Drew v. 2.0

[glasspusher]

Is the error itself in game two +17, or the whole play? Even without the error, it's a hit, and first base no out in a walk off inning at home. It was a one base error, not two. It was still a hit. I am surprised one base makes a +17 difference. Is it possible that number includes the single? In any event, I think knowing the teams and the situation and the emotional way one watches a game affects how one views a play.. Even without the error in game 2, the Sox felt in control. The error in game 5 tipped the balance, and seemed to turn the game from one in which the Tigers were in control to one in which the Sox were in control.

Put it another way. The Sox were probably already in the high sixties with the lead off single in the 9th in game 2, so the error made a high WP higher, which is great but not as dramatic. The game 6 error, on the other hand probably changed the horse race and moved the Tigers from the probable winner to the probable loser.

I'm trying to figure out how he fucked up Ells' ball. It was in his glove! As near as I can tell, going with the two handed catch, if you don't get your hand over the ball in the glove, the act of smacking your hands together launches the ball out of the glove, as opposed to a one handed scoop where the ball and the glove stay at the same relative speed and holding on to it. If he does that, the Tigers get one out, at least. I guess he had proper form, just didn't execute it properly. Incredible.

 

[smastroyin]

A friend theorized that he was going for the flip toss to infante.

 

[YTF]

A friend theorized that he was going for the flip toss to infante.

This was mentioned in game as well, but the replay didn't really show any sort of movement toward making that sort of play. Originally I thought that perhaps the ball going directly over second base (white over white) may have cause Iggy to momentarily lose the ball, but the replay clearly shows the ball right in the glove and he really doesn't look as if he was distracted or looking toward anything other than the play he was trying to make. Did he simply have a brain fart?

 

[Red(s)HawksFan]

This was mentioned in game as well, but the replay didn't really show any sort of movement toward making that sort of play. Originally I thought that perhaps the ball going directly over second base (white over white) may have cause Iggy to momentarily lose the ball, but the replay clearly shows the ball right in the glove and he really doesn't look as if he was distracted or looking toward anything other than the play he was trying to make. Did he simply have a brain fart?

 
It looked to me like he was trying to make a quick transfer in order to throw the ball (whether to Infante or to first will never be known) and simply missed the ball with his throwing hand.  His right hand is going into the palm of the glove as the ball bounces out of the webbing and past his fingers.
 
I would think if he was considering trying to flip the ball with his glove to Infante, his throwing hand wouldn't have been in the picture at all.

 

[TheBenzingerGame]

Yeah, the sweet play on Cabrera in the top of the 7th was worth -0.48 WPA - his subsequent strike out with Gomes on 2nd for the 1st out in the bottom of the 7th was worth -0.90.....

 
Drew's defensive play in the top of the 7th contributed 4.8% to the Sox' WPA as measured in hindsight, yes. But what this particular view of WPA fails to capture is the alternative: what if the shortstop in that scenario did not make the play?
 
Had the ball gotten through, the game state would have changed from 2 out, 12-, DET by 1 to 2 out, 12-, DET by 2. That outcome would have presented a 12.6% swing in the other direction (and the inning would have continued, so who knows what else unfolds from there).
 
Is this a fair way to measure the impact of one discrete event? Because if so, the swing in WPA of that play was actually 17.4% (with the potential for more).

 

[Kevin Jewkilis]

 
Drew's defensive play in the top of the 7th contributed 4.8% to the Sox' WPA as measured in hindsight, yes. But what this particular view of WPA fails to capture is the alternative: what if the shortstop in that scenario did not make the play?
 
Had the ball gotten through, the game state would have changed from 2 out, 12-, DET by 1 to 2 out, 12-, DET by 2. That outcome would have presented a 12.6% swing in the other direction (and the inning would have continued, so who knows what else unfolds from there).
 
Is this a fair way to measure the impact of one discrete event? Because if so, the swing in WPA of that play was actually 17.4% (with the potential for more).

 
FWIW, the Leverage Index measures that.  The LI of the play (which accounts for possibilities like a home run that at that point weren't going to happen) was 1.87 (where 1.0 is average).  That puts it as the 16th most important play of this particular game.  For context, the at bat where Victorino hit the grand slam had an LI of 5.08.

 

[TheBenzingerGame]

 
FWIW, the Leverage Index measures that.  The LI of the play (which accounts for possibilities like a home run that at that point weren't going to happen) was 1.87 (where 1.0 is average).  That puts it as the 16th most important play of this particular game.  For context, the at bat where Victorino hit the grand slam had an LI of 5.08.

 
Cool, thanks, KJ. And here's a link to the LI data from last night's game, for anyone else interested.
 
Of course, even LI still just tells us how important the moment was - not how difficult the play may have been. Seems to me you'd have to do something like combine LI with PMR (or some other range-based defensive value) to really judge the value of that play in the context of the game. Is that being done anywhere we know of?

 

[smastroyin]

WPA only assigns value to the pitcher or batter/baserunner.
 
Maybe someone is working on something to quantify how much a fielder contributes to each play, but right now even for batters for instance if you reach on an error, you get the same credit as the single and the pitcher gets the same penalty.

