Archive for the 'Splits' Category

Some tidbits about the Phillies and Dodgers pitching:

First, check out this table for relief pitching in 2009 (regular season):

Dodgers relievers appeared in the 4th most games this year, which is pretty odd for a playoff team. The Phillies' relievers made 67 fewer appearances than the Dodgers' and all the other playoff teams are well below the Dodgers as well. (Rockies are 12th, Cardinals are 14th, Twins are 16th, Red Sox 18th, Yankees 19th, and Angels 27th.)

When I think about how-off playoff teams are, I always look at the bullpen status. These guys get tired and can often be less effective in the playoffs than they were in the regular season.

Now here are the save stats for the 2009 regular season:

Much has been made of Brad Lidge's poor performance (and Ryan Madson's bad stint as closer in Lidge's absence) but the Phillies ended up about average in save conversion percentage. It was the Dodgers' 26 blown saves that was 3rd most in baseball. Again notice that the 6 other playoff teams are all above average. (This isn't too surprising...if a team blows a lot of saves, it's hard to get a lot of wins and make the playoffs...therefore playoff teams should, on average, have fewer blown saves.)

I also wanted to point out something that happened in the Phillies-Rockies NLDS that shows why the National League is so much more interesting.

In the bottom of the 7th inning of Game 4 Raul Ibanez made a huge error, dropping a fly ball (lost in the lights) and allowing Seth Smith to reach second base. At the time, the Phillies were up by 1 run with one out in the inning.

Why would I label this error as huge? The Rockies ended up not scoring in the inning and the error seemed to have no effect on the game. But...it did. After Clint Barmes flied out for the second out, pitcher Ubaldo Jimenez's spot came up in the lineup. With the tying run on second base, Jim Tracy had no option but to pinch hit for Jimenez. Ryan Spilborghs made out, but nobody would argue with Tracy's decision.

Up to that point, Jimenez had pitched very well, allowing just 6 hits and 2 walks in 7 innings while striking out 7. The runs came on two gopher balls.

In the 8th inning, the Phillies nearly scored off the Rockies' bullpen. Perhaps if Jimenez had still been pitching, he would have breezed through another inning, saving the Rockies' relief pitchers for the 9th. Instead, they used up guys because they had to.

Weirdly, Ibanez's two-base error actually seems to have helped the Phillies win the game! Isn't it amazing how an error by one team's fielder can get the other team's pitcher out of the game? This is the beauty of the National League and makes NL games so much more interesting than AL games.

All of this is a long-winded way of saying two things:

• In the NLCS, keep your eye on when each manager decides to lift the starting pitcher, especially if you are a relatively new fan of baseball. It's one of the most fascinating things about the National League playoffs.
• The designated hitter concept really stinks, in my opinion. We've had some great DHs over the years, but personally I feel it's time to get rid of it. The MLB players union does not want to allow this as it would cost roster spots for players who fit only as DHs, such as David Ortiz, Hideki Matsui, Jim Thome, etc. In my opinion, the DH should be eliminated, effective in 2014. This would give teams 5 years to change their rosters, etc, to compensate for the coming change.

Today's MLB Oriole's Game Notes has a paragraph chronicling all the doubles lists that Brian Roberts is climbing this season. Here is a closer look courtesy of the Baseball-Reference.com Play Index:

Read the rest of this entry »

Check out Raul Ibanez's splits before and after the All-Star break this year:

So the guy's power and productivity have been largely diminished in the second half. It's also interesting to look at his career splits before and after the break:

It's interesting just how similar his numbers are before and after, especially his OPS being identical. His lopsided performance this season actually brought those numbers in line to each other for his career.

Check out Joe Blanton's numbers so far this season. Not including his nice start tonight, he's 9-7 in 27 starts with a 4.11 ERA. But against the Nationals (again, before tonight) he was 0-2 in 4 starts with a 9.45 ERA over 20.0 innings. By doing a little math, that means he was 9-5 with a 3.40 ERA against everybody else.

I guess things really do average out over time, though, as Blanton tonight pitched 6 scoreless ininngs against the Nats and picked up the win.

Check out the worst slugging percentage by 4th-place hitters for each team (stats through last Thursday...sorry...wrote this one a little early.)

