Guillen as a Weaver-Style, January, Replacement Level Player

...............

Seen at Bill James Online, in the "Hey Bill" section:

I always thought that perhaps Weaver's system worked, but not necessarily for the reason people thought it worked.   Weaver collected outfielders, first basemen and DH types, occasionally catchers and third basemen, with severe limitations but identifiable strengths. . .left-handed .320 hitters who couldn't run, right-handed power hitters who couldn't throw, short switch-hitters who walked every day but struck out twice. He rotated these players based on matchups, true, but he might have been equally successful if he had rotated them in and out of the lineup by pitcher height, hair color and biorhythms.

     And, by the way, that system would work better now than it did then.

The enormous expansion of the bullpens since the 1970s has trapped in the minor leagues dozens of the types of players that Weaver worked with, the two-dimensional outfielders.   It's a lot easier to find those guys now than it was in Earl Weaver's day.

This is an interesting insight. In the 1970's, teams ran 10- and even 9-man pitching staffs. Now you see 12-man staffs and, correct me if I'm wrong, but haven't there been some 13-man staffs at times?

The difference between a 10-man staff, in Weaver's day, and a 12-man staff nowadays, well, that's the difference between a 6-man bench and a 4-man bench. But! Hold on now - you've got to have a backup catcher and a backup shortstop. So you're talking about four bench players vs. two.

It hadn't really occurred that we're talking about half as many jobs for the Carlos Guillens of the world. Maybe 50% of the 1975 Carlos Guillens in Major League Baseball are now in the minor leagues?

Meaning that JANUARY "replacement level players" -- guys you can get in January, not in June, now -- face a very high bar.

............

Carlos Guillen, when last seen, was quite a hitter for a bench infielder, especially by Seattle Mariner standards. He's been, essentially, a 5.0 runs per 27 outs player. The two guys who did that for the M's last year were Dustin Ackley (5.6) and Mike Carp (4.9). Adam Kennedy finished at 3.5, after the late collapse that Geoff Baker prophesied.

Mariners 2011 3B	.195/.252/.275	50 runs created
Chone Figgins 2011	.189/.241/.246	11 RC in 77 games (!)
Carlos Guillen 2009-11	.250/.330/.400	70 RC per season

Chone Figgins hit like a pitcher, literally like a pitcher, as did the M's center fielders in 2011. Carlos Guillen, by contrast, is an average-solid ML hitter, one who knocks in runs at a normal rate of speed - 80 RBI per 160 games. Replacing a pitcher in the lineup, with an ML hitter, is a big deal.

Guillen is definitely a "severely limited" player "with identifiable strengths." If the M's can keep him on the field, he'll help them win games. Baker warned the M's about riding Adam Kennedy too hard; if the M's use Guillen 2-3 times per week, and don't get tempted to ride his hot streaks, they might keep him out there. It's not like he's never posted 300 at-bats.

.......

James was asked why the bullpens have expanded so much. He replied, ONE reason is that ballclubs have realized that there are many, many useful relief pitchers in baseball. There are scads of guys who cannot face twelve hitters in a game and survive, but use them 1-2 innings and you get good results.

Wilhelmsen, Kelley, Delabar, Ruffin, Furbush, Sherrill, League, Beavan, and who knows who else going for 1-2 innings at a time ... the M's chances to get good performances out of their #4-9 hitters, and #4-9 pitchers, are increasing as we go along.

Cheers,

Dr D

Printer-friendly version

Categories:

Sports

Tags:

Carlos Guillen

Replies

This illustrates a big difference between '11 and '12

Last year, Kennedy was legitimately one of the M's best hitters to start the year. However, since he was just a scrub, over-the-hill bench player pulled off the scrap heap, that was a very bad thing. With Figgins, Jack Cust, Milton Bradley and Jack Wilson as starters, Kennedy and Smoak had to carry the offense at the beginning of the season and there was no way that could last.

This year, if Guillen hits well it will serve to supplement a whole bunch of other quality offensive players instead of being a main driving force. That's a drastically better situation.

