Friday, December 30, 2011

Tigers Interested in Garza

When Nick Cafardo recently wrote in the Boston Globe about the Tigers possibly being interested in Cubs right hander Matt Garza, it sounded like speculation to me so I didn't take it too seriously.  There apparently might be something to it though as Jon Paul Morosi of Fox Sports is now reporting the same thing.  As you might expect, there are several other teams who are interested including the Yankees and Blue Jays.

The 28-year-old Garza posted a 3.32 ERA and impressive 197/63 K/BB ratio in 198 innings last year.  He also finished among the National League's top ten pitchers with a 2.95 FIP and 20 Base Runs Saved.  More importantly, Garza has maintained consistent success since becoming a regular starter averaging 197 innings and a 3.72 ERA over the last four seasons.  Each year, his ERA has been below 4.00. 

Garza's arsenal features a fastball with an average velocity of 93.8.  For comparison purposes, Max Scherzer averages 93.1 MPH.  Garza's second best pitch is his slider and he also throws a curve ball and change up.

Garza is clearly a very solid number two or three starter on a contending team.  The catch is that he is just two years away from free agency and will be costly in terms of prospects.  The Tigers might not have to gut their entire system for two years of Garza as they would have for four years of Gio Gonzalez.  However, the Cubs would likely require Tigers prized righty Jacob Turner as part of the deal. 

Should the Tigers give up Turner for Garza?  One argument is that they would be trading six years of control over Turner for two years of Garza.  On the flip side, Turner may never be as good a Garza already is.  I don't worry too much about years of team control when it comes to pitchers, because they are so fragile and unpredictable, you just don't know where they'll be two years from now.  Never mind six. 

The bigger question for me is, do you want to trade your best prospect trading chip for two years of Garza or do you want to save him for something bigger?  It might depend on what else they can do.  If they are able to acquire Garza and also make an upgrade or two offensively, this would put them in really good position to win it all the next couple of years. Otherwise, they might be better off waiting to see how things look prior to next season's deadline.

Thursday, December 29, 2011

Shutdowns and Meltdowns: Tigers Top Relievers Since 1974

Going back to my earlier Shutdown/Meltdown theme, I wanted to look at Tigers relievers historically.  Remember that a pitcher gets credit for a shutdown (SD) whenever he substantially enhances his teams chance of winning a game (sum of Win Probability Added for the game is 6% or better).  He is charged with a Meltdown (MD) if he significantly hampers his team's chance of winning  (sum of Win Probaility Added for the game is less than -6%).

Before I present the historical SD/MD leaders, there are a couple things to keep in mind.  First, Win Probability Added data are only available back to 1974.  So, anybody pitching prior to that year will be excluded.

Another issue is that relievers were used a lot differently back in the 1970s and even into the 1980s.  In contrast to today where closers typically enter games with nobody on base and pitch one inning, earlier top relievers would often come into games with runners on base and pitch multiple innings.  So, earlier relievers had more opportunities for Meltdowns than contemporary relievers and thus typically didn't do as well on the SD/MD statistic. 

Table 1 lists the Tigers SD/MD leaders since 1974.  1984 MVP winner Guillermo Hernandez tops the list with an amazing 10.2 (41/4) ratio.  The second best ratio belongs to Jose Valverde with a 7.6 (38/5) ratio in 2011.   

Table 1: Tigers Single-season SD/MD Leaders, 1974-2011


Pitcher
Year
SD
MD
SD/MD
Guillermo Hernandez
1984
41
4
10.2
Jose Valverde
2011
38
5
7.6
Todd Jones
2000
35
5
7.0
Mike Christopher
1995
12
2
6.0
Fernando Rodney
2009
35
7
5.0
Jose Valverde
2010
23
5
4.6
Todd Jones
1999
27
6
4.5
Aurelio Lopez
1979
34
8
4.2
Jerry Don Gleaton
1990
21
5
4.2
Todd Jones
2007
29
7
4.1
Todd Jones
1997
32
8
4.0
Joel Zumaya
2006
35
9
3.9
Todd Jones
1998
26
7
3.7
Mike Henneman
1991
29
8
3.6
Jose Lima
1996
14
4
3.5
Data from: FanGraphs.com

Todd Jones had five of the top 15 seasons - 1997, 1998, 1999, 2000 and 2007.  There are a couple of factors involved in Jones' success.  First, the Roller Coaster was not used in a lot of situations where he had chances for Meltdowns.  Just as importantly, he was very good at his role and therefore was able to get a good number of Shutdowns.

