### Wednesday, April 06, 2005

## Should Some Lineups be Tweaked Already?

Due to spring training samples, or a reluctance to go with youth at a certain position, there are many position battles that I would not consider finished even though the season has started. I'll explain my take on these in detail, as well as throw a few hypothetical ones in there for kicks at some point.

Let's start off with Mark Ellis versus Keith Ginter in Oakland for the second base job. Ellis is the superior defended, but Ginter is the superior bat. How many runs would each be worth then, since defensive and offensive performance have to be put together to define their true value. Using Marginal Lineup Value Rate (MLVr) we can see the difference between the two offensively. MLVr puts the player into a lineup along with 8 league average hitters to determine how much of an effect this one player would have on runs scored. A league average hitter's MLVr would be 0.00. Keith Ginter's 2004 MLVr was .101, and Mark Ellis' was not available due to his missed time last year. So obviously we can see Ginter would win using his 2004 MLVr on the strength of his bat alone, but we want to project the future, so let's take a look at their PECOTA Projections. Ginter projects, as a weighted mean, to a .005 MLVr; a huge dropoff from 2004, but that probably has to do with his rising strikeout and falling walk rates, and the fact that to be above average he needs to slug homeruns (his 90th percentile projection has 21 homeruns and an MLVr of .181). Ellis', on the other hand,

I'm using Keith Woolner's formula found on the BP Statistic Glossary, which can be found here if your interested. You plug the hitter into the #5 spot in the lineup, because they are hypothetically supposed to get 1/9 of the total plate appearances; being in the middle makes this closer to the truth in the formula.

League Average = .270/.340/.440

League Average Runs and Runs Per Game: 847 runs, 5.2258 runs per game.

That figure for some reason seems rather high, but I'll go with it for now since this will all be relative. If someones knows what might be the matter, please comment. Even if whats the matter is my head isn't on straight today.

Team with Keith Ginter (weighted mean projection): 846 runs, 5.2230 per game (-.0028 from League Team).

Team with Mark Ellis (weighted mean projection): 833 runs, 5.1420 per game (-.1018 from League Team)

It was known Ginter was the better bat, but now we see by exactly how much over the course of a season: 13 runs if they both play to their weighted mean projection. How about their gloves though? Ginter projects to -8 Runs Above Average at second base, while Ellis projects to -1 Runs Above Average. That would make Ginter's run contribution over Ellis 6; a slim margin. Now to make things more confusing, the Oakland Athletics defensive ratings are different than BP's, and they view Ellis as a superior defensive second basemen. So now it depends on who you believe: BP or Billy Beane. Ginter is only 6 runs better than Ellis total if you use the BP method, so maybe trying Ellis out and hoping for increased plate production and the A's defensive ratings is worth a shot. If the A's are right, the slim 6 run margin disapears quickly and Ellis is clearly the superior player. This is how the Athletics are thinking at the moment, and they also like Ginter's ability to play third base (he's no Eric Chavez over there defensively, but he was at -1 Runs Above Average in 40 games from 04'). So after doing this, I guess I would take Ellis even though Ginter seems to provide more to the lineup based on his positional flexibility, power coming off the bench, and the fact that Ellis very well maybe an excellent defender depending on who you listen to.

On to the next roster battle of importance: Russell Branyan versus Wes Helms in Milwaukee. They are platooning, with Branyan getting the bulk of at bats against righthanders. Without the same calculations as last time, because you will see they are absolutely unnecessary. Russ Branyan's projected MLVr is .101, with Wes Helms at -.063. Obviously this won't be as close as the Ginter/Ellis battle, and even when you throw in defensive ratings Branyan is the superior, with a -5 Runs Above Average at third against Helms' -7. For a few interesting numbers, Branyan's Secondary Average (SecA) from 2004 was .424 (considering Secondary Average can be compared to regular average in order to determine effectiveness, you should all be bowing to the leader of the Three True Outcomes) and his Isolated Power (IsoP) was an impressive .291. He also had a .280 EqA...so he is the opposite of Alex Sanchez, owner of an extremely empty batting average.