 

[DennyDoyle'sBoil]

Clutchness type stats and leverage stats don't seem really to apply to fielders so much.  A guy that gets to a certain batted ball most of the time is, for most of the purposes you want to know stats for, going to get to it most of the time with a 9-0 lead or a 1-0 lead.  I think everyone is potentially susceptible to getting the yips in a very big moment, but by and large when you see a guy like Drew make the play he made in the seventh inning, you sort of just have to conclude, "that's a player who can make that kind play, hopefully he'll do it in a big spot."  There doesn't seem to be a very large branch of sabermetrics that I'm aware of that tries to figure out whether guys field better or worse depending on the LI of the at bat.  Maybe there is?
 
To me the exercise is more about trying to figure out how a particular fielding play impacted a game -- this is something one would usually do with an unusally good or unusually bad play.  And for that purpose, whatever the LI was for the situation when the batter was standing in the box, it doesn't seem nearly as interesting or significant as taking the result of the play and comparing it to the likely result of the play if a normal play is made, and then seeing how it affected win percentage up (if it was an abnormally good play) or down (if it was an abnormally bad play).  
 
Consider the home team leading by 1 run in the bottom of the ninth with two out and bases loaded.  If a fielder makes an outstanding play or a bad play in that moment, I'm not sure the fact that it was a 10 or whatever it was in terms of LI tells you very much at all about the player, unless the player has a habit of making bad plays in that situation, which you probably don't need stats to tell you if you watch the team enough.  Also, whatever the LI is when the batter steps in the box, it certainly can change once the ball is batted.  Consider the same situation, with a screaming line drive hit 8 feet high right at the first baseman.  As soon as the ball leaves the bat, the leverage of the play is whatever corresponds to "total leverage" and the WPA difference, in that moment, is a spread of 100 percent.  Over his head, they lose.  In his glove, they win.  I suppose there are a few handful of unlikely other outcomes too, but by and large, in the moment, that play was the ball game, one way or the other.  The fielder's ability to execute or not was the difference between a win and a loss.  I think spread in win percentage between making the play and not making the play in less black and white circumstances is equally interesting when going back over a game and trying to figure out what the impact of the play was.  One of the reasons you can do this is that LI, as I understand it, takes into account all of the possible outcomes, but when you actually look back at a play after the fact, you can with good accuracy identify what the result would have been but for the abnormally good or abnormally bad play.

 

[Koufax]

According to the link provided above, there were 6 plays that were that were favorable to the Sox that were higher leverage than Drew's play in the 7th inning. 
 
Most of them involved taking advantage of unforced errors on the part of the Tigers:  Two walks, Ellsbury's reaching base on Iggy's error and Pedroia taking third on a wild pitch. While I have a lot of respect for X working a walk from an 0-2 count, I don't think that qualifies as a more clutch and impressive play than Drew's play at SS.  
 
The other two were hits -  the Gomes double and of course the Victorino GS.  The LI of Gomes's hit was just a hair higher than that of Drew's play, leaving Drew's play as the third highest leverage positive RS play of the game. 
 
I'm not suggesting that we sign Drew for next year, only that we have in Drew what we always said we wanted with Iggy -- a top-notch, light-hitting SS.  We should appreciate what we have while he's here.

 

[lambeau]

Drew's strikeout in the seventh reduced the RS win probability by 10% (no question his beautiful fielding play in the top of the inning saved an equally important run).
 
In a .111 OPS slump he needs to bunt Gomes over; the win probability declines 1% for an average hitter--which he's currently not--but the net increase is 9% relative to a K.

 

[OzSox]

 
It looked to me like he was trying to make a quick transfer in order to throw the ball (whether to Infante or to first will never be known) and simply missed the ball with his throwing hand.  His right hand is going into the palm of the glove as the ball bounces out of the webbing and past his fingers.

 
I agree, and I also have an additional theory.
 
The most bizarre thing about the play is that the ball was in his glove, in the right spot in his glove, and yet somehow it came out, despite being an unbelievable defensive player. What he didn't do was appear to make any attempt to close his glove around the ball, he just went straight in with his throwing hand. If you watch closely he went in with his throwing hand to the part of the glove where the ball hit originally, but had since moved on from.
 
So perhaps that's the secret to Iglesias' lightning-quick transfers, especially on slow grounders - maybe he 'cheats' on some balls by avoiding the closing and reopening of his glove altogether to gain him more time. In this case he just messed it up.

 

[Red(s)HawksFan]

So perhaps that's the secret to Iglesias' lightning-quick transfers, especially on slow grounders - maybe he 'cheats' on some balls by avoiding the closing and reopening of his glove altogether to gain him more time. In this case he just messed it up.

 
I absolutely buy that theory.  I remember seeing a piece on Omar Vizquel years ago.  It might have been a Sportscenter thing or maybe a TWIB story, I can't be sure.  What stood out to me is they revealed the "secret" of his lightning quick turns at the bag on double plays.  He used to catch the throw from the second baseman by trapping it between the outside of his closed glove and his throwing hand instead of opening the glove and catching it the "normal way".  By doing that, he was more or less barehanding the toss and had the ball in his throwing hand instantly to make the relay throw to first.
 
