I’ve run my defensive projections for 2011 using UZR as the primary input data source.  For a general feel for the methodology read this post.  A couple changes from last year

  1. Instead of using the average FSR for bins of players I’m using the FSR converted to runs as found on Fangraphs as my scouting component number.
  2. I took into account multiple positions this year.  I used the Fangraphs standard position adjustments to convert all UZR values from all positions to a player’s primary position.

As mentioned, this set of data is run with UZR as the primary data input.  Over the next few weeks/month I’ll be running the same process with DRS as an input, TZ as an input, and an average as an input.  I’ll post those files as they become available.

Click here for an excel download
Click here for the google docs version

Quick Edit:  Didn’t explicitly say this anywhere.  The numbers in the files are Runs/150 Games.

Somehow I missed on Monday that Dan at Baseball Think Factory posted his Cardinals Zips projections.  Here are the projections converted to wOBA first for the MLBers (all calculations use the PAs in Zips and a 0.335 league average for RAA calcs)

Player wOBA RAA
Albert Pujols 0.434 55
Matt Holliday 0.376 23
Lance Berkman 0.368 14
Colby Rasmus 0.344 5
Allen Craig 0.337 1
Jon Jay 0.326 -5
David Freese 0.318 -6
Yadier Molina 0.315 -9
Skip Schumaker 0.311 -11
Tyler Greene 0.297 -17
Ryan Theriot 0.294 -23
Gerald Laird 0.284 -16

and then for the MiLBers

Player wOBA RAA
Nick Stavinoha 0.315 -7
Daniel Descalso 0.317 -10
Mark Hamilton 0.320 -5
Matt Carpenter 0.318 -8
Bryan Anderson 0.311 -7
Aaron Luna 0.306 -11
Andrew Brown 0.300 -13
Daryl Jones 0.297 -17
Thomas Pham 0.301 -14
Adron Chambers 0.291 -18
Steve Hill 0.292 -17
Tony Cruz 0.291 -18
Pete Kozma 0.285 -28
Donovan Solano 0.276 -25


These tables highlight the folly of the Brendan Ryan for Ryan Theriot swap. Theriot projects to only hit slightly better than Pete Kozma. Let that sink in for a while.  They also show the extreme lack of middle infield options, with Descalso posting the best wOBA projection.


The projection for Berkman is alright, but it is likely not good enough to overcome the defensive deficiencies.  On the whole the offensive projections do not inspire a whole lot of optimism.


On the positive side, the projection thinks that if David Freese were to get injured again, Matt Carpenter could step right in and be a suitable replacement.  As I’ve mentioned previously, that’s a sentiment I share.  Bryan Anderson would be a suitable backup with that line, with some potential to be more.  With the versatility of some of the bench options (Greene, Craig if he can backup at 3rd) would it be worth it to carry 3 catchers?  Probably not, but it’s something to think about.

Over in the comments at VEB I posted a notional chart that attempts to illustrate the “there’s no such thing as a pitching prospect” adage.  The basic idea behind the chart is that pitching prospects will have a lot more variance in their projection that would hitters.  This additional variance leads to a lower mean projection if you want to assume that two players have the same peak and same peak probabilities (i.e. both have the potential to be 6 WAR players).  Here’s the chart I used to illustrate my point (the values are completely notional)

In this case the red is the position player and the blue is the pitcher.

Before the season, CHONE projected the Cardinals were on course to win 91 games and enjoy a cake walk in a relatively easy division.  Someone will run off with this and say, “see, this is why projections are worthless”.  But before you go off an anti-metrics rant, let’s remember what projections are. In a nutshell, among other things, projections  use regression to the mean, age adjustments and weighted averages to derive their results. Projections are “50th percentile” projections; there are always players who are exceptions to the rule due to good or bad luck, or some sort of overhaul to their swing mechanics (hello, Jose Bautista!) or injury. It’s not divination.

It’s common for a team to have several players under or overshoot their projection. It’s just that in the Cardinals case, the under achievers have been particularly damning to the team. For this post, let’s just take a look at the hitters. The cut off is 100 plate appearances. I took the players individual CHONE projections and then adjusted them for their plate appearances and not their projected PA’s. Then I did a little color scaling for your eyes, because I’m a nerd like that.