The biggest surprises on the list are Mike Christopher in 1995 and Jose Lima in 1996.  Neither was very good according to traditional statistics, but a look at their game logs shows that they pitched well in high leverage situations.  It's probably just a fluke that they pitched their best when they were needed the most, but their good SD/MD ratios indicate that they probably had more value to their teams than the more traditional measures indicate.

Wednesday, December 28, 2011

Regular-Season Performance Versus Post-Season Success - Part 2

In my last post, I looked at the correlation between regular-season win totals and post-season success and found no clear association between the two.  Teams seemed to have equal likelihood of winning their divisional series, winning a pennant and winning the World Series regardless of their win totals in the regular season.

A better measure than wins of the dominance of a team is run differential (runs scored minus runs allowed), so I'm going to do a similar analysis using that.  Here, the 128 playoff teams between 1996-2011 are divided into approximately equal-sized groups (called quartiles) based on run differential.  The first quartile contains the bottom 25 percent of teams, that is, teams with a run differential of 77 or less.  The second, third and fourth quartiles include teams with run differentials of 78-114, 115-149 and 150+ respectively.

Table 1 shows the relationship between run differential and post-season performance.  The first rows tell us the following:  33 teams had run differentials of 77 or less, 16 (48%) triumphed in the divisional series, 7 (21%) won the pennant, and 5 (15%) were world champions.  As we saw with wins, the percent of teams winning in the first round of playoffs does not change very much across quartiles. 

However, teams with run differentials of 150 or more were almost three times as likely as the 77 and under group and nearly twice as likely as the middle quartiles to win the pennant.  The 150+ quartile was also most likely to win the World Series, but not by much: 18% for 150+ versus 15% for 77 and under. The results are not statistically significant, in part because 16 years is not a large sample. 

Table 1: Post-Season Performance by Regular Season Run Differential

Regular
Season
 Run Diff
Teams
Divisional Series
Pennant
World Series
<= 77
33
16(48%)
7 (21%)
5 (15%)
78-114
33
16 (48%)
10 (30%
3 (9%)
115-149
29
13 (45%)
4  (31%)
2 (7%)
150+
33
19 (58%)
11 (58%)
6 (18%)

Data from Baseball1.com

In Table 2, the teams are divided into groups according to where they finished in run differential among the four teams in their league each year.  Here, we see stronger evidence (although still not statistically significant) of more dominant regular-season teams doing better in the post-season.  The top-ranked teams had a higher likelihood than the bottom-raked teams of winning the divisional series (66% versus 47%), pennant (38% versus 19%) and World Series (22% versus 9%).

The results of this analysis are more in line with what you might expect.  The less dominant teams do have a decent chance of post-season success. Indeed, nearly half (47%) of the fourth-ranked teams got by the first round of the playoffs.  Additionally 7 of the 16 World Series winners ranked third or fourth in run differential.  Similarly, the strongest teams can fail in the playoffs - about a third (34%) of the top-ranked teams did not get past the first round.  However, we are seeing some evidence that teams with the best run differential do have a somewhat better chance at playoff success.   


 Table 2: Post-Season Performance by Regular Season Run Differential Rank



Regular
Season
Run Diff
Teams
Divisional Series
Pennant
World Series
Fourth
32
15 (47%)
6 (19%)
3 (9%)
Third
32
16 (50%)
7 (22%
4 (12%)
Second
32
12 (38%)
7 (22%)
2 (6%)
First
32
21 (66%)
12(38%)
7 (22%)

Data from Baseball1.com

Tuesday, December 27, 2011

Regular Season Records Say Nothing About Post-Season Success

I often hear fans say something along the lines of: "The Tigers are good enough to get into the playoffs, but they don't have enough to win the World Series."  This suggests that the Tigers might have enough talent to win 90 games and a division title, but are not strong enough to beat a stronger team from another division.  Recent history has shown, however, that it doesn't take a dominant team to win the World Series.

Two examples of teams with modest regular-season success winning the World series happened the last two times the Tigers were involved in the playoffs.  When the 95-win Tigers won the American League pennant in 2006, they went on to lose the World Series to the Cardinals, a team which won just 83 games during the regular season.

This past season, the Cardinals miraculously seized the wild card spot on the last day of the season with 90 wins.  The Red Birds proceeded to eliminate the Phillies (winners of 102 regular season games) and eventually won the World Series again.

Despite the examples above, one would think that the most dominant regular-season teams would win most often in the post-season.  To test this theory, I looked at every playoff team since 1996.  This includes all years since the expanded playoffs began except the strike-shortened 1994 and 1995 seasons.

