, 4.18 xFIP, 3.67 SIERA, 0 fWAR, 1.4 RA-9 WAR

Let's compare those numbers to the numbers of recently released Ronald Belisario:
K'd 20.1% of hitters, walked 9.7%, gave up line drives 21% of the time, 3.24 ERA, 3.36 FIP, 3.46 xFIP, 3.16 SIERA, .7 fWAR, .9 RA-9 WAR

*stats from FanGraphs

Belisario is, more than arguably, better than Perez, but the Dodgers decided to let Belisario walk away. I don't understand.

But, that's the past.  How about the future?
STEAMER and Oliver projections project:
Perez for 21.4% K's, 8.3%BB's, 4.12 FIP and -.1 WAR per 55 innings
Belisario for 19.4% K's, 10.1 BB's, 3.73 FIP and .2 WAR per 55 innings

Belisario still comes out better.

Terms of Perez' deal haven't been released, but hopefully it's nothing more than a couple of million with incentives.




Jamey Wright also signed.  I like this move better.
How do his numbers look?
Last 3 seasons:
K'd 19% of hitters, BB'd 9.4%, gave up line drives 19% of the time, 3.32 ERA, 3.60 FIP, 3.69 xFIP, 3.41 SIERA, .6 fWAR, 1.3 RA-9 WAR.

and STEAMER and Oliver projections?
18.8% K's, 9.0% BB's, 3.64 FIP, 0.9 WAR per 55 innings.

*stats from FanGraphs

He looks like the best of the bunch.


That leaves the Dodger bullpen looking something like:
Closer - Kenley Jansen 68.3 innings, 13.0 K/9, 2.9 BB/9, 2.18 ERA, 2.35 FIP, 1.2 WAR
HL R - Brian Wilson 32.3 innings, 8.6 K/9, 3.6 BB/9, 3.27 ERA, 3.39 FIP, 0.3 WAR
HL L - Paco Rodriguez 46.7 innings, 10.9 K/9, 3.6 BB/9, 3.02 ERA, 3.10 FIP, 0.4 WAR
ML L - JP Howell 50.7 innings, 7.9 K/9, 3.7 BB/9, 3.42 ERA, 3.64 FIP, 0.1 WAR
ML R - Chris Perez 55.7 innings, 8.3 K/9, 3.2 BB/9, 3.88 ERA, 4.12 FIP, -0.1 WAR
LL R - Jamey Wright 42 innings, 7.3 K/9, 3.5 BB/9, 3.61 ERA, 3.64 FIP, 0.1 WAR
LL R - Brandon League 61.7 innings, 6.2 K/9, 2.9 BB/9, 3.76 ERA, 3.72 FIP, -0.1 WAR

With the following waiting in AAA
Chris Withrow - 64.3 innings, 9.5 K/9, 4.7 BB/9, 3.58 ERA, 3.81 FIP, 0.1 WAR
Jose Dominguez - 49.3 innings, 8.9 K/9, 5.3 BB/9, 4.15 ERA, 4.15 FIP -0.1 WAR
Onelki Garcia 47.7 innings, 8.0 K/9, 5.6 BB/9, 4.37 ERA, 4.53 FIP, -0.4 WAR
Scott Elbert - 24.3 innings, 8.5 K/9, 3.7 BB/9, 3.52 ERA, 3.84 FIP, 0.0 WAR
Javy Guerra - 55.3 innings, 7.3 K/9, 3.7 BB/9, 3.98 ERA, 3.95 FIP, -0.2 WAR

*Combined projections from STEAMER, Oliver, and ZiPS

These moves push Chris Withrow to AAA.
Elbert and Guerra are out of options and will need to be traded or released.
Rule V pick Seth Rosin doesn't look to have a spot and will have to be offered back to the Mets (along with $25000) or a trade will have to be worked out.



Saturday, December 21, 2013

2013 Pitcher Plus

I've taken the following stats: line drive %, ground ball %, fly ball %, infield fly ball %, home run per fly ball %, walk %, strike out %, SIERA for every player that pitched in the majors in 2013 and divided their rate by the rate for their league.  I further separated starters and relievers.

Two examples,
*click to embiggen



For Kershaw, his LD%+ of 108% means that he gave up line drives at 108% of the league rate - 8% more than the average NL starter.  His BB%+ means that he walked hitters at 66% of the rate of the average NL starter.

Kenley Jansen's SIERA indicates that his SIERA was 41% of the league average SIERA.


Here's the link to the complete spreadsheet.


Wednesday, December 18, 2013

A quick note on lead-off hitters

The prevailing wisdom for a long time was that lead-off hitters needed to be fast and steal lots of bases.

The modern wisdom is that lead-off hitters need to get on base.


I was involved in a discussion regarding this and broke out the Excel sheet so I figured that I'd do a quick post.


I took a look at how often runners score compared to their ability to steal bases and their ability to get on base.  I looked at all hitters from 2013 who had more than 200 PA's.  I looked at runs scored/PA.  I looked at SB/PA.  And I looked at OBP.

Here's what I found.



You can see that in the top graph of Runs & OBP.  As OBP (the horizontal axis) gets bigger the number of runs per plate appearance also gets bigger.  

In the bottom graph as the number of stolen bases per plate appearances per plate appearances gets bigger so does the number of runs per plate appearances.  But the relationship is not nearly as well defined as in the top graph.

The number in each graph, the r-squared value, tells the strength of the relationship between the two variables.  A perfect correlation where in each instance as one variable goes up so does the other would have an r-squared of 1.  A correlation where the two variables had nothing to do with each other would have an r-squared of 0.

We can see that the r-squared for the bottom graph (stolen bases and runs scored) is .09.  That's very close to zero, meaning there's very little relationship between stolen bases and the number of runs scored.  In the top graph, the r-squared is .34.  That's a much stronger relationship between OBP and number of runs scored.