Will Carroll has been expressing his frustration with the sabermetrics community in a of couple of articles as of late. The second article will not go down as his best work and it was surprisingly bitter and petty. Many have responded including Tom Tango at the Book Blog, Carson Cistulli at FanGraphs, Josh Garoon at Baseball Time in Arlington and Minda Haas at Royal Blue. They have done a good job critiquing his inaccuracies and unfair generalizations, so I won’t dwell on that. If you can get past all of the hyperbole though, Carroll does bring up up interesting points about the issue of communicating the world of sabermetrics to a broader audience. Sabermetrics is a very difficult subject and its creators have not always done a great job explaining the ideas very clearly. The sabermetrics community is coming up with new statistics seemingly every week and many, if not most, readers have been overwhelmed at times. The main reason I wrote Beyond Batting average is to clear up much of the confusion for those that do have an interest in learning sabermetrics My book was not targeted to the mainstream, but rather a niche audience of knowledgeable fans with an interest in learning more about statistics. Beyond Batting average has been pretty well received and I have been told that it is quite accessible. I think I have achieved my goal of educating intelligent and curious readers. If you have already decided that you want to learn more about sabermetrics, then my book is a good place to start. If you don’t have any interest in sabermetrics, then my book is probably not going to convert you into a believer. I believe Will Carroll is interested in engaging the second group. He’d like us to teach sabermetrics to a more mainstream group of fans, who are not yet convinced that they need sabermetrics. He points out that very few fans use even the most simple sabermetrics measures. He believes that sabermetrics writers have not done a good job of making sabermetrics interesting to the average baseball fan and that we need to tell a more compelling story: Moneyball told a good story and brought some advanced measures to a wider audience. In the shadow of that book, statheads lacked a Michael Lewis to carry their message, and worse, didn’t understand why the book was popular…Statheads need to “stop making sense” and start making strides. He also uses Freakonmics as an example of a book which made a difficult subject – economics – more appealing to a wide audience. He doesn’t think that we are doing that with sabermetrics. I would argue that writers such as Joe Posnanski and Rob Neyer have certainly made strides into the mainstream and many bloggers have also contributed.. If you have spent a lot of time reading and participating in various internet forums, you would know that interest and understanding of sabermetrics is at least ten times what is was five years ago. Advanced statistics have also made it into the broadcasts of Dan Dickerson (Tigers), Vin Scully (Dodgers) and Chris Welsh (Reds) among others. Interest in sabermertrics is growing for sure, but Carroll is right that it’s not really part of the mainstream yet. He claims that “99% o f baseball fans still don’t use OPS, let alone a more advanced measure.” Once you go beyond the internet and a few curious media people, there isn’t a great deal of enthusiasm for the field among baseball fans. Tango, Mitchel Lichtman and others say that they don’t care if the audience gets bigger. They want to learn more about the game through statistics for the interest of themselves and their niche group of readers. I certainly don’t want analysts of the caliber of Tango and Lichtman spending their time trying to entertain fans with stories Personally, I don’t want us to dumb down sabermetrics to the point where they are no longer useful. I didn’t like what happened to ESPN when the went from a television station for hardcore sports fans to an entertainment network for more casual fans. That may not be a perfect analogy but I’m not interested in contributing to that kind of transformation. However, I do think that Carroll has a point. If we are going to complain about the media’s lack of understanding and awareness of sabermetrics in Hall of Fame and awards voting, then we do probably need to make more progress in making sabermetrics accessible and interesting. I think we've made strides, but there is more work to be done.
The current Tigers are done with their season but there are quite a few former Tigers making an impact on the post-season this year. Curtis Granderson leads all hitters with a .455 batting average collecting 5 hits in 11 at bats. He has a double, a triple and three RBI. Some Tigers faithful don't know whether to be happy for the former fan favorite or unhappy that he's helping the hated Yankees. Personally, I'm glad to see him having success. Another Yankee Marcus Thames is 2 for 7 with a homer and a double. Another personal favorite Carlos Pena of the Rays didn't have a big season and is trying to make a splash in the playoffs prior to free agency. Pena is 4 for 10 including a double, a triple and a homer. He also has a post-season leading 4 RBI. Cody Ross is helping the Giants going 4 for 14 with a homer, a double and 3 RBI. He isn't the only ex-Tiger in San Francisco. Ross, Omar Infante (4 hits), Aubrey Huff (4 hits), Andres Torres (2 hits) and Edgar Renteria (2 hits) have combined for 16 of the Giants 28 hits. Other former Bengals seeing action in the playoffs so far include Placido Polano and Chad Durbin of the Phillies, Matt Joyce of the Rays and Kyle Farnsworth of the Braves. Why can't the Tigers ever get guys like that?
Many Tigers fans anguished about the Tigers inability to drive home runs this year. They batted a league leading .270 with the bases empty, but hit only .256 with runners in scoring position. Batting average only explains a a portion of run scoring though. The question is could they have scored more than 751 runs if they were more efficient with their hits, extra base hits and walks? The Tigers finished 8th in the American League in runs and seventh in OPS. So, based on OPS, it doesn’t appear that they should have ranked a lot better in runs than they did. As I’ve explained before though, OPS does not do the best job of summarizing an offense. First, it does not consider baserunning at all. It also weights OBP and slugging equally when OBP contributes about 80% more to run scoring than slugging.