Man, check out just how awful the Royals' cleanup hitters have been. They are nearly 100 points below the second-worst team, and the entire range of everybody else is about 200 points. It's not like their producing in any other way, either. The Royals' #4 batters have a collective batting average of .211 and OBP of .278. Sheesh. The cumulative OBP for all of MLB is .333, meaning that the guys on the Royals who should be in the upper echelon of power hitters have an OBP that's more than 50 points below leave average. And they have struck out a ton. Umm...yeah.

Let's not miss the Mets. Although they have an average SLG out of their cleanup hitters, they have just 14 HR, the second-fewest in baseball.

Here are the performances by the guys who have hit 4th the most for the Mets.

The loss of Delgado hurts for sure. All of the other guys have hit pretty well, truth be told, and 97 RBI over the course of the season is not too bad. For some reason, though, the dingers have been infrequent.

If you check out the major-league batting splits for 2009 to date, you can find the batting splits based on leverage.

For a brief description of leverage, you can check out the glossary here. Basically, each plate appearance can be assigned an index that indicates how big of an influence it has on the outcome of the game.

I've reduced the table to some stats of interest:

First, you can see that medium-leverage and low-leverage plate appearances have about the same frequency, whereas high-leverage plate appearances are about half as frequent. This tells you that in an average game, only about 1 out of every 7 plate appearances can have a large impact on the game's outcome. Interesting, huh? (That's not to suggest that runs scored during low- or medium-leverage situations don't matter--they just have less average effect on the outcome. Many of these are, for example, solo homers late in a game where the run differential is already 5.)

Calculating RBI per plate appearance, high-leverage situations come in at 0.22, medium at 0.10, and low at 0.08.

These numbers might be confusing. Some people interpret them to mean that players focus more during high-impact situations and drive in more runs. (Another way to refer to this is "clutch hitting.") In reality, the reason there are more RBI in high-leverage situations is that there are more runners on base. In other words, a situation with 2 runners on base is more likely to be high-leverage, whereas a situation with empty bases is more likely to be low-leverage. A homer hit in each case results in more RBI in the high-leverage case.

This explanation is backed up by the HR numbers. Here are the HR hit per plate appearance in each situation. High: 0.025, Medium: 0.027, Low: 0.029. They are pretty similar. I would presume that HR (as well as batting average) are slightly lower in high-leverage situations because they often see an effective relief pitcher (such as a lefty specialist or a closer) come in to face the batter.

Anyway, I will be looking at the performance of some specific players in these situations. I just wanted to introduce the numbers for starters.

Check out Evan Longoria's splits by opponent this year:

Granted he has played far more games against Boston than most other teams, but his numbers are staggering, with 8 HR and 26 RBI in just 58 at-bats.  He has a 1.280 OPS against. Compare those numbers with his performance against the other team he's faced a lot this year, Toronto. No homers and 7 RBI in 57 at-bats with a .634 OPS.

Anyway, currently this site (nor any other) doesn't have the capability to search in such a fashion, but can you think of other instances of a player having at least 8 HR or 26 RBI against a single opponent in any one season?

(sorry for the very late post today...internet problems.)

Another indicator I like to look at to evaluate a team's performance is the innings pitched by starters vs. relievers. There are lots of reasons why any given team's starters might throw more of fewer innings. The two most common are:

• If the starters aren't as good or as durable, they tend to throw fewer innings per start
• If the starters are very young, their managers tend to have them throw fewer innings. This is piece of baseball wisdom that was largely ignored for many years but most managers now abide by.

So let's start in the American Leage by looking at splits by pitching role. The second set of data shows the number of games and innings pitched depending on whether the pitcher was a starter or a reliever. By clicking on the red text you can get an expanded breakdown by team such as this.

I extracted some of the key data and put it into an Excel spreadsheet. See here:

Starts    Relief    IP/start    IP/RG
KCR     82    208    6.09    1.08
MIN     83    223    6.05    1.06
TEX     81    204    6.04    1.15
TOR     84    235    6.00    1.08
LAA     81    216    5.97    1.08
CHW     82    203    5.96    1.15
BOS     82    234    5.95    1.04
SEA      82    208    5.89    1.21
DET     82    220    5.81    1.12
TBR     83    239    5.78    1.03
NYY     82    233    5.76    1.13
CLE     83    231    5.59    1.17
OAK     81    236    5.58    1.17
BAL     83    231    5.40    1.22

The four columns here are game starts, games in relief (total number of relief appearances) and then 2 numbers I calculated in Excel: innings pitched per start, and innings pitched per relief appearance.