I want to say they also showed him handling a lot of groundballs, routine ones anyway, without ever actually closing his glove around the ball.  He secured the ball by putting his throwing hand in there and making the transfer faster, much like you describe with Iglesias.  Must be some sort of secret of the trade for these elite infielders.

 

[smastroyin]

I'm not trying to be a dink but one problem with super slo mo is that routine things like this get magnified.  All infielders have many tricks to speed up the transfer.  It would be almost impossible to turn double plays if they didn't.  Just watch a few instructional videos and you'll see things like reaching in before the glove is closed, flipping the ball from the glove into the throwing hand, etc.  As well, there are transfer errors all the time, it's not like it is unusual for a ball to be in the web and then have a subsequent error.  I realize there is some shock that Iglesias would do it, but it's not like it is a specifically unusual play in and of itself.
 
The fact that this stuff gets pulled off so routinely is what makes these guys major leaguers.

 

[Bucknahs Bum Ankle]

Yeah, that's not really a trade secret at all.  Kids in many Latin American countries have routinely used oven mitts or the like to field grounders.  Basically just use them to shield the impact of the ball and get their bare hand there quickly to secure it.  I've heard of the technique beginning to gain popularity even at the little league level in the US.

 

[Skiponzo]

FWIW I'm a little league coach and I have my kids begin each practice by taking grounders with no glove.

 

[Al Zarilla]

FWIW I'm a little league coach and I have my kids begin each practice by taking grounders with no glove.

When coaching the first year on a new field and they hadn't prepared the infield well, a kid at short got a grounder in the mouth, off a rock I presume. Be careful with that. Maybe you roll the ball instead of hitting rocket one hoppers.   

 

[TheYaz67]

I know it is useless to argue that Drew be benched, since his defense has been more than good so far in the playoffs and having youngsters at SS and 3B is a bit risky (and you loose a left handed bat), but again, more at bats last night that show how poor his approach has been and/or he is not seeing the ball well / guessing or something.  Its not just the horrible results, its the shitty approach which is bothering me.  I mean, what "advantage" does being a lefty confer if you just have a bad approach?
 
First at bat ("single" - aka Wainwright brain fart):
 
1st pitch:  Called FB strike, bottom edge of zone (fine approach - they are all taking 1st pitch first at bat as they should)
2nd pitch:  Lays off close high & outside ball (good)
3rd pitch:  Swings at ball two - high FB (bad)
4th pitch:  Lays off bounced curve in dirt that missed badly
5th pitch:  Swings at what should have been ball four, popping up to pitcher (bad)
 
So as the leadoff batter in the 2nd, he should have been out for swinging at mostly balls and making poor contact (when everyone else was mostly squaring up Wainwright's stuff).
 
Second at bat (strike out looking):
 
1st pitch:  Called strike (curveball) in lefty wheelhouse (ok, not looking for that one, so fine take)
2nd pitch:  Takes FB that misses badly (good)
3rd pitch:  Fouls back curve that is almost dead center in middle of the zone (bad)
4th pitch:  Swings at low inside curveball that was a ball, strikes out (bad)
 
 
Third at bat (Axford pitching - strike out swinging):
 
1st pitch:  FB cookie down middle, foul ball (bad)
2nd pitch:  Low outside pitch - ball according to FX, called strike (not his fault obviously)
3rd pitch:  Badly bounced curve, ball 1
4th pitch:  Badly bounced FB, ball 2
5th pitch:  Pitch up in zone, fouled off (bad)
6th pitch:  Curve well below strike zone, swing and miss (bad)
 
Fourth at bat (Martinez pitching - ground out to first):
 
1st pitch:  Sinker outside, ball 1
2nd pitch:  Sinker outside, ball 2
3rd pitch:  FB low inside, fouled off
4th pitch:  Sinker same location, ground out to first
 
I mean, the only positive thing I can say about his efforts at the plate last night was that he made them throw 19 pitches in his 4 at-bats, which beat out Napoli and X's 18.
 
You could go back through all his postseason at bats and see the same things, over and over - taking obvious strikes, swinging at balls out of the zone, striking out looking, or making weak contact for an out, often on a pitch out of the zone.  I mean, even the vast majority of pitches he laid off in this game were no where near the strike zone, so hard to give him huge credit for his "patience" - you and I would not have swung at most of those offerings.  And he got several gift pitches that were dead in the middle of the zone, and could do nothing with them.
 
He just looks completely lost out there at the plate, despite getting the vast majority of at bats against righties.  Thankfully does not seem to be affecting his glove work, meaning perhaps Farrell and others are giving him extra props for his defense in the dugout and in doing so helping to keep his head in this, which is good if that is the case.

 

[tims4wins]

He may be lost at the plate, but I thought he had a tremendous defensive game last night

 

[geoduck no quahog]

If any lineup in baseball can accommodate a weak bat at a key defensive position - it's the 2013 Red Sox.
 
Then again, we could have Kozma (allegedly one of the best defensive short stops in the NL), who is out-hitting Drew.
 
No thanks.

 

[smastroyin]

I thought he got a few good looks at pitches last night and against Scherzer in game 6 of the ALCS.  There are only a few games left and I don't know if it means anything, but he didn't look hopelessly lost like he did against the Tigers for the first 5 games.