Based on the results of thousands of games, we know that there are more accurate ways to weight offensive events. According to linear weights theory, the average single single is worth 0.47 runs. In other words, if one single is added to a team’s hit total in each game for 100 games, that team would be expected to add 47 runs to their season total. Other events are weighted as follows:
1B 0.47 2B 0.77 3B 1.04 HR 1.40 BB 0.31 IBB 0.17 HBP 0.33 SB 0.20 CS -0.42 outs -.097 The formula for linear weights runs created (RC) looks like this: RC = (0.47 x 1B + 0.77 x 2B + 1.04 x 3B + 1.40 x HR +0.31 x BB + 0.17 x IBB + 0.33 x HBP + 0.20 x SB – 0.42 x CS - .097 x (AB-H) If you plug a team’s numbers of singles, doubles and everything else into the formula, you will arrive at an estimate of how may runs a team should have scored. Table 1 below tells us that the Tigers should have scored 756 runs with their offense. That is just five more runs than they actually scored. So, they were a little inefficient with their offense but probably not as much as some fans suggested. Table 1: Runs Versus Runs Created for AL Teams, 2010 | Team | R | RC | R-RC | % Diff | | Rays | 802 | 739 | 63 | 7.8 | | Angels | 681 | 644 | 37 | 5.5 | | Yankees | 859 | 833 | 26 | 3.0 | | Rangers | 787 | 765 | 22 | 2.8 | | White Sox | 752 | 731 | 21 | 2.8 | | Twins | 781 | 769 | 12 | 1.6 | | Athletics | 663 | 661 | 2 | 0.3 | | Blue Jays | 755 | 759 | -4 | -0.5 | | Tigers | 751 | 756 | -5 | -0.7 | | Indians | 646 | 658 | -12 | -1.8 | | Red Sox | 818 | 836 | -18 | -2.3 | | Royals | 676 | 704 | -28 | -4.1 | | Mariners | 513 | 542 | -29 | -5.6 | | Orioles | 613 | 648 | -35 | -5.7 |
Five American League teams undershot their runs created by more than the Tigers did. The least efficient team was the Orioles, who scored 35 runs fewer than expected. Part of that was baserunning. Beyond stolen bases and caught stealing, the Orioles were 7 runs below average on the bases (according to the Equivalent Baserunning stats at Baseball Prospectus). A .661 OPS with runners in scoring position probably did not help either.
The most efficient team in the league was the Rays, who scored 63 more runs than expected. They they helped themselves with an estimated 9 runs on the bases (other than SB and CS). They also hit a lot better with runners in scoring position (.790 OPS) than with the bases empty (.699). The National League results are shown in Table 2.
Table 2: Runs Versus Runs Created for NL Teams, 2010 | Team | R | RC | R-RC | % Diff | | Astros | 611 | 575 | 36 | 5.9 | | Padres | 665 | 637 | 28 | 4.2 | | Dodgers | 667 | 646 | 21 | 3.2 | | Marlins | 719 | 702 | 17 | 2.4 | | Cardinals | 736 | 719 | 17 | 2.3 | | Phillies | 772 | 761 | 11 | 1.4 | | Braves | 738 | 730 | 8 | 1.1 | | Cubs | 685 | 679 | 6 | 0.9 | | Mets | 656 | 651 | 5 | 0.8 | | Giants | 697 | 695 | 2 | 0.3 | | Rockies | 770 | 768 | 2 | 0.3 | | Reds | 790 | 788 | 2 | 0.2 | | Nationals | 655 | 661 | -6 | -0.9 | | Pirates | 587 | 593 | -6 | -1.1 | | Diamondbacks | 713 | 724 | -11 | -1.5 | | Brewers | 750 | 773 | -23 | -3.1 |
I think the lesson to be learned here was that the Tigers did not have a major problem getting the most out of their offensive output. Thus, they should not necessarily be looking to be more efficient next year. Instead, they should aim to add more hits, walks and extra base hits. If they can get players who do that, they’ll score more runs even if they don’t always get their hits “at the right time”.
The 2009 Tigers barely outscored their opponents 745 to 743. You would expect a team with such a small run differential to finish with 81 wins, but the Tigers managed to win 86 games. The reason was because they had a record of 52-33 in games decided by just one or two runs. At the time, I theorized that they won so many close games, in part, because of Fernando Rodney’s high save conversion rate (37 for 38), seven walk-off wins and some luck.
This year, the Tigers had remarkably similar runs scored and runs allowed totals to 2009. They outscored their opponents 751 to 743, but won just 81 games. Pete Palmer – author of The Hidden Game of Baseball – estimated that a differential of ten runs is worth approximately one win. In other words, adding ten runs to a team’s season total increases their expected wins (EW) by one and subtracting ten runs decreases their expected wins by one. Palmer developed the following formula for expected wins based on that principle:
EW = games/2 + run differential/10
According to this formula, you would expect the Tigers to have 82 wins. The results for other American League teams are shown in Table 1 below. You can see that every team in the league finished within 4 wins of their expected wins. So, the formula fit reality amazingly well in the American League this year. This means that winning was largely a function of the obvious – scoring more runs and allowing fewer runs. Table 1: Wins and Expected Wins in AL in 2010 | Team | R | RA | W | EW | W-EW | | Royals | 676 | 845 | 67 | 64 | 3 | | Orioles | 613 | 785 | 66 | 64 | 2 | | White Sox | 752 | 704 | 88 | 86 | 2 | | Twins | 781 | 671 | 94 | 92 | 2 | | Blue Jays | 755 | 728 | 85 | 84 | 1 | | Angels | 681 | 702 | 80 | 79 | 1 | | Red Sox | 818 | 744 | 89 | 88 | 1 | | Rays | 802 | 649 | 96 | 96 | -0 | | Tigers | 751 | 743 | 81 | 82 | -1 | | Rangers | 787 | 687 | 90 | 91 | -1 | | Indians | 646 | 752 | 69 | 70 | -1 | | Mariners | 513 | 698 | 61 | 63 | -2 | | Yankees | 859 | 693 | 95 | 98 | -3 | | Athletics | 663 | 626 | 81 | 85 | -4 |
The results for the National League are shown in Table2. The Houston Astros were the only team falling more than five wins from their expected wins. They won seven more games than expected (76 versus 69). The reason was because they lost a lot of blow outs. They were 1-13 in games with margins of eight runs or more. The Cardinals won five fewer games than expected (86 versus 91). In their case, they were 24-7 in games decided by six or more runs. | Team | R | RA | W | EW | W-EW | | Astros | 611 | 729 | 76 | 69 | 7 | | Pirates | 587 | 866 | 57 | 53 | 4 | | Phillies | 772 | 640 | 97 | 94 | 3 | | Cubs | 685 | 767 | 75 | 73 | 2 | | Dodgers | 667 | 692 | 80 | 79 | 2 | | Brewers | 750 | 804 | 77 | 76 | 1 | | Padres | 665 | 581 | 90 | 89 | 1 | | Giants | 697 | 583 | 92 | 92 | -0 | | Reds | 790 | 685 | 91 | 92 | -1 | | Braves | 738 | 629 | 91 | 92 | -1 | | Marlins | 719 | 717 | 80 | 81 | -1 | | Mets | 656 | 652 | 79 | 81 | -2 | | Nationals | 655 | 742 | 69 | 72 | -3 | | Rockies | 770 | 717 | 83 | 86 | -3 | | Diamondbacks | 713 | 836 | 65 | 69 | -4 | | Cardinals | 736 | 641 | 86 | 91 | -5 |
Getting back to the Tigers, why did they win five fewer games this year when their run differential was six runs more than it was in 2009? The reason was because they had less success in close games. After finishing 19 games above .500 in games decided by one or two runs last year, they were 33-38 this year.