Kansas City's starters are averaging 6.07 IP per start, meaning just barely into the 7th inning. Baltimore's starters are avering 5.40 IP per start, meaning that they get, on average, a bit more than 1 out in the 6th inning. This might not sound like a big difference, but so far Baltimore's bullpen has pitched 282.1 innings while Kansas City's has pitched only 224. That's nearly 60 innings of extra work being loaded onto Orioles relievers.

The point is this: with the exception of the closer and sometimes the setup guy, pitchers in the bullpen are usually the worst major-league pitchers. I'm not saying they are bad, but these are often guys that haven't been able to develop the pitching technique or stamina to be starters and are usually the worst performers on their teams. The more innings these guys pitch, the worse it usually is for their team because A) they are the worst pitchers and B) they have less stamina.

Point B) is the really significant one. It means that if they get used more over the course of the season, they are quite likely to see a significant decline in performance come August & September. (As an aside, sometimes playoff teams see a boost when they call up an extra bullpen guy or two in September but then can't use those guys in the playoffs and have to resort to their tired-arm relievers.)

When I look at the list above, the team the jumps out to me as being in trouble is Tampa Bay. They are averaging only 5.78 IP/start, one of the worst figures in the AL plus their relievers are pitching only 1.03 IP/game, the lowest figure in the game. Their relievers have already appeared in 239 games, most in the AL. More appearances by relievers means more times these guys warm up in the bullpen, more pitches thrown, and tired arms later in the season.

The Yankees and Tigers don't look much better. Both have IP/start figures similar to the Rays and have also seen many relief appearances from their bullpen.

Alternatively, the Rangers looks great. Their starters have one of the highest averages, as do their relievers. It means fewer innings and fewer pitches for their relievers. The Twins and Red Sox also look fine.

So what does this mean for the playoffs? Well for Tampa Bay, it's the opposite of what I wrote yesterday. Based on run scoring differential, they seemed to be shy a few wins and I expected to make them up later in the season. But based on what their pitching staff is doing, I also expect to see the team fade down the stretch. By combining those factors, I'd expect to see a similar record in the second half, meaning they'll finish with about 87 wins and out of the playoffs.

Yesterday, I liked the Twins, and today even moreso. I predict a fade for the Tigers and a third-place finish between the Twins and the White Sox.

And in the West, I like the Rangers.

Because of the strength of the Angels pitching, I'm taking them ahead of the Rays or the Yankees to win the AL Wild Card.

Now here's the same table for the National League.

Starts    Relief    IP/start    IP/RG
STL     85    258    6.13    0.90
SFG      83    222    6.12    1.02
CHC     81    236    6.11    0.99
COL     83    232    6.11    0.98
ARI     84    273    5.97    0.96
PIT     84    236    5.95    1.00
ATL     83    257    5.94    0.98
CIN     82    245    5.94    1.05
NYM     82    266    5.77    0.96
PHI     81    251    5.75    1.06
WSN     82    270    5.71    0.94
FLA     85    262    5.71    1.05
LAD     83    262    5.69    1.08
HOU     83    239    5.69    1.13
MIL     83    243    5.64    1.08
SDP     83    255    5.55    1.13

Right away we see some interesting differences between the leagues. The IP/start is just about the same between leagues (5.863 for NL and 5.840 for AL) but relievers in the NL tend to pitch fewer innings per appearance.  I assume this is due to the DH--in the NL if a pitcher gets in trouble and needs to be relieved, the reliever is going to come out of the game as soon as his place at bat comes up. In the AL, this concern doesn't come up. This probably means that it's a little more common for a reliever in the AL to come in to finish an inning and stay in to face one or more batters the next inning.