The results of all 128 playoff teams between 1997-2011 are shown in Table 1 below.  The ninth row of the table is read as follows: There were 12 playoff teams with 90 wins during the season, four got past the divisional series, three won the pennant and one was the world champion.  The tables tell us that the regular-season win totals of the 16 world champions were: 83, 87, 90, 91, 92, 92, 92, 92, 92, 96, 98, 98, 99, 99, 103, 114.  Just looking at the win totals of the World Series winners makes it appear as if their not a strong association between winning in the regular season and winning the World Series.

An examination of the divisional series and pennant columns reveals a similar story.  In fact, the correlation between regular season wins and post-season success is only 0.11 and is statistically insignificant.  In other words, how often a team wins in the regular season tells us nothing about what they will do during the playoffs.

Table 1: Post-Season Performance by Regular Season Win Totals

Regular Season Wins
Teams
Divisional Series
Pennant
World Series
82
1
0
0
0
83
1
1
1
1
84
2
1
0
0
85
1
0
0
0
86
1
1
1
0
87
2
1
1
1
88
7
4
0
0
89
4
2
1
0
90
12
4
3
1
91
9
2
1
1
92
10
6
5
5
93
5
2
1
0
94
7
3
1
0
95
16
10
3
0
96
10
5
3
1
97
12
6
1
0
98
5
4
3
2
99
3
2
2
2
100
4
1
0
0
101
5
3
1
0
102
3
0
0
0
103
4
2
2
1
105
1
1
1
0
106
1
1
0
0
114
1
1
1
1
116
1
1
0
0
Totals
128
64
32
16

Data from Baseball1.com

To illustrate the above correlation (or lack thereof) more clearly, regular-season win totals can be divided into categories: < 90 wins, 90-94, 95-99, 100+.  Table 2 below shows the data for all 128 playoff teams.  The first row is read as: 19 teams with less than 90 wins, 10 (53%) took the divisional series, four (21%) won the pennant and two (11%) were world champions.

You can see that the percentages don't change very much as you move up the ladder from teams with fewer than 90 wins to 100-plus-win teams. The results in row are fairly close to what you would expect if every series were a coin flip: 50% chance of winning the divisional series, 25% chance of winning the pennant and 12.5% chance of winning the World Series.

Table 2: Post-Season Performance by Regular Season Wins Category

Regular
Season
 Wins
Teams
Divisional Series
Pennant
World Series
< 90
19
10 (53%)
4 (21%)
2 (11%)
90-94
43
17 (40%)
11 (26%)
7 (16%)
95-99
46
27 (59%)
12 (26%)
5 (11%)
100+
20
10 (50%)
5 (25%)
2 (10%)
Data from Baseball1.com

Post-season success is looking even more random than I expected, but let's do one more thing.  Instead of just taking win totals across seasons, we can rank the post-season teams within each season. For example, the following ranks would be assigned to NL teams in 2011: (1) Phillies 102 wins, (2) Brewers 96, (3) DiamondBacks 94, (4) Cardinals 90.  

The results for all years are shown in Table 3 below.  The first row reads as follows: There were 36 teams ranked first in wins (including ties), 21 (58%) took their divisional series, 10 (28%) won the pennant and 5 (14%) were world champions.  As in Table 2 above, there is not a lot of variation from row to row.  The bottom ranked teams are as about as likely to win post-season as the top ranked teams. Again, the results are similar to what you would see if the series were determined by coin flips. 


Table 3: Post-Season Performance by Regular Season Rank

Regular
Season
Rank
Teams
Divisional Series
Pennant
World Series
First
36
21 (58%)
10 (28%)
5 (14%)
Second
32
14 (44%)
8 (25%
5 (16%)
Third
33
13 (39%)
7 (21%)
3 (9%)
Fourth
27
16 (59%)
7 (26%)
3 (11%)

Data from Baseball1.com
 

So, it seems clear that regular-season win totals give us no indication of what to expect in the playoffs.  That's not to say that the playoffs are a total crapshoot and that there is no difference between a World Series contending team and a playoff-contending team that doesn't win the World Series.  It's entirely possible that certain types of teams have a greater likelihood of success in the playoffs regardless of win totals.  Perhaps, teams with very wide run differentials (more so than their wins would indicate) or teams with strong pitching or any number of other types of teams have a greater likelihood of success in post-season.  I'll explore some of those possibilities in the future.

Still, the total lack of correlation between regular-season wins and playoff wins is pretty remarkable.

Sabermetrics Book

Sabermetrics Book
One of Baseball America's top ten books of 2010

Blog Archive

Subscribe

501 Baseball Books

501 Baseball Books
Recommended by Tiger Tales

Stat Counter

Site Meter