It’s hard to blame their bullpen which was better than last’s year’s crew according to most statistics including ERA, FIP and WPA. They also had their share of dramatic victories early in the season. Some might attribute the difference in close game performance to poor managing or less desire to win, but I’m going guess that it was just a random thing.
The tweets are coming in fast and furious from Jason Beck today as the Tigers get an early start to their hot stove season. Here is the summary: - Johnny Damon was told that he will not be returning in 2011
- Gerald Laird will also not be returning, probably by mutual agreement.
- The Tigers will not exercise the $15 million option on Magglio Ordonez's contract. They can still try to sign him at a lower salary.
- The Tigers announced that they will probably not be re-signing Jeremy Bonderman.
- Brandon Inge has been offered a multi-year deal by the Tigers. For Dave Dombrowski to broadcast that publicly, I have to think that he'll be signing soon.
- The Tigers want Jhonny Peralta back and are talking to his agent. He should be their shortstop next year.
- Carlos Guillen will go into the season without a regular position.
- Phil Coke will join the starting rotation.
So, it appears that the Tigers infield is set for 2011 except to determine whether Scott Sizemore or Will Rhymes will be the second baseman. The starting rotation is also pretty much set unless there is competition with Armando Galarraga for the fifth spot. We can expect them to pursue, corner outfieders, DHs and relievers.
In the first big news of the Tigers "off-season", Jim Leyland announced that Phil Coke will join the starting rotation next year. The 27-year-old southpaw will get his first start for the Tigers today but that's because because none of their regular starters are available. He is not expected to pitch more than two or three innings this time. Next year, the plan is for him to be a full-time starter. Coke will be the sole lefty in the rotation behind Justin Verlander, Max Scherzer and Rick Porcello. Armando Galarraga is the leading contender for the fift spot. The move indicates two things. First, Jeremy Bonderman is not likely to be back next year which is not a surprise. It also suggests that they probably won't pursue a starting pitcher during the off-season. Instead, we can expect them to focus on hitters and relievers.
Whether or not you think second baseman Will Rhymes has a future as a major league regular, it's hard not to root for him. The diminutve infielder has made a lot of fans with his hustling play and solid contribution offensively and defensively down the stretch. One of his biggest fans is Kurt Mensching of Bless You Boys and Kurt got the opportunity to interview Will earlier this week. Among other things, you'll learn that Rhymes is a biologist and that he makes himself very accessible to internet fans. It was particularly interesting to hear Rhymes's thoughts on the competitiveness of each minor league level. He feels that there is not a big difference between Short season ball (Connecticut Tigers), level A (West Michigan WhiteCaps) and high level A (Lakeland Flying Tigers). While he thinks that the jump to double-A (Erie SeaWolves) is the first really big challenge, he does not feel as if it's as big as the jump to Triple-A (Toledo MudHens:) To me the lower levels are roughly equivalent, short season through high A. Obviously you are continually learning at each level and people can make improvements or struggle at different levels but from a competition stand point, it is very similar. AA begins to separate the men from the boys a little but I disagree with the jump to AA being the biggest. To me AAA is a whole different animal. You have a lot of veteran pitchers who really know how to pitch. You also see a lot of major league-ready arms, and the bullpens are much improved. Overall defense is better, and all of these things make it harder to hit.
Former Detroit Tigers beat writer and current Fox analyst Jon Morosi is supporting Miguel Cabrera for the MVP award over Rangers outfielder Josh Hamilton. The part of the article which piqued my interest was the list of criteria which writers are given in considering the MVP: 1. Actual value of a player to his team, that is, strength of offense and defense. 2. Number of games played. 3. General character, disposition, loyalty and effort. 4. Former winners are eligible. 5. Members of the committee may vote for more than one member of a team.
The last two criteria are just reminders and don’t have any bearing on a choice between Cabrera and Hamilton. Criterion three is interesting in that each has experienced substance abuse problems. Both have apparently conquered these issues however, and there is no evidence that either has an advantage in general character.
So, that leaves us with items one and two. Cabrera certainly has the upper hand over Hamilton in games played (146 versus 130). In fact, Hamilton has only started 114 of those 130 games. If he misses the rest of the season with injured ribs, he will have failed to start almost a third of his team’s games. That’s a big chunk and it will probably be considered fairly heavily in MVP voting.
The tricky part is criterion one – value of a player to his team. Morosi gives the following definition of value:
I believe a player’s “value” is best defined as how difficult he would be for the team to replace.
I don’t disagree with this definition. It’s essentially the same criteria used by many modern statistical analysts. It fits nicely with the Wins Above Replacement (WAR) concept. Where I disagree is how he determines value. He reasons that the Rangers would still have a solid line-up without Hamilton, but that the Tigers would be lost without Cabrera:
Take away Hamilton, and the Rangers still have a very good lineup.
Take away Cabrera, and Ryan Raburn is the potential cleanup man. He correctly states that Cabrera has a larger percentage of his team’s home runs and RBI than Hamilton. There is no question that Cabrera is the centerpiece of the Tigers line-up and that Hamilton shares the load with the likes of Vladimir Guerrero, Nelson Cruz and Michael Young. That’s a classic interpretation of what “valuable” means in MVP. It has some merit, but I’m most interested in the number of runs Cabrera and Hamilton contribute to their teams.
One way to estimate run contribution is with the Batting Runs statistic which I discussed earlier this week. I’ll use the following weights for events:
NIBB 0.33 IBB 0.18 HBP 0.33 1B 0.47 2B 0.77 3B 1.04 HR 1.40 Out –0.27
Hamilton leads Cabrera 56.2 to 54.8 in Batting Runs. If we attempt to adjust for home parks (Hamilton plays his home games in a more friendly hitters park than Cabrera), they are essential even: Hamilton 54.4 and Cabrera 54.3.