 

[underhandtofirst]

Drew is now 4 for 42 including 2 infield singles and just 1 BB in 12 games.  He hasn't put more than 2 balls in play since Game 3 vs TB.
 
The lack of walks is troubling.  He's gone into cold spells, but he could always get on base.  Walks by month: 8,15,5,4(in 11 games),14,8.
 
With all that said, he looks very confident and smooth in the field.  I can't see any way they sit him at this point especially the way defense has played a huge part in the postseason.

 

[Koufax]

If sitting him down resulted in adding a bat like Nava's to the lineup, it might be worth doing.  But is Will Middlebrooks an upgrade at this point?  I have a hard time saying that it is, although Will couldn't be worse.

 

[JakeRae]

If sitting him down resulted in adding a bat like Nava's to the lineup, it might be worth doing.  But is Will Middlebrooks an upgrade at this point?  I have a hard time saying that it is, although Will couldn't be worse.

 
Will could be worse. We aren't comparing Middlebrooks' expected future performance to Drew's past performance. The valid comparison is to Drew's expected future performance. Middlebrooks hit .204/.244/.412 against RHP this year for a wRC+ of 69. His career line is .239/.273/.443 for an 88 wRC+. Drew, in contrast, has a .284/.377/.498 line against RHP this year, for a 137 wRC+ and a career line of .275/.343/.451 with a 105 wRC+.
 
Slumps are not predictive. Preferring the clearly inferior offensive and defensive player (who was also slumping mightily when he was last playing, in case people forgot) to the clearly superior offensive and defensive player simply because the latter has hit poorly for the past couple weeks is not rational. There is, admittedly, some chance that Middlebrooks would outperform Drew over the next 3-5 games if both were given the chance to play. But, it is far more likely that Drew would outperform Middlebrooks because Drew is a much better baseball player than Middlebrooks as long as a RHP is on the mound.

 

[dhellers]

I thought he got a few good looks at pitches last night and against Scherzer in game 6 of the ALCS.  There are only a few games left and I don't know if it means anything, but he didn't look hopelessly lost like he did against the Tigers for the first 5 games.

I saw that also.
 
As noted in other threads, with Ortiz at first, the upside of a good ss is increased -- since you won't be able to count on your firstbase man scooping up the bad throws
 
(does anyone remember the pre 84 team, when they got Buckner? This was a team with an atrocious fielding infield.  Buckner excellent ability to handle the bad throws turned it around: the infield went from utterly untrustworthy to merely not very good)

 

[Al Zarilla]

I saw that also.
 
As noted in other threads, with Ortiz at first, the upside of a good ss is increased -- since you won't be able to count on your firstbase man scooping up the bad throws
 
(does anyone remember the pre 84 team, when they got Buckner? This was a team with an atrocious fielding infield.  Buckner excellent ability to handle the bad throws turned it around: the infield went from utterly untrustworthy to merely not very good)

Ortiz is a lot better fielder than he is given credit for, including scooping the ball. Range? Can't imagine it's a lot worse than Napoli's.

 

[JimD]

As frustrated as I am with Drew's lack of hitting, it sure is gratifying to watch as every routine ball hit to short and most of the difficult ones are converted to outs.

 

[Savin Hillbilly]

Ortiz is a lot better fielder than he is given credit for, including scooping the ball. Range? Can't imagine it's a lot worse than Napoli's.

 
Napoli's range is average at worst. But the Sox have had excellent-to-extraordinary range at first base every year since 2006. We're spoiled. 
 
Papi handles most of what he can reach, and he's decent at scooping throws. But he's far less mobile around the bag than Napoli. It's a downgrade, but hopefully won't become a game-changer.

 

[Koufax]

Maybe Napoli can play shortstop. 

 

[Drek717]

 
Will could be worse. We aren't comparing Middlebrooks' expected future performance to Drew's past performance. The valid comparison is to Drew's expected future performance. Middlebrooks hit .204/.244/.412 against RHP this year for a wRC+ of 69. His career line is .239/.273/.443 for an 88 wRC+. Drew, in contrast, has a .284/.377/.498 line against RHP this year, for a 137 wRC+ and a career line of .275/.343/.451 with a 105 wRC+.
 
Slumps are not predictive. Preferring the clearly inferior offensive and defensive player (who was also slumping mightily when he was last playing, in case people forgot) to the clearly superior offensive and defensive player simply because the latter has hit poorly for the past couple weeks is not rational. There is, admittedly, some chance that Middlebrooks would outperform Drew over the next 3-5 games if both were given the chance to play. But, it is far more likely that Drew would outperform Middlebrooks because Drew is a much better baseball player than Middlebrooks as long as a RHP is on the mound.

You're leaving a pretty big home/road split in both their numbers out of the equation though.  WMB home against RHP had a 41 wRC+.  Drew for that same split was at 170 wRC+.  Their road numbers however were 91 for WMB and 100 for Drew.  A slight edge, but not something you'd hang your hat on to predict the next 3-5 games.
 