Not out of the question

All Guillen has to do is revert to the form he showed in '09 and '10 and he's a big help. He was a 300 AB/100 OPS (roughly) bat for those years. THAT would be a tremendous help. That level of performance is certainly not out of the question.

In fact, it is probably to be expected.

Personally, I would ratehr see those 300 ABs go to Liddi (who would be a great lefty/righty platoon partner with Seager) but I think bringing in Guillen probably indicates that isn't going to happen.

And I suppose Guillen's future with the M's is really dependent on Figgins' ST, if we keep him that long.

moe

Ya, there's a reason they're going to Guillen ...

Personally, I would ratehr see those 300 ABs go to Liddi (who would be a great lefty/righty platoon partner with Seager) but I think bringing in Guillen probably indicates that isn't going to happen.

... if Zduriencik were in rebuild mode - the kind of rebuild that the internet means, saying "No Mas" in February - he wouldn't have Guillen (or Sherrill, or Jaso, or Millwood, or ...) here, right?

I think it'll be more important

To make sure Guillen never, ever sees left-handed pitching. Since 2009

vs. Left: .228/.291/.331

vs. Right: .260/.332/.434

Even last year, in an admittedly miniscule sample size;

vs. Left: .185/.179/.185

vs. Right: .250/.297/.441

It makes him a terrible platoon partner for Seager, I imagine him getting a handful of starts at 3rd, Left, 1st, and probably as many as all those totalled at DH for days when Carp is in Left and Wells is in Center.

And, of course

Montero at Catcher

Exactly

Exactly. Z's trying to put together a team that could win a bunch of games if everything goes right. 300+ Guillen AB's, with 15 doubles and 8 taters would go a long way to helping.

Two main things scare me about the M's offense right now: #1. That we get to watch too much of Figgins. Over the past 4 seasons, Figgins has had 1 horrible one, 2 lousy ones, and one good one. The good one was driven completely by an outlier # of walks. That # of walks will never happen again because every pitcher just flat out challenges him. He can't hurt them so they don't nibble. #2. That we get to watch too much of Guti in CF against RHP, whom he flails against terribly. But I suppose I'm beating that point to death.

Used correctly, we have enough bats to score runs.

The Guillen addition doesn't hurt.

moe

Ignore platoon splits

This is a classic case of how having more information makes us understand a player less. Platoon splits of right-handed hitters are almost always misleading and should be ignored. Even with 2000 PA's against southpaws (which would take over 10 full seasons), you would have to regress a righties observed platoon split by 50%.

Gutierrez has only 700 PA's against lefties so his splits aren't anything to make hay over. If he can hit lefties he can hit righties; they aren't two completely separate skills. If he hits like he is capable of then he will be a very good full-time player.

Raises hand.

Could you elaborate on why righties splits are misleading and even at 2000 pa you would need to regress them?

Short answer:

Because that is what the measured variance of Major League platoon stats suggests.

Long answer:

There are far more right-handed pitchers than lefties, especially at lower levels of the game, like high school and college. This results in the right-handed hitters who reach the Major Leagues having very little difference between them when it comes to platoon splits. There are two reason for this.

First, batters are selected based upon their ability to hit right-handers. If a guy is strong against lefties but weak against righties, his scouting report would not read "Strong against left-handers; weak against right-handers". Instead, his scouting report would just say "Can't hit".

Second, they are forced to face tons of pitchers who have the advantage over them which means by the time they reach the majors they will have maximized their ability to hit those types of pitchers.

Think about penmanship. There is a big difference between how nice some people can write compared to others. There is also a big difference between how most people write with their strong hand compared to their weak hand. Now imagine if everyone were forced to write with their left hand throughout the entire time they were in school. If after their senior year they were then required to take a test measuring their penmanship with both their right hand and left, what would be the result?

For the natural left-handers, you would likely see a wide range of differences between how they write with their right as opposed to left hand because they had so little practice with their weak hand. So a person with very good penmanship could still be really bad with their right hand while person with sloppy hand writing might be just a little worse with their right hand.