Now, let’s add equivalent baserunning runs from from Baseball Prospectus. Hamilton has contributed an estimated 3.3 runs with his base running while Cabrera has cost the Tigers an estimated 1.3. The tally is now Hamilton 57.7 and Cabrera 53.0. That is how many offensive runs the two batters have contributed above what you would have been expected from an average batter. So, according to these statistics, Hamilton has had slightly more value while he has been in the line-up than Cabrera has had.
FanGraphs estimates that Hamilton has cost his team 2.4 runs versus Cabrera by not being in the line-up (18.6 replacement value 21.0) So, the gap closes to 2.3 runs. It’s really very close offensively and you could make an argument for either player on that basis.
Where Hamilton has a bigger advantage is in run prevention. FanGraphs estimates that Hamilton has a 6 run advantage simply by playing outfield instead of first base. That sounds fair enough to me. What about quality of defense? If we take the average of Total Zone, Defensive Runs Saved and Ultimate Zone Rating, Hamilton is an estimated +3 runs defensively and Cabrera is –5. That all adds up to a 14 run difference on defense which is a lot. Even if you don’t trust the defensive statistics at all and want to assume that all players are average defensively, it’s still a six run edge for Hamilton.
Is there any way for Cabrera to close the gap? What about situational hitting? Win Probability Added (WPA) is a statistic which takes into account a player’s performance in various scenarios. For example a home run to break a tie in the ninth inning would be worth more than a home run in a blow out. Cabrera has a 6.5 WPA versus 5.9 for Hamilton. So, Cabrera appears to have done a little better in clutch situations and that may be enough to swing the offense slightly in his favor. I don’t know if it’s enough to make up for the fielding gap though.
We’ll get back to the one area where Cabrera has the clear lead – playing time. We kind of incorporated that into our calculations already but it’s still hard to justify giving an MVP to a player with 114 starts. I think Hamilton has been the better player overall and I’m leaning that way at the moment, but if he doesn’t play another game and Cabrera finishes strong, I think you can make a decent case for Cabrera.
Now, I don’t want to give the impression that Hamilton and Cabrera are the only candidates. Jose Bautista, Robinson Cano, Adrian Beltre and Evan Longoria are in the mix as well. Based on what I’ve read though, it seems that voters are giving Cabrera and Hamilton the most consideration.
In a recent FanGraphs article, Dave Cameron discussed how Miguel Cabrera has not been as valuable as some of his numbers might indicate. His reasoning was that Cabrera has been walked intentionally 30 times and intentional walks are not as valuable as non-intentional walks:
Intentional walks are issued in situations where the opposing team believes it is more valuable to have the batter on first base than at the plate. It is a strategic move, based on the situation at hand, that is aimed at reducing the offense’s chance of scoring a run, or multiple runs, in a given inning.
From the examination of thousands of games , it has been determined that the average non-intentional walk (NIBB) contributes about 0.33 runs. In other words, if one NIBB is added to a team’s total in each game for 100 games, that team would be expected to add 33 runs to their season total. An intentional walk (IBB), on the other hand contributes about 0.18 runs on average (according to The Book by Tom Tango, Michel Lichtman and Andrew Dolphin). Other events have the following approximate values (called linear weights):
HBP 0.33 1B .47 2B .77 3B 1.04 HR 1.40 Out –.27
These linear weights can be inserted into a formula to calculate Batting Runs (BR):
BR = 0.33 x NIBB + 0.18 x IBB + 0.33 x HBP + 0.47 x 1B + 0.77 x 2B + 1.04 x 3B + 1.40 x HR – 0.27 X Outs
Sometimes, other events such as errors, stolen bases and caught stealing are included in the formula, but those are not needed for this particular discussion. Cabrera had the following numbers as of yesterday:
NIBB 55 IBB 30 HBP 3 1B 90 2B 45 3B 1 HR 34 Outs 342
According to the above formula, Cabrera had 57.8 BR heading into today’s action. This tells us that he has contributed 57.8 runs above what an average batter would have been expected to contribute given the same number of outs. Suppose, we did not distinguish between NIBB’s and IBBS and credited Cabrera 0.33 runs for all of his walks. In that case, we would add 30 x .15 = 4.5 runs giving him a total of 62.3 BR.
Whether or not we distinguish between different types of walks is important in comparing his batting to Josh Hamilton, who is considered his main competition for the MVP award. If we give batters less credit for intentional walks, then Cabrera leads Hamilton (38 NIBB, 5 IBB) in BR 57.8 to 56.2. If we consider all walks to be worth 0.33 runs, then Cabrera leads 62.3 to 56.9. We wouldn’t stop there in determining their value of course. We would look at baserunning contribution, park effects, position and defensive contribution. That’s for another post though.
The question is should we distinguish between type of walk in the calculation of batting runs? Tigers fans have protested that Miguel Cabrera is being unfairly punished just because he has had less protection in the batting order than Hamilton. They also reason that he gets walked a lot because pitchers fear him and would rather not face him. This indicates something positive about Cabrera’s hitting skill, not something negative.
The purpose of Batting Runs though is not to punish or reward a player. It’s also not meant to determine the best hitter. It is supposed to determine how much value in terms of runs a player contributes to his team. Thus, when a pitcher walks a batter to reduce the potential of scoring a run, it makes sense that the value of that walk might be reduced accordingly.
This practice is not universally accepted though. Tango and FanGraphs distinguish between walks but Pete Palmer (the creator of the linear weights system) and Gary Gillette do not distinguish between NIBBs and IBBs in The ESPN Baseball Encylopedia. Baseball-Reference follows the lead of Palmer on that issue as well.
A reason for not cutting the value of an IBB might be that many NIBBs are essentially IBBs. There are many instances where batters are pitched around and receive a walk without it being an intentional walk. This is especially true when considering hitters of the caliber of Hamilton and Cabrera. So, do we cut the value of all their walks in high leverage situations because we know pitchers are usually not giving them anything to hit?
I don’t think the solution to the intentional walk problem is cut and dried. I typically look at FanGraphs Batting Runs (wRAA) before I go to Baseball-Reference but in the case of batters like Cabrera or Albert Pujols, who receive a lot of intentional walks, I think it’s important to look at it both ways. There is never one number that gives you the final answer in statistical analysis.