If we take it a step further and look at pitch types (getting into clearly less predictive data here) via Fangraphs' PitchF/x breakdowns per hitter for Drew and WMB respectively we can see that Drew for the past three years has a pretty legitimate negative indicator against sliders and is bad against change ups but mitigates that by doing quite well against fastballs.  WMB meanwhile has minor negatives against most pitches with a positive against cutters near twice those  negatives and a positive against curves equal to them and reinforced by his 2012 numbers (as is the cutter number).  The big change for him from 2012 to 2013 is that he couldn't hit fastballs as effectively, though given his better second half the hypothesis that he has overcome this isn't a bad starting point, unfortunately I can't find pitch splits by half for a player.
 
the upcoming Cardinals starters feature the following repertoires:
Joe Kelly:
FB - 67%
SL - 10.6%
CB - 8.3%
CH - 14.2%
 
Lance Lynn:
FB - 73.2%
SL - 12.8%
CB - 10.1%
CH - 4.0%
 
For the record here are the numbers for the two we've faced:
 
Wainright:
FB - 40.5%
CU - 28.5%
CB - 27.3%
CH - 3.8%
 
Wacha;
FB - 64.9%
CT - 2.7%
CB - 5.0%
CH - 27.4%
 
So just using these home/road and pitch histories I'd personally like to see the Sox give WMB the start tomorrow (as I don't think either he or Drew are a superior matchup v. Kelly) and then again in game 5 against Wainright, who's pitch arsenal falls pretty well in WMB's wheelhouse.  Then should the series make it back to Fenway you play Drew's incredibly strong Fenway splits and depending on how WMB did in St. Louis you potentially mix up he and Bogaerts.

 

[SaveBooFerriss]

Maybe Napoli can play shortstop. 

 
That is stupid.  Why downgrade two positions by playing Ortiz at 1b and Napoli at SS?  Just leave Napoli at 1b and play Big Papi at SS.  

 

[Al Zarilla]

 
That is stupid.  Why downgrade two positions by playing Ortiz at 1b and Napoli at SS?  Just leave Napoli at 1b and play Big Papi at SS.  

Because Papi might field a grounder and throw to third base instead of first.

 

[JakeRae]

You're leaving a pretty big home/road split in both their numbers out of the equation though.  WMB home against RHP had a 41 wRC+.  Drew for that same split was at 170 wRC+.  Their road numbers however were 91 for WMB and 100 for Drew.  A slight edge, but not something you'd hang your hat on to predict the next 3-5 games.
 
If we take it a step further and look at pitch types (getting into clearly less predictive data here) via Fangraphs' PitchF/x breakdowns per hitter for Drew and WMB respectively we can see that Drew for the past three years has a pretty legitimate negative indicator against sliders and is bad against change ups but mitigates that by doing quite well against fastballs.  WMB meanwhile has minor negatives against most pitches with a positive against cutters near twice those  negatives and a positive against curves equal to them and reinforced by his 2012 numbers (as is the cutter number).  The big change for him from 2012 to 2013 is that he couldn't hit fastballs as effectively, though given his better second half the hypothesis that he has overcome this isn't a bad starting point, unfortunately I can't find pitch splits by half for a player.
 
the upcoming Cardinals starters feature the following repertoires:
Joe Kelly:
FB - 67%
SL - 10.6%
CB - 8.3%
CH - 14.2%
 
Lance Lynn:
FB - 73.2%
SL - 12.8%
CB - 10.1%
CH - 4.0%
 
For the record here are the numbers for the two we've faced:
 
Wainright:
FB - 40.5%
CU - 28.5%
CB - 27.3%
CH - 3.8%
 
Wacha;
FB - 64.9%
CT - 2.7%
CB - 5.0%
CH - 27.4%
 
So just using these home/road and pitch histories I'd personally like to see the Sox give WMB the start tomorrow (as I don't think either he or Drew are a superior matchup v. Kelly) and then again in game 5 against Wainright, who's pitch arsenal falls pretty well in WMB's wheelhouse.  Then should the series make it back to Fenway you play Drew's incredibly strong Fenway splits and depending on how WMB did in St. Louis you potentially mix up he and Bogaerts.

The single season samples are small enough, as is, that I'm far more inclined to look at career performances. Also, the idea that the home/road split is as substantial as the numbers you've provided indicate doesn't pass the smell test. There is no good reason to think that Middlebrooks becomes an awful hitter at home or that Drew is only good in Fenway, particularly the latter given that his career also supports his just being better against RHP than Middlebrooks is. 
 
As for the idea that Middlebrooks should start tomorrow because you don't see a substantial matchup advantage for either. That is absurd. Middlebrooks requires a substantial offensive advantage to be the preferred starter. Drew provides significant defensive value over Middlebrooks. If the offensive considerations are close, Drew should be starting.

 

[Drek717]

The single season samples are small enough, as is, that I'm far more inclined to look at career performances. Also, the idea that the home/road split is as substantial as the numbers you've provided indicate doesn't pass the smell test. There is no good reason to think that Middlebrooks becomes an awful hitter at home or that Drew is only good in Fenway, particularly the latter given that his career also supports his just being better against RHP than Middlebrooks is. 
 
As for the idea that Middlebrooks should start tomorrow because you don't see a substantial matchup advantage for either. That is absurd. Middlebrooks requires a substantial offensive advantage to be the preferred starter. Drew provides significant defensive value over Middlebrooks. If the offensive considerations are close, Drew should be starting.