But for the natural righties, you wouldn't see as big a range of differences between right and left because they would have so much practice with their weak hand that they were as good as they could be with that weak hand.

We can extend this analogy further. If colleges required students to have good penmanship with their left-hand to be admitted, what would happen? Number one, it would be far easier for natural lefties to get in. Number two, natural righties would have a small variance between how they write with their left-hand compared to their right hand because the couldn't get in if they struggled with their weak hand.

This is the way it is in the Majors: lefty batters are far more prevalent than they are in the general population and righty batters have very similar actual (as opposed to measured) platoon splits.

Not so, Papa

Splits are not something to be ignored. In nearly 1800 career PA's, Guti is a .246-.294-.354 vR player. 1800 is a significant sample. Brendan Ryan, in 1200 career PA's vR, is .252-.303-.341! 1200 PA's is a significant sample. Like it or not, Guti = Ryan, vRHP.

Ryan projects to just 12 homers in 1800 PA's vR, while Guti has 30. But Ryan projects to 80 doubles and 15 triples, while Guti has 71 and 7. 107 extra-base hits for Ryan, 108 for Guti.

1800 PA's is three full seasons of PA's. It a guy is a .648 OPS player after three full seasons, it suggests he isn't a very good hitter.

Guti isn't a very good hitter vs. RHP.

Now Wells has only had 178 MLB PA's vs RHP. That is a quite small sample. But in it he's a .800+ OPS guy. As the season starts, Wedge should give him the bulk of the vR starts. There is the hint that he is a significantly better bat than Guti, in that situation. We need a significantly better bat.

Splits count.

moe

An explanation...and the truth about Guti.

When someone says you need to be careful about the vL split for a right handed batter...they certainly don't mean ignore their actual performance against right handed pitching. They mean that because right handers more often face their disadvantaged platoon side, it's statistically less likely that that platoon side will actually be a significant disadvantage. They mean...ignore the vL performance and assume the righty's actual ability is much closer to his vR split.

If Guti is a .240 hitter vR...then he's a .240 hitter. Period. Which means he should never play at all. If he has a better year than that in 2012, it won't be because he's a "lefty masher" playing only against the lefties...it'll be because he hits better all around.

BTW, I think you folks are also reading WAY too much into Guillen's signing (even Doc...sorry, dude...can't follow you down this rabbit hole).

I don't think it's at all clear that he'll make the club, let alone that the Mariners have some grand plan to win in 2012 with crafty veretan leadership like Kennedy. I view Guillen as veteran competition for rookie ABs...a guy who'll force Liddi and Seager and a couple of the other young bats to earn their big league roster spots and their playing time by outhitting a replacement level vet. Zduriencik likes to force his kids to earn their keep rather than handing them jobs. Even in a rebuilding year, he wants them to have to work for their pay. I think this is a good thing...and I really do not think it means no shot for Alex Liddi.

Maybe we should discuss regression...

especially as it applies w/r/t Guti's lefty-righty splits...

Regression is a more elusive concept even than the megastar sabermetricians seem to grok.

Bottom line, Dr. D groks that Guti's platoon splits will close up some, yes .... his rationale for that should be worth a free paragraph or twenty... >:-]

"1800 is a significant sample."

You can not say that something is a "significant sample" if you don't know what the spread in the population is in regards to that skill. Some stats require a relatively small sample to be meaningful because there is large spread in the measured skill, like groundball-rates. Other stats require a massive sample (almost a whole career) to tease anything out because there is almost no difference between players, like with clutchness. In the case of right-hander platoon splits, it is much closer to the latter. Read The Book to understand what I am talking about, that's where I got the reqression number. For lefty batters the regression is about half that of righties (meaning their is a larger variation among them) and for switch-hitters it's even less, with about 600 PA's needed to regress half-way (which makes sense because they are almost like two different hitters).