Ryan Raburn’s 2010 season is looking very familiar. Last year, the Tigers outfielder finished strong down the stretch and ended up with an impressive .892 OPS in 291 plate appearances. He batted only .253 with a .778 OPS in spotty playing time through the first four months. Most of his damage was done in August and September when he batted .342/.403/.640 and was one of the Tigers’ key hitters in their central division title chase. It was felt by some observers that Ryan’s strong finish earned him more regular playing time for 2010, but he found himself back on the bench for much of the first four months. He did not play well in his utility role batting just .211 with a .329 slugging average in 182 PA through July. Some embarrassing mishaps in the outfield did not help his cause either. When Magglio Ordonez went down with an ankle injury in late July, Raburn got an opportunity for more playing time in left field with rookie Brennan Boesch switching to right field. Raburn has responded by batting .340/.389/.647 in 172 PA since August 1. The only American Leaguers with a higher slugging average during this period have been Blue Jays surprise slugger Jose Bautista (.664) and White Sox first baseman Paul Konerko (.656). Raburn’s late season surge leaves some Tigers fans wondering once again whether be deserves a shot at a starting job. His .837 OPS over the last two seasons ranks him third among American League outfielders with 500 or more PA. For his career, he has posted a .857 OPS versus left-handed pitchers and a .753 OPS versus right-handers. So, he probably wouldn’t do quite as well in a full-time role. There are concerns about his defense. He looks very awkward at times, but I think he makes up for some of that with a strong arm and decent range. His numbers don’t support the contention that he is an awful outfielder. In four partial seasons, he has a combined –3.5 Ultimate Zone Rating and +2 Fielding Bible Defensive Runs Saved. He’s certainly not a graceful outfielder but he seems adequate enough if he hits and may be a better defender with regular playing time at one position. Ultimately, I think the 29-year-old Raburn will be a utility man again, backing up all three outfield positions and second base. Another season of 300 PA seems like a good bet, but a starting job is not out of the question at this point. With only center fielder Austin Jackson locked into a role for next season, the opportunity could be there. If Raburn gets out of the gate fast for a change, he could seize a regular job.
One of my favorite news items of the off-season used the be the announcement of next season’s schedule. If I remember correctly, this would typically occur in the middle of December. Seeing next season’s schedule on a frigid winter day was enough to warm a fan up with thoughts of summer. For the last few years though, the schedule has been coming out in August or September. Yesterday, the Tigers released their 2011 Schedule. While it doesn’t have quite the same effect as a December release, it’s still fun to anticipate 2011 given that the Tigers are out of contention in 2010. The first thing I noticed is that the season ends on Wednesday, September 28 versus the Indians. This is unusual because as long as I can remember every regular season has ended on a Sunday. The likely reason is an effort to start and end the playoffs a little earlier. This year, the World Series may extend into November. The Tigers season opens on Thursday, March 31 in Yankee Stadium, which means that fans will probably be in winter coats, hats and gloves for the opener. The Thursday start to the season is also a bit unusual. In recent years, the season opened with a game or two on Sunday and a full slate of games on Monday. The first Comerica Park game is versus the Royals on April 8. Other Scheduling Thoughts The Tigers have interleague games at Pittsburgh, Colorado and Los Angeles next year. They play New York, Arizona and San Francisco at home. I’m not a fan of interleague play, but the Tigers generally have so much success versus the National League that it’s hard to complain too much. I don’t like west coast games very much because they end too late causing me to stay up beyond 1:00 AM. In a good year, they are limited to one or two trips to Oakland, Seattle and Los Angeles. This year, they’ve got four of them spread throughout the year – seven games in Oakland and Seattle in April, a series in Los Angeles versus the Dodgers in June, another in Los Angeles against the Angels in July (including July 4th which I hope will be an afternoon game) and one series at Oakland in September. If I go to see the Tigers at Fenway this year, my choice will be limited to two games – a Wednesday or Thursday in the middle of May. A night game on May 18 in Boston is almost certainly going to involve a chilling sea breeze. They have another dreadful four day all-star break this year from July 11-14. They are off again on July 18. Their toughest stretch of the year may be 15 games versus the Rays, Red Sox, Twins, White Sox and Rangers in late May and early June. On the other hand, they finish the season with eight games versus the Royals, Orioles and Indians.
In an earlier post, I discussed the pros and cons of two commonly used Wins Above Replacement statistics for pitchers: I then took the average of those two WARs for each Tigers starter. In a follow-up post, I did the same calculation for American Leaguers starters and found Seattle left-hander Felix Hernandez to be the top pitcher. Today, I’ll look at National League starters The fWAR statistic is based on FIP and innings pitched. As such, it favors pitchers with a high number of innings pitched and strong peripherals – high strikeout rates and low walk and home run rates. It is forgiving to pitchers with high ERAs relative to their peripherals. Table 1 shows the National League fWAR leaders in 2010. Roy Halladay of the Phillies and Josh Johnson of the Marlins are tied for the league lead with 6.2 fWAR. Rockies fire baller Ubaldo Jimenez (5.8) and Cardinals right-hander Adam Wainwright (5.4) are also more than five wins above replacement.
Table1: NL fWAR Leaders | Pitcher | fWAR | | Halladay, Phi | 6.2 | | Johnson, Fla | 6.2 | | Jimenez, Col | 5.8 | | Wainwright, Stl | 5.4 | | Lincecum, SF | 4.2 | The rWAR statistic is based on innings pitched and runs allowed with adjustments made for team defense (as measured by Total Zone) behind a pitcher. It is friendly to a pitcher with a high number of innings and a low RA - which is the same as ERA except it considers all runs rather than just earned runs. A pitcher with a low RA, despite weak defensive support, will do particularly well on this statistic. Rally’s WAR is unaffected by peripherals. We can see in Table 2 that Halladay sits atop rWAR leader board at 6.7. He is followed by Johnson (6.3), Jimenez (5.9) and Braves comeback pitcher Tim Hudson (5.7).