Drew has had a strong home/road split all season.  I'd say that's a large enough sample size.  Meanwhile WMB's career sample indicates the opposite.  Neither is a great sample size but we're talking about 3-5 games with Drew being abysmal so far, and it's not like we're working with 30 or 40 ABs.  They both have over 130 road ABs against RHP.  At some point you take a small bit of data and use it as the basis for a hypothesis you then have to test.
 
As for playing WMB tomorrow,  to me that isn't even about matchup.  That's about simply getting Drew a day off and letting WMB have a crack at doing something.
 
I also don't see the significant defensive value argument here.  WMB is a better 3B than Bogaerts.  We have no evidence to prove that Drew is a better fielder than Bogaerts at SS.  Therefore we can't really put the math together on the defensive plus/minus.
 
The whole "you make moves based on what you expect them to do, and for that we should look at career/season" splits might work great building a roster for the second half of a season, but this is now a best of five series.  The variability in the numbers VERY OFTEN does not normalize within that small a sample, and Drew doesn't look like he's on the verge of breaking out. At some point you have to make a judgement call based on the full extent of the symptoms, not just looking at the numbers and betting that they're due to change, and hat time is with a  maximum of five games left in the season.  He isn't even making legitimate contact or drawing walks.  Those aren't the signs you see from a guy just about to bust out of his slump.  Meanwhile WMB has had time to work on some things, has seen hot streaks that suggest a correlation with pitcher unfamiliarity being a real benefit to him, and has the ability to change a game when he dumb luck runs into one.  So instead of taking the guy we've seen struggle outside Fenway all season and have seen struggle all playoffs and giving him more starts why not go with something with some greater variability?  The defensive risk isn't that great and the offensive downside is literally no worse than what we're getting now from Drew.

 

[dbn]

Because Papi might field a grounder and throw to third base instead of first.

 
The old "no triples" play when leading late in the game.

 

[KillerBs]

I don't agree that because there are only a few games left, you throw out the career/season numbers and rely more heavily on recent performance. Yes of course there is more variability in results over the next 5 games, then over a longer period, and obviously, given the extreme importance of the games, we should be starting over the next 5 days, the guys who are considered to be the best players RIGHT NOW, but it does not follow (at least to me) that you necessarily look to a smaller sample size than usual in trying to determine (imperfectly for sure) who those guys are.
 
A quick check of bbref shows Drew has had at least 2 slumps of this magnitude already this year, one in May (5-40) and the other in June (6-44, with 18 Ks). In both case, he came out of it, without warning, going 12-33, .667 SLG after the May skid and 5/9 and then 11/33 after the June slump. This has no predictive import of course, but I think it illustrates the point. If SD goes 5/9 in the next 2 games we probably win the WS.
 
To pluck another example of many, there was a majority opinion, at least, on this board in July/August (understandably, by the way) that Napoli could simply no longer hit big league pitching. Mike Carp simply must play against RHP, it was said.
 
Reaching a bit further back in time, I think it of real importance that Drew is an 837 OPS guy in the regular season since July 20. This continues to give me comfort amid the whiffs etc. Isn't it likely that whatever has changed in Drew's abilities as a hitter since the regular season is wholly ephemerable, unknowable and therefore subject to immediate reversal.
 
Middlebrooks recent performance also leaves a lot to be desired too. Yes, he may "get lucky and run into one" but that doesn't strike me as very convincing argument. There is too much waiting around for someone at the bottom of the order to get lucky and run into one for my liking at this stage.
 
Drew also been a veritable rock defensively. My sense is that there is a significant defensive drop off if Drew is benched, with there being no real basis to be confident that WMB is a better 3b defensively than XB, while the SD:XB defensive drop off is apparent and marked.  
 
I agree he looks terrible up there now and if the staff concludes that WMB/XB is a better combo than SD/XB or SD/WMB, then I am inclined to defer to that on the premise that they know a hell of a lot more than I do about what is up with Steven Drew. But based on all the info that appears available, I could not advocate for this.
 
Now, if we are behind after 5,6,7 then unless there are signs of life in his bat, and in any event against a LHPer, I think it obvious they should be PH for him. 

 

[Bigpupp]

Maybe Napoli can play shortstop. 

Frank Thomas started a game at short towards the end of his career.* If he can do it anyone can!



*He hit 3rd in the top of the first and was subbed out in the bottom of the inning.

 

[DennyDoyle'sBoil]

It's 1-1 in the World Series.  I don't like how all this got so complicated based on such small sample sizes to begin with.  It's down to the last five games of the year, and if we are to go down, I want to go down with our best over the long haul.  
 
We won 97 games.  I want Drew at SS.  I want WMB at 3B.  And I want Nava in LF against RHP.  Enough of the hot hand stuff.  Xander is our PH for Drew against lefties.  Gomes and Nap (when he doesn't start) from the right too, in the right situation where needed.  I want Stephen Drew in the field as our SS.  I want WMB at the position, 3B, he's played 16 times as many starts in the big leagues as Bogaerts.  I think that getting down on Drew for his recent woes at the plate and getting too high on Bogaerts for his recent success is doubling down on small sample sizes.  
 