Here's the kicker though: Helm's batted .306/.398/.444 against southpaws last year, while Branyan hit .167/.212/.267. Helm's 3 year splits against lefties (including a horrid 2002 that looks kind of like Branyan's 2004 against lefties) are .274/.365/.467. That is pretty good for a platoon, considering Branyan's three year splits of .227/.333/.465 versus righties (with 34 homers in 571 at bats, to go along with 90 (!!!!) walks). Branyan isn't some Tony Batista wannabe; he is a valuable player waiting for a starting job, and even if the strikeouts drive the fans crazy the souveniers will more than make up for it. Platooning Branyan and Helm's is a very smooth move by Doug Melvin and manager Ned Yost, as they now have a third basemen (when combined) that should hit well above league average and provide great production from a traditionally weak hitting position (try and remember Joe Randa and Bill Mueller and others man the hot corner as well as Scott Rolen and Alex Rodriguez; not everyone is a serious bopper.) Speaking of platoons...

Calvin Pickering deserves to be the starting DH for the Royals...but does Ken Harvey deserve to platoon? The Royals are already platooning at

Let's start off with Mark Ellis versus Keith Ginter in Oakland for the second base job. Ellis is the superior defended, but Ginter is the superior bat. How many runs would each be worth then, since defensive and offensive performance have to be put together to define their true value. Using Marginal Lineup Value Rate (MLVr) we can see the difference between the two offensively. MLVr puts the player into a lineup along with 8 league average hitters to determine how much of an effect this one player would have on runs scored. A league average hitter's MLVr would be 0.00. Keith Ginter's 2004 MLVr was .101, and Mark Ellis' was not available due to his missed time last year. So obviously we can see Ginter would win using his 2004 MLVr on the strength of his bat alone, but we want to project the future, so let's take a look at their PECOTA Projections. Ginter projects, as a weighted mean, to a .005 MLVr; a huge dropoff from 2004, but that probably has to do with his rising strikeout and falling walk rates, and the fact that to be above average he needs to slug homeruns (his 90th percentile projection has 21 homeruns and an MLVr of .181). Ellis', on the other hand,

*only*MLVr on the positive side of the spectrum is at his 90th percentile projection, and is .096; less than what we could expect from a 2004 Ginter repeat (of course, Ginter may not repeat 2004 unless he corrects those falling rates).I'm using Keith Woolner's formula found on the BP Statistic Glossary, which can be found here if your interested. You plug the hitter into the #5 spot in the lineup, because they are hypothetically supposed to get 1/9 of the total plate appearances; being in the middle makes this closer to the truth in the formula.

League Average = .270/.340/.440

League Average Runs and Runs Per Game: 847 runs, 5.2258 runs per game.

That figure for some reason seems rather high, but I'll go with it for now since this will all be relative. If someones knows what might be the matter, please comment. Even if whats the matter is my head isn't on straight today.

Team with Keith Ginter (weighted mean projection): 846 runs, 5.2230 per game (-.0028 from League Team).

Team with Mark Ellis (weighted mean projection): 833 runs, 5.1420 per game (-.1018 from League Team)

It was known Ginter was the better bat, but now we see by exactly how much over the course of a season: 13 runs if they both play to their weighted mean projection. How about their gloves though? Ginter projects to -8 Runs Above Average at second base, while Ellis projects to -1 Runs Above Average. That would make Ginter's run contribution over Ellis 6; a slim margin. Now to make things more confusing, the Oakland Athletics defensive ratings are different than BP's, and they view Ellis as a superior defensive second basemen. So now it depends on who you believe: BP or Billy Beane. Ginter is only 6 runs better than Ellis total if you use the BP method, so maybe trying Ellis out and hoping for increased plate production and the A's defensive ratings is worth a shot. If the A's are right, the slim 6 run margin disapears quickly and Ellis is clearly the superior player. This is how the Athletics are thinking at the moment, and they also like Ginter's ability to play third base (he's no Eric Chavez over there defensively, but he was at -1 Runs Above Average in 40 games from 04'). So after doing this, I guess I would take Ellis even though Ginter seems to provide more to the lineup based on his positional flexibility, power coming off the bench, and the fact that Ellis very well maybe an excellent defender depending on who you listen to.