Table 2: NL rWAR Leaders | Pitcher | rWAR | | Halladay, Phi | 6.7 | | Johnson, Fla | 6.3 | | Jimenez, Col | 5.9 | | Hudson, Atl | 5.7 | | Wainwright, Stl | 5.5 | Fans who don’t want to choose between runs allowed and peripherals ( I fall in this category) might prefer a combined WAR. One way to do this is to compute a simple average of fWAR and rWAR as displayed in Table 3. Using this approach, Halladay (6.4 Avg WAR) is the top pitcher, followed by Johnson (6.2), Jimenez (5.8) and Wainwright (5.4).
Another thing to note is that some pitchers do much better on one stat than the other. For example, the Giants’ Tim Lincecum is fifth with 4.2 fWAR but has only 2.7 rWAR. This is because he has strong peripherals relative to his RA. Other pitchers such as the Hudson (3.4 vs. 5.7) do better on rWAR. Table 3: NL Average WAR Leaders | Pitcher | fWAR | rWAR | Avg WAR | | Halladay, Phi | 6.2 | 6.7 | 6.4 | | Johnson, Fla | 6.2 | 6.3 | 6.2 | | Jimenez, Col | 5.8 | 5.9 | 5.8 | | Wainwright, Stl | 5.4 | 5.5 | 5.4 | | Hudson, Atl | 3.4 | 5.7 | 4.6 | | Myers, Hou | 4.1 | 4.7 | 4.4 | | Santana, NY | 3.7 | 4.6 | 4.2 | | Hamels, Phi | 3.5 | 4.5 | 4.0 | | Latos, SD | 3.7 | 4.3 | 4.0 | | Kershaw, LA | 4.0 | 3.9 | 4.0 |
In my previous post, I discussed the pros and cons of two commonly used Wins Above Replacement statistics for pitchers: The fWAR statistic is based on FIP and innings pitched. As such, it favors pitchers with a high number of innings pitched and strong peripherals – high strikeout rates and low walk and home run rates. It is forgiving to pitchers with high ERAs relative to their peripherals. Table 1 shows the American League WAR leaders in 2010. Those who favor this FIP based WAR are probably choosing between Francisco Liriano (6.1 fWAR), Cliff Lee ( 6.0) and Felix Hernandez (5.9) in a close Cy Young race.
Table1: AL fWAR Leaders | Pitcher | fWAR | | Liriano, Min | 6.1 | | Lee, Sea-Tex | 6.0 | | Hernandez, Sea | 5.9 | | Weaver, LA | 5.1 | | Lester, Bos | 5.0 | The rWAR statistic is based on innings pitched and runs allowed with adjustments made for team defense (as measured by Total Zone) behind a pitcher. It is friendly to a pitcher with a high number of innings and a low RA - which is the same as ERA except it considers all runs rather than just earned runs. A pitcher with a low RA, despite weak defensive support, will do particularly well on this statistic. Rally’s WAR is unaffected by peripherals. As seen in Table 2, supporters of rWAR likely see Hernandez as the clear Cy Young favorite.
Table 2: AL rWAR Leaders | Pitcher | rWAR | | Hernandez, Sea | 5.6 | | Weaver, LA | 5.1 | | Liriano, Min | 4.7 | | Pavano, Min | 4.6 | | Price, TB | 4.6 | Fans who don’t want to choose between runs allowed and peripherals ( I fall in this category) might prefer a combined WAR. One way to do this is to compute a simple average of fWAR and rWAR as displayed in Table 3. Using this approach, Hernandez (5.8 Avg WAR) looks like the top pitcher, followed by Liriano (5.4), Weaver (5.1) and Lee (4.8).
Another thing you may notice is that some pitchers do much better on one stat than the other. For example, Liriano (6.1 fWAR versus 4.7 rWAR) and Lee (6.0 vs. 3.5) do far better on fWAR because of their strong peripherals and good but not great RAs. On the other hand, hurlers such as David Price (3.7 vs. 4.6) and Carl Pavano (3.6 vs. 4.6) do better on rWAR. Table 3: AL Average WAR Leaders | Pitcher | fWAR | rWAR | Avg WAR | | Hernandez, Sea | 5.9 | 5.6 | 5.8 | | Liriano, Min | 6.1 | 4.7 | 5.4 | | Weaver, LA | 5.1 | 5.1 | 5.1 | | Lee, Sea-Tex | 6.0 | 3.5 | 4.8 | | Lester, Bos | 5.0 | 4.3 | 4.6 | | Price, TB | 3.7 | 4.6 | 4.2 | | Wilson, Tex | 3.9 | 4.4 | 4.2 | | Danks, Chi | 4.1 | 4.2 | 4.2 | | Pavano, Min | 3.6 | 4.6 | 4.1 | | Sabathia, NY | 4.0 | 4.0 | 4.0 | | Verlander, Det | 4.5 | 3.4 | 4.0 |
You may have noticed that there are two commonly used Wins Above Replacementnt (WAR) statistics for both hitters and pitchers:
Today. I’m going to discuss the pitching version of each. There are pros and cons to both measures and the statistical community is still debating the merits of the two. fWAR is based on FIP and therefore only considers events which a pitcher essentially controls – K, BB, HR, IP. It also adjusts for a pitchers home park.
The fWAR statistic has some potential drawbacks. First, it considers BABIP to be completely out of control of the pitcher. It also gives a pitcher no credit for distribution of base runners or sequencing events. For example, a pitcher that pitched well with runners on base would not get credit for that.
The rWAR statistic was invented by Sean Smith and is found on Baseball-Reference. It starts with total runs allowed by a pitcher and then adjusts for team fielding behind that pitcher. It considers IP run average (RA) - which is the same as ERA except it considers all runs rather than earned runs - and the Total Zone statistic. It also adjusts for the pitcher’s home park. The advantage is that it tries to tease out defense rather than completely ignore balls in play.
A potential shortcoming of rWAR is that the measurement of team defense in a single season is still shaky. It’s also possible that rWAR gives a pitcher too much credit for limiting hits on balls in play, distribution of base runners and sequencing.
So, which one is better? I prefer the concept of rWAR better. I think it’s good to start any kind of run prevention evaluation with runs allowed. The next step would be to determine how much of run prevention is pitching and how much is fielding. rWAR attempts to do that but the measurement of team defense in a single season is still too tenuous for me to use rWAR by itself.