Bogaerts' performance so far in the playoffs is stunning.  He seems beyond his years.  It gives me warm fuzzies about the future.  That he has gotten on base six times against RHP in the playoffs -- and good to great RHP at that -- is remarkable.  There are some reasons to think it was lightening in a bottle.  He was .133/.188/.276 against RHP this year.  He's 0-7 so far in the WS with 4 Ks.  He also is inexperienced and has a downside -- he failed to score from second base on a double and he had no idea what to do on the pickle play with Fielder.  In both cases his teammates picked him up.  But if Sanchez catches the soft liner from Ellsbury, that's an expensive run.  WMB is the better baserunner, by all accounts, and after seeing how the players on both sides are gripping the ball tighter and how a single mistake can cost the game -- indeed the season -- I want the guy who has experience at the position in there.
 
Drew's woes at the plate are infuriating.  But he has a career of demonstrated ability to hit RHP over nearly 3k plate appearances.  It's nut cutting time.  Drew is our short stop.  WMB is tough to watch too, but he has had a few extra base hits and walks this postseason.  And as frustrating as his at bats have seemed at times, are we putting too much emphasis on the strike out?  Yes, he's had a lot of them and Ks always make you wince -- but then again, Bogaert's K/PA rate in the playoffs isn't that different.  (WMB has 9 Ks in 26 ABs, and Bogaerts has 5 in 18.)  WMB's OBP against RHP for the year is .060 better than Bogaerts and his slugging is .140 better.  It's more than just hoping he runs into one.  He may not be great, but on balance, I think he's the steadiest option in a series where I really feel like our steady team gives us the best chance to win.

 

[radsoxfan]

The single season samples are small enough, as is, that I'm far more inclined to look at career performances. Also, the idea that the home/road split is as substantial as the numbers you've provided indicate doesn't pass the smell test. There is no good reason to think that Middlebrooks becomes an awful hitter at home or that Drew is only good in Fenway, particularly the latter given that his career also supports his just being better against RHP than Middlebrooks is. 
 
As for the idea that Middlebrooks should start tomorrow because you don't see a substantial matchup advantage for either. That is absurd. Middlebrooks requires a substantial offensive advantage to be the preferred starter. Drew provides significant defensive value over Middlebrooks. If the offensive considerations are close, Drew should be starting.

 
Agree completely. The one year home/road stuff is most likely flukey in Drew's case.  There are hundreds and hundreds of guys in the majors in any one year, and any random distribution is going to have some guys have big splits.  Drew just happens to be one of those guys this year.  I don't think he has some particular skill-set or swing thats made only for Fenway, but doomed to failure on the road. Same randomness likely applies to Middlebrooks the other direction (plus his career home/road is very similar).  This is almost surely looking for causation when one doesn't exist.
 
Drew has been tough to watch at the plate, no doubt.  If the Cards had a lefty starter, I think there would be a good argument for a Xander/WMB combo on the left side of the infield.  But the Cards don't.  WMB has been bad recently, has bad overall season stats, and has been particularly horrific against righties.  He has power, so there is always the chance he could get lucky and run into one.  But I still find nothing close to a compelling reason to replace Drew with WMB against a right handed starter. 
 
 


 

It's 1-1 in the World Series.  I don't like how all this got so complicated based on such small sample sizes to begin with.  It's down to the last five games of the year, and if we are to go down, I want to go down with our best over the long haul.  
 
We won 97 games.  I want Drew at SS.  I want WMB at 3B.  And I want Nava in LF against RHP.  Enough of the hot hand stuff.  Xander is our PH for Drew against lefties.  Gomes and Nap (when he doesn't start) from the right too, in the right situation where needed.  I want Stephen Drew in the field as our SS.  I want WMB at the position, 3B, he's played 16 times as many starts in the big leagues as Bogaerts.  I think that getting down on Drew for his recent woes at the plate and getting too high on Bogaerts for his recent success is doubling down on small sample sizes.  
 

 
As to the argument that WBM has been the 3B all year so he should stay the 3B in the playoffs....
 
First of all, he has sucked for the majority of the year, and was sent to the minors for a good chunk of it, and is particularly horrible against right handed pitching. The Red Sox won despite Middlebrooks, not because of him.
 
Also, are you really comparing OBP and SLG vs RHP between WMB and Xander when Xander had 29 at bats in the majors against RHP?  He isnt a 1.089 OPS hitter vs LHP just like he isn't a .464 guy vs RHP as his regular season numbers would suggest. That is a useless sample. You say you don't want to double down on Xander's small positive sample in the ALCS, but then you quoted numerous meaningless tiny sample sizes.  If you really care about tiny samples, it's worth nothing that even though he is 0-7 in the World Series, he has hit 3 line drives.  
 
Even ignoring age, Bogaerts has outperformed Middlebrooks in the minors, and I think he is a better offensive player at this moment.  Unless you think his unfamiliarity with 3B is a reason to bench him (I think he's been fine so far), Xander is the obvious play.

 

[DennyDoyle'sBoil]

 
 
Also, are you really comparing OBP and SLG vs RHP between WMB and Xander when Xander had 29 at bats in the majors against RHP?  