On to the next roster battle of importance: Russell Branyan versus Wes Helms in Milwaukee. They are platooning, with Branyan getting the bulk of at bats against righthanders. Without the same calculations as last time, because you will see they are absolutely unnecessary. Russ Branyan's projected MLVr is .101, with Wes Helms at -.063. Obviously this won't be as close as the Ginter/Ellis battle, and even when you throw in defensive ratings Branyan is the superior, with a -5 Runs Above Average at third against Helms' -7. For a few interesting numbers, Branyan's Secondary Average (SecA) from 2004 was .424 (considering Secondary Average can be compared to regular average in order to determine effectiveness, you should all be bowing to the leader of the Three True Outcomes) and his Isolated Power (IsoP) was an impressive .291. He also had a .280 EqA...so he is the opposite of Alex Sanchez, owner of an extremely empty batting average.

Here's the kicker though: Helm's batted .306/.398/.444 against southpaws last year, while Branyan hit .167/.212/.267. Helm's 3 year splits against lefties (including a horrid 2002 that looks kind of like Branyan's 2004 against lefties) are .274/.365/.467. That is pretty good for a platoon, considering Branyan's three year splits of .227/.333/.465 versus righties (with 34 homers in 571 at bats, to go along with 90 (!!!!) walks). Branyan isn't some Tony Batista wannabe; he is a valuable player waiting for a starting job, and even if the strikeouts drive the fans crazy the souveniers will more than make up for it. Platooning Branyan and Helm's is a very smooth move by Doug Melvin and manager Ned Yost, as they now have a third basemen (when combined) that should hit well above league average and provide great production from a traditionally weak hitting position (try and remember Joe Randa and Bill Mueller and others man the hot corner as well as Scott Rolen and Alex Rodriguez; not everyone is a serious bopper.) Speaking of platoons...

Calvin Pickering deserves to be the starting DH for the Royals...but does Ken Harvey deserve to platoon? The Royals are already platooning at

*both*outfield corners, and manning three platoons is difficult, so I think they may be banking on Pickering learning to hit lefties. Hopefully this turns out more like Eric Chavez 2004 (.306/.412/.481 after years of struggles) than Trot Nixon's career (.220/.293/.348 versus southpaws) as far as hitting lefties goes, if not this year than maybe next. Harvey's three year splits against lefties: .306/.347/.489 are much better than his righty numbers over the same time (.263/.316/.382). Pickering has hit .214/.250/.310 against lefties over the same time, while smashing righties for .263/.378/.600(!). Obviously this team could use another platoon split in order to hit better, but with the two platoons in the outfield corners (Terrence Long and Eli Marrero in left, while Matt Stairs and Emil Brown platoon in right) it makes it difficult, unless you carry 11 pitchers. Why any team is carrying 12 pitchers is beyond me, unless your the Rockies, but that is a whole different matter I can address another time. If the Royals would carry 11 pitchers then they would have 3 platoons and good production out of three controlled environments. Maybe later in the year when Pickering struggles against lefties than "Grimace" can come up from Omaha to lefty mash. By then maybe one of the outfield platoons will be unnecessary, or the DH platoon would just have more benefit. I'm hoping for it, because Harvey does need to be used in the role he is suited for, and as much as I love Pickering on the roster Harvey's place is not in AAA.**Note:**Right before I went to publish this article, I saw this on ESPN.com. Hopefully nothing is seriously wrong with Tito and he'll be back soon; make sure you send some good vibes out Tito's way.