Over the course of a career or several seasons, rWAR becomes more reliable because variation in team defense evens out and we learn more about a pitchers ability to control BABIP, base runner distribution and sequencing over time.
fWAR is valuable because it tells us how good a pitcher was at events over which he has the most responsibility. Thus, it is better statistic than rWAR for identifying pitching talent and projecting into the future. However, it makes too many assumptions about events which a pitcher does not control by himself for me to use it as a stand alone measure.
So rWAR is best for career measurement and fWAR is good for projection. How about evaluating which pitcher did best last season? I would not use either by itself, but rather I would make a judgment using both. One way to do that would be to take an average or weighted average of the two statistics. For now, I’ll look at the Tigers starters in 2010 using a straight average (Table 1).
Table 1: WAR for Tigers Starters in 2010
| Pitcher | fWAR | rWAR | AverageWAR | | Verlander | 4.5 | 3.4 | 4.0 | | Scherzer | 3.3 | 2.9 | 3.1 | | Galarraga | 0.9 | 1.8 | 1.4 | | Porcello | 1.7 | -0.1 | 0.8 | | Bonderman | 1.5 | -0.1 | 0.7 |
We can see that Justin Verlander was 4.5 wins above a replacement player according to Fan Graphs and 3.4 using Baseball-Reference. That averages out to 4.0 WAR. Based on that, we would say that he has added an estimated four wins to the Tigers beyond what an average player would contribute. Max Scherzer is next in line with an average WAR of 3.1.
All of the Tigers did better on fWAR than rWAR except Armando Galarraga. Galarraga does poorly on f WAR because he has not done a good job on things that a pitcher most controls (fWAR). He does better on rWAR because he has a low BABIP and relatively low ERA.
In future posts, I am going to look at average WAR for pitchers on other teams.
Last year, the voters for the Cy Young award selected 16-game winner Zach Greinke over 19-game winners Felix Hernandez, CC Sabathia and Justin Verlander. Based on ERA and sabermetric statistics such as Wins Above Replacement (WAR), they clearly made the right choice. In the the National League, they selected 15-game-winner Tim Lincecum instead of 19-game winner Adam Wainwright. There would have been no shame in giving the award to Wainwright, but again I think the made the correct selection. Those were two of the three lowest win totals for starting pitchers in the history of the award (Brandon Webb also won 16 games in 2006).
Traditionally, the writers who vote for the award have made games won a major criteria. In 1990, for example Clemens had 21 wins and led the American League with a 1.93 ERA, but lost out to 27-game winner Bob Welch, who had an ERA a full run higher (2.95). That appears to be changing as they seem to be recognizing more than ever that wins is not the fairest way to evaluate pitchers. This year will be particularly interesting though as they may have to choose between a potential 20-game winner and pitcher with fewer than 15 wins.
As it stands now, Yankees left-hander CC Sabathia has a 19-6 record. Felix Hernandez of the Mariners stands at 11-10. Readers of this blog don’t need me to explain the failings of won-loss record for pitchers, but this case is so extreme it’s worth mentioning the run support argument. While Sabathia has benefitted from 6.07 runs per game, King Felix has only received 3.16 runs per game. That’s almost three more runs per game for Sabathia! That’s hardly a fair comparison
Sometimes, pitchers accumulate a lot of wins because they are workhorses which pitch deep into games. One might guess that a 19-game winner would be pitching deeper into games than an 11-game winner, but, in this case, Hernandez actually has more innings pitched (219 1/3 versus 209). So, the wins statistic is not going to work here at all.
Since the main job of a pitcher is to give up as few runs as possible, a good starting point in pitcher evaluation is ERA. Table 1 shows that Hernandez leads Sabathia in ERA by a wide margin (2.30 versus 3.14). The ERA leader is Clay Buchholz of the Red Sox at 2.25. However, Buchholz has pitched only 151 2/3 innings, so he has not helped his team as much as Hernandez.
Table 1: AL ERA Leaders
| Pitcher | IP | ERA | | Buchholz, Bos | 151 2/3 | 2.25 | | Hernandez, Sea | 219 1/3 | 2.30 | | Cahill, Oak | 165 2/3 | 2.72 | | Price, TB | 178 2/3 | 2.87 | | Wilson, Tex | 177 | 3.10 | | Weaver, LA | 189 | 3.14 | | Sabathia, NY | 209 | 3.14 |
In order to give pitchers credit for quantity of innings pitched as well as quality, Pete Palmer introduced the Pitching Runs statistic in 1984. Pitching Runs tells us the number of runs saved by a pitcher compared to league average. In it’s purest form, it is based on a pitcher’s IP and earned runs (ER) and the league ERA (Lg ERA):
Pitching Runs = IP x Lg ERA/9 – ER.
The Baseball-Reference version also adjusts for ballpark. Hernandez has 39 Pitching Runs (See Table 2) which means that he has saved his team 39 runs compared to what the average pitcher would have saved in his place. Because his greater workload is taken into account, Hernandez leads Buchholz by 7.7 runs. Sabathia is only 10th in the league with 19.5.
Table 2: AL Pitching Runs Leaders
| Pitcher | IP | Pitching Runs | | Hernandez, Sea | 219 1/3 | 39.0 | | Buchholz, Bos | 151 | 31.3 | | Price, TB | 178 2/3 | 24.6 | | Cahill, Oak | 165 2/3 | 23.9 | | Wilson, Tex | 177 | 22.2 | | Weaver, LA | 189 | 21.3 | | Lester, Bos | 182 | 20.9 | | Liriano, Min | 172 1/3 | 19.9 | | Gonzalez, Oak | 179 2/3 | 19.7 | | Sabathia, NY | 209 | 19.5 |
A shortcoming of both ERA and Pitching Runs is that the don’t consider defensive support behind a pitcher. The WAR statistic developed by Sean Smith and listed at Baseball-Reference attempts to take fielding into account. WAR is calculated as follows: -
Determine how many runs a pitcher allowed.
-
Calculate average runs allowed by pitchers facing the same teams for the same number of innings as the pitcher of interest.
-
Adjust for team defensive runs saved based on Total Zone. Total Zone is computed from plays made, errors, which fielders fielded each out and hit, batted ball type, handedness of pitcher and batter, and park adjustments.
-
Multiply by 1.22 to get replacement level for an AL starter.
-
Subtract (4) from (2) to get WAR.