 
No, I was trying to make two different points (though agree it's not too clear) -- the first was to talk about what to think about Bogaerts' amazing performance against RHP in the playoffs so far and the second was to try to make my case for WMB.  The attempt was not to use the small number of ABs against RHP in the regular year to compare it to WMB's abilities against RHP (such as they are), but to compare it to Bogaerts' own performance over a similar number of ABs in the playoffs.   It is that if we're going to try to predict what Bogaerts will do at the plate in the next 5 games against RHP, I see as many reasons to believe it will be poor as I do to believe it will be good.  Given WMB's relative experience advantage, the fact that he has played the position for nearly a full season of games, and the fact that they bat from the same side of the plate, I would prefer to see WMB get the starts and to have Xander as a PH for Drew against lefties, which I think may well be his highest and best use.

 

[radsoxfan]

I agree neither Xander's playoff numbers nor his tiny sample at the end of year are very predictive.
 
Thats why I think its safer to look at season and career numbers.  The numbers have already been posted, so I wont repeat them.  But Xander has been a better minor league player throughout his career than Middlebrooks and I think he is a better offensive player right now. Obviously when you account for age its no contest, but even throwing that out, I think its reasonable to conclude Xander is currently a better hitter.  Add that to the fact that WMB REALLY struggles vs. right handers (.244 OBP) this season, and I don't think there is a good rationale to play Middlebrooks if Xander can hold his own on D.
 
I also think WMB can PH for Drew against lefties just as easily as Bogaerts can.  You still end up with the same defensive alignment. I'd rather give Xander 4 ABs and WMB 1 AB than the other way around.

 

[Rasputin]

It's probably been mentioned - tho I missed it if it did - is at the very least Drew should be able to hopefully draw a few walks as the presumably pitch around him a bit in the 9 hole. Those may not directly lead to many runs but hey, at least his obp will go up a bit.

I almost wonder if it Mike make sense to hit Drew 7th and Bogaerts 8th because Boagaerts is more likely to send a fastball a long way.

 

[Mugsy's Jock]

Unless (God willing) some drastic turnaround unfolds, is Drew playing himself out of a Quakifying Offer (and, in the process, costing the Sox a #1 draft pick)?

Put another way, if you're another GM, would you still give Drew a Multi-year offer that beats the one year $14 million or so the Sox offer would be?

 

[moondog80]

Put another way, if you're another GM, would you still give Drew a Multi-year offer that beats the one year $14 million or so the Sox offer would be?


A guy whose ERA+ the past two years was 76 and 68 just got 35 mil for two years. So yeah , I could see giving Drew 28 mil. Lots of people have bad postseasons.

 

[smastroyin]

I don't know that there is a lot of indication that teams price for post season performance anyway.

 

[radsoxfan]

I don't know that there is a lot of indication that teams price for post season performance anyway.

 
Personally I don't think he should be punished much for it, and I wouldn't project his performance over the next few seasons that much differently based on a very poor 43 PAs. But I would be curious if there is any study on how teams tend to view it.  Human nature is probably to overvalue very recent performance, and I'm sure all FO's are watching these games. In college basketball for example, there are studies about how performance in the NCAA tournament can cause a relatively big change on a player's draft status compared to their pre-tourney projection (though it's unclear if this shift is too small or too big).
 
Drew's season line is .253/.333/.443/.777 and pretty much in-line with his career numbers. His playoff line is a putrid .095/.116/.143/.259, but its just really bad 43 PA stretch than can happen to almost anyone. I wouldn't argue with adding it to his season numbers (though post season pitching is tougher than average), which would make him a .240/.317/.417/.734 player this season overall.   Not sure if that small drop in a single season performance would really make a big difference in his FA contract though.

 

[Dogman]

I almost wonder if it Mike make sense to hit Drew 7th and Bogaerts 8th because Boagaerts is more likely to send a fastball a long way.

 
 
The other side of the coin is that hitting X in the 8th spot may very well be moot given who is hitting behind him.  I think we see the 8th spot pitched around in the early innings. Based on X's baserunning blunders this postseason, I'd rather have Drew hitting 8th to minimize the potential baserunning gaffe's.  

 

[radsoxfan]

 
 
The other side of the coin is that hitting X in the 8th spot may very well be moot given who is hitting behind him.  I think we see the 8th spot pitched around in the early innings. Based on X's baserunning blunders this postseason, I'd rather have Drew hitting 8th to minimize the potential baserunning gaffe's.  

 
I might just be forgetting, but has he made more than one?

 

[Dogman]

Now that I think about it, no. 

 

[Mighty Joe Young]

The other side of the coin is that hitting X in the 8th spot may very well be moot given who is hitting behind him. I think we see the 8th spot pitched around in the early innings. Based on X's baserunning blunders this postseason, I'd rather have Drew hitting 8th to minimize the potential baserunning gaffe's.


This is making the (at this stage) rather unrealistic assumption that Drew will actually be on base. Clearing the pitcher is a big deal in NL baseball. Unfortunately hitting Drew 7th or 8th really doesn't make much difference. Actually , hitting him 8th might entice Metheny into actually pitching to him in a situation where, otherwise, he would be walked.