So in the NL East, all three leading teams (Phillies, Mets, and Marlins) are putting up some unimpressive pitching numbers. All three teams have low averages for IP per start. The Phils and the Fish, at least, have a higher average IP per relief appearance. The Phillies have one more ace up their sleeve, though, which is that the two biggest names in their bullpen, Brad Lidge and J.C. Romero, missed significant time to far this season, meaning that they will be fresher later in the season. Given that yesterday's run-scoring analysis suggested that the Phillies are already deserving of a bigger this, this reaffirms that they will run away with the division.

In the NL Central, both the Cardinals and Cubs have put up fantastic starting pitching so far, at least in terms of innings per start. Coupled with yesterday's results, this puts the Cardinals as the team most likely to win the divison.

The NL West gives us our first big surprise, where the Dodgers rank quite low on the list. The reason, though, is one of those exceptions I mentioned at the top--they've been holding up some of their youngsters from accumulating too many innings. After all, their team ERA is fantastic and their relievers are pitching just about the longest of any NL team. The Giants pitching has been great as well, but not enough to catch the Dodgers.

I do, however, like the Giants for the NL Wild Card.

So the picks are all the same results as from yesterday's analysis except for the AL Wild Card. I think the locks are Philly, LA, Boston, and Minnesota. I feel pretty good about the Rangers and Cardinals, as well as the Giants as the NL WC.

2008 Tampa Bay Rays Pitching Splits - Baseball-Reference.com

Fixed a couple of bugs in the team splits. Tooltips for team splits do not show up for splits we don't keep for players, e.g., "May, GS". They would always be blank.

Second, I fixed the bug that was causing the tooltip for the pitching stats output (e.g., the second output for home/road stats) to fail.

Yesterday Ken Rosenthal was reminding Red Sox fans how badly Carlos Delgado started out last year.

Indeed, check out the sum from his first 43 games last year (though May 21.) He had 5 HR and 20 RBI, to go along with 36 Ks, a .217 BA, and a .650 OPS.

Through his first 43 games this year, David Ortiz had 1 HR and 18 RBI, 45 Ks, a .194 BA, and a .594 OPS. Worse numbers than even Delgado last year. But you know what else? Delgado finished 2008 with 38 HR, 115 RBI, and an .871 OPS. His slow start is nothing but a distant memory. If Ortiz can have 80% of the season that Delgado had last year, he'll be fine. That bomb he hit against the Yankees last night helps.

Speaking of the Yankees, on ESPN radio, I heard the guys talking about how much better Derek Jeter is than Alex Rodriguez during critical times in games. (I think the theory came from all the recent discussion of how much better A-rod hits when games are out of reach, i.e. how many of his homers & RBI come during garbage time.)

Thanks to the new Leverage splits that were added during the recent site update, this is quite easy to measure.  Take a look at Derek Jeter's career leverage splits. For those who do not know, leverage calculations use Win Probability Added to determine how much of a swing any given situation has in the outcome of the game. It looks at how all games in the same situations have turned out historically and can therefore calculate the increased or decreased likelihood that a player's team will win that particuar game based on what the player does. It is, in my opinion, the best metric out there for measuring the true value of a player's offensive performance.

Anyway, you can see that Jeter has hit better in high- and medium-leverage situations than he has in low-leverage situations. Both his BA and OBP are significantly higher and his fractional OPS (tOPS+) shows that more of his own production comes in the higher leverage situations. In terms of RBIs, he has 0.24 RBI per PA in high-leverage, 0.10 RBI per PA in medium-leverage, and 0.07 RBI per PA in low-leverage situations. (This general trend would be expected as situations with runners on base are by definition higher leverage, so there are more RBI opportunities in high-leverage situations.)

With A-rod, the picture is a little different. He hits virtually the same in all three cases, with BA, OBP, and tOPS+ virtually identical for high-, medium-, and low-leverage situations. He has 0.37 RBI per PA in high-leverage, 0.16 RBI/PA in medium, and 0.12 RBI/PA in low. We see that his RBI production, while still shifted like Jeter's is somewhat closer to being equal in all three cases.

This supports the general argument that a greater fraction of Derek Jeter's production helps his team win games than for A-rod. That being said, A-rod produces so many more runs that he is still likely helping his team win more games than Jeter. The point being made on ESPN radio, though, was this this is why the fans adore Jeter and hate A-rod--because the perception (which really is truth) is that Jeter shines more when the game is on the line.