The reality is that there is very little variation amongst right-handed hitters (at least as far as major leaguers are concerned) when it comes to their platoon split. So if you see a guy like Guti who has hit lefties extremely well but poorly against righties, then he is almost certainly a better batter against righties than his numbers indicate (and worse against lefties than his numbers suggest). Therefore, platooning him with another right-hander like Wells wouldn't do you any good, especially since Franklin is a better defender.

agreed...except...

...for three words in your post which I do not agree with.

"better against righties"

Nope...that's not my understanding of The Book at all. My understanding is...you regress the small split toward the big one a heck of a lot more than you regress the big split to the small. So...if by better hitter you mean he'll bat .244 instead of .240...then sure. Otherwise, I'll stick with my interpretation that Guti is, essentially, a .240 hitter against all comers so far in his career and his lefty split is probably inflated. I am not saying I would necessarily assume he'll hit .240 next year. I left room for the possibility that he could improve his real performance, leading to a global batting statistic improvement...but he'll have to actually hit better to get better results. If you could somehow guarantee that he'd only face lefties next year...I'd be projecting him to hit about .250...not .290.

This must be right

My understanding is...you regress the small split toward the big one a heck of a lot more than you regress the big split to the small. That must be right. It will be much easier for Guti to raise or lower his vL split because he has less than 40% of the total chances there that he has against righties.

4 years of Guti hitting righties at a .262-.309-.372 clip (his vR line in '09, his break out year) to raise his career avg/OBP/slg 8 pts each!

His career vR #'s aren't going anywhere fast.

moe

It's not about raising one or lowering another

They aren't completely separate skills. If you have two right-handed hitters and both of them are "true" talent .280 batters against right-handers, then you should expect them to hit lefties about equally well regardless of how they had hit lefties in the past.

Or let me put it this way. If you have two righties who both have an .800 OPS over the last 3 years, with one having a very large platoon split, the largest in the game, and the other having no split at all (he had an .800 OPS against both lefties and righties), what should you expect going forward? You should expect them to have essentially the same split going forward. That means it's very possible that the guy who had a huge split could suddenly have almost no split and the guy with no split have a giant split.

Take a look at Edgar Martinez. In most years, he would hit lefties better, sometimes dramatically so, than righties. But in a few years, like in 1998 and 2001, he actually had a reverse platoon split, doing a lot worse against southpaws than righties. That's why we should just focus on a players totals instead of slicing up his numbers into smaller samples that have more noise.

no disagreement

I'm just saying that if we expect Guti's splits to move toward some new level...that level will be much closer to his vR split than the vL split.

Sure

But that means that platooning him with another right-hander like Wells doesn't do you any good. If Wells is a better hitter versus righties than Gutierrez, then he's also a better hitter versus lefties.

That's exactly what I'm saying

Ignore the lefty/righty numbers and just look at his overall stats.

Hope you're right, Ghost

Ghost, I would love to see a Seager/Liddi 3B platoon.

In the 6 seasons where Guti has more than 97 PA's vR, he has never been a league average hitter against righties. In '07 he OPS'ed (relative to the league vR) 97. Other than that he has OPS'ed 74, 84, 88, 85 & 54. Even if you discount last year, he is still well below the league average vR. He slugged .429 in .07 vR, but other than that he hasn't been above .372....so he doesn't mash the ball either.

On the other hand, he does OPS 130 in relation to the league average against lefties. He creams lefties. Against lefties, his OBP is 50 points better and he slugs 100 points better.

He's the perfect definition of a perfect platoon player, if the M's have a capable CF who can hit righties. I think that is Wells.

If I've done it correctly (which I may not have), based on career numbers, Gti creates 42 runs in 400 AB's vR, and Wells (smaller sample, I know) creates 63.

If correct, that's significant. Does Guti save 20+ runs over Wells on defense? I don't know.

I'm suspect we see a bunch of Guti early on (return on investment, I suppose). I hope he's a bat vR. I'm just not betting much on it.

moe

Seattle Sports Insider

You are here