Table 3 shows that Felix Hernandez was the clear leader in WAR with 5.7. This says that Hernandez contributed close to six wins more than you would expect from a replacement level player. Sabathia is 10th in the league with 4 WAR.
Table 3: AL Starting Pitcher WAR Leaders
| Pitcher | WAR | | Hernandez, Sea | 5.7 | | Weaver, LA | 4.8 | | Buchholz, Bos | 4.8 | | Liriano, Min | 4.7 | | Wilson, Tex | 4.6 | | Price, TB | 4.6 | | Pavano, Min | 4.4 | | Lester, Bos | 4.3 | | Danks, Chi | 4.2 | | Sabathia, NY | 4.0 |
Based on ERA, Pitching Runs and WAR and every other statistic other than wins, Hernandez appears to be the superior pitcher to Sabathia this year. It will be very interesting to see how the vote turns out if Sabathia wins 20+ and Hernandez wins only 11 to 13 games.
Edit: St. Pierre did not actually score the winning run. Brennan Boesch pinch ran for him and he scored the run.
I’ve been to a lot of minor league parks over the years. I have followed Tigers prospects in Lowell, MA, Manchester, NH, Pawtucket, RI, Oneonta, NY and Lakeland, FL among other places. I’ve even attended some Arizona Fall League games. I typically go specifically to watch the top prospects like Justin Verlander, Rick Porcello, Curtis Granderson and Cameron Maybin. I could not help but notice though that, for the longest time, I’ve been penciling Maxim St. Pierre’s name into box scores.
There is a reason it seems that Max has been around forever. He was drafted by the Tigers in the 26th round of the First Year Player draft in June, 1997. The Montreal native has spent the last14 years in the minors – 13 in the Tigers system – without ever playing a game in the majors. He has made stops with the Gulf Coast League Tigers, Oneonta, West Michigan, Lakeland, Erie, Toledo and Huntsville (one year with the Brewers). How long is 14 years? When he was drafted in 1997, Raul Casanova was the Tigers regular catcher that year.
Earlier in the week, St. Pierre finally realized his dream when the Tigers recalled him for the final month of the season. With Laird battling a sore back, the Tigers felt they could use a third catcher as rosters expanded. The 30-year-old St. Pierre has had a good season batting a combined .274/.353/.479 for Erie and Toledo and is a strong defender. Just as important, the out-of-contention Tigers are rewarding him for his many years of service.
Today, Max played in his first major league game. The pitcher was Rick Porcello, who was eight years old when St. Pierre signed with the Tigers organization. St. Pierre grounded out to shortstop in his first major league at bat. He failed to hit in his first three at bats, but got one more chance in the eighth with the score tied at four. The rookie catcher responded by poking a single to center field. He eventually came around to score what proved to be the winning run in a 6-4 win over the Royals.
The Tigers have had a lot of rookies get their first major league hits this year – Austin Jackson, Brennan Boesch and Casper Wells to name a few. But St. Pierre’s first hit was more special than any of them.
There has been much discussion on the internet about Jhonny Peralta’s potential role on the Tigers in 2011. He has a club option that would pay him $7 million next year. If they want him back, they can either allow the option to kick in or decline the option and try to sign him at a lower salary. The other question is whether he would play third base or shortstop.
With Brandon Inge’s contract coming to an end after this season, the Tigers could replace him with Peralta. According to the Fan Graphs weighted runs created statistic (wRC or RC), Peralta has 57.7 RC in 506 Plate appearances (PA) this year. Over 650 PA, that would be 74 RC. Similarly, Inge’s 52.5 RC over 466 PA translates into 73 RC over a full season. So, there is only a one run difference between the two. Performing similar calculations on 2009 data yields 65 RC for Peralta and 70 RC for Inge. So, based on the last two seasons, Peralta would not be an upgrade over Inge offensively.
Defensively, I’ll look at the Ultimate Zone Rating (UZR) and Defensive Runs Saved (DRS) over 2009-2010 because it’s always a good idea to use multiple fielding measures and multiple years when evaluating defense. Peralta had the following numbers at in 2009-2010:
DRS –4 +2 UZR –2 –4
Averaging those together yields an estimate of 0 defensive runs saved. Now for Inge:
DRS +8 -1 UZR +7 +3
That averages out to +4 runs saved. So, I’ll estimate that Inge is 4 runs better than Peralta defensively. Peralta is only one run better offensively this year, so he’s clearly not an upgrade overall. They are probably better off with Inge at third for next year, but it’s not a big difference.
What about Peralta at shortstop. Would you rather have Peralta or Ramon Santiago/ Danny Worth? Some of you might be thinking “none of the above”, but let’s look at those two options. The table below looks at the runs created numbers for the three shortstops in 2010.
| Player | PA | RC | | Peralta | 506 | 57.7 | | Santiago | 334 | 34.6 | | Worth | 115 | 9.7 |
We saw above that Peralta would have 74 RC over 650 PA. Suppose, we give 400 PA to Santiago and 250 to Worth. Their estimated RC are in the following Table
| Player | PA | RC | | Peralta | 650 | 74 | | Santiago | 400 | 42 | | Worth | 250 | 21 |
Over a combined 650 PA, Santiago and Worth would have 63 RC. That is 11 runs fewer than Peralta. Do Santiago and Worth make up for it defensively? We are dealing with a lot of small sample sizes here, so I’ll go back five years. First, Peralta:
DRS –6 –6 –2 –6 +5 UZR –6 –11 –12 –1 +3
That averages to –4 runs which Peralta cost his team. Now, Santiago:
DRS +1 +8 –6 –2 +9 UZR +2 +3 0 –1 +6
That comes out to +2 runs saved on average. The data on Worth is too limited, so I’ll assume that he is average defensively, which based on observation and reputation seems fair. So, let’s say that Santiago/Worth are +1 defensively. That makes them five runs better than Peralta. Combine that with the offense and Peralta is six runs better.
My conclusion is that the 2009-2010 version of Peralta would not be an upgrade at third base, but would be a slight upgrade at shortstop assuming his defense does not fall off a cliff. Now, if Peralta can get back to his 90 RC production of 2008, then he would be a big upgrade at either position. My original thought was that he would be ok at third but not at shortstop. I'm still leery of the Tigers using a below average defender at shortstop and I hope they look elsewhere. However, it looks like shortstop might actually be a better idea than third base.
|
|
|