  • The infield trio of death of Brendan Ryan, Skip Schumaker and Felipe Lopez has really hurt this team. All three were not expected to be some sort of offensive force behind Pujols-Holliday-Rasmus, but neither were they expected to hit for sub-.300 wOBAs, either. (Schumaker is at .301, being fair). I thought Flip was a good Freese-insurance signing, and instead he pooped the bed. Ryan and Schumaker picked a bad time to do the same thing.
  • Molina and Pujols also both were one WAR apiece worse than projected. Molina’s bat backslid after two good seasons, and Pujols has been a little off from his normal numinous standard. And yet he’s still an MVP candidate.
  • The sheer waste of roster space of the likes of Feliz, Winn, Miles and Stavinoha has been pretty frustrating to watch.
  • The most pleasant surprise has been Matt Holliday, who is doing his best to show that he earned the ginormous contract he got this past winter.

All told, the Cardinal’s hitters are five wins worse than we would have expected. Combine that with losing Brad Penny for the year, and Kyle Lohse being hurt and then horrible, you have a recipe for a pretty disappointing year. It’s really not that hard to figure out.

There’s been a decent amount of discussion regarding Brendan Ryan’s offense and his early season slump.  With that in mind I wanted to frame the discussion a little bit on what would be required for him to be an average MLBer (2 WAR) or do what he did last year (3 WAR).  For the playing time portions of WAR I assumed 625 PAs which would be abut 150 games started.  With that in mind, here’s a visual representation

The interpretation is simply that the lines give pairs of wOBA and defensive runs saved above average (DRAA, think his UZR value) that correspond to 2 and 3 WAR. For example, if Boog were to have a 0.290 wOBA (think Joe Thurston) he’d need to save a little over ten runs (13) to be average and 23 to be 3 WAR. Raw data can be found here if you want the exact numbers.

Continue reading »

I’m a little late to the party here.  I started this post the night the Mauer signing was announced, but slacked on finishing it.  Anyhow, here’s what I came up with.  Joe Mauer’s deal was formally announced at 8 yrs $184M.  The immediate question that comes to mind for Cardinals fans is, “How does this deal affect the pending Pujol’s negotiations?”  We can’t guess at that until we make an attempt to compare the players over the lifetime of such deals.   First what could Mauer look like during the entirety of his deal?  The following tables attempt to give insight into that question.  First if he can remain at C during the entire deal

Year Age Pos Pos adj rep def off War
2011 28 C 7 21 6.5 36.0 7.1
2012 29 C 7 21 5.5 35.3 6.9
2013 30 C 7 21 4.5 35.3 6.8
2014 31 C 7 21 3.5 34.4 6.6
2015 32 C 7 21 2.5 33.8 6.4
2016 33 C 7 21 1.5 31.9 6.1
2017 34 C 7 21 0.5 28.3 5.7
2018 35 C 7 21 -0.5 24.2 5.2

Then if he has to switch to 1B at the end of the deal

Year Age Pos Pos adj rep def off War
2011 28 C 7 21 6.5 36.0 7.1
2012 29 C 7 21 5.5 35.3 6.9
2013 30 C 7 21 4.5 35.3 6.8
2014 31 C 7 21 3.5 34.4 6.6
2015 32 C 7 21 2.5 33.8 6.4
2016 33 1B -11 21.5 2 34.6 4.7
2017 34 1B -11 21.5 1 30.7 4.2
2018 35 1B -11 21.5 0 26.3 3.7

And a similar table constructed for Albert

Year Age Pos Pos adj rep def off War
2011 31 1B -11 21.5 5 60.0 7.6
2012 32 1B -11 21.5 4 59.4 7.4
2013 33 1B -11 21.5 3 57.3 7.1
2014 34 1B -11 21.5 2 53.4 6.6
2015 35 1B -11 21.5 1 49.0 6.0
2016 36 1B -11 21.5 0 45.2 5.6
2017 37 1B -11 21.5 -1 39.7 4.9
2018 38 1B -11 21.5 -2 35.4 4.4

You may be asking “Where did you come up with those tables?”  Which would be a valid question.  I used CHONE to get a feel for the offensive production levels, CHONE for catcher D, my own projections for Albert’s D, and the JC Bradbury criteria (read as star player with slow aging) aging curve that MGL developed using the delta method.

Now that we have all of the assumptions out of the way, let’s talk about the actual impact.  The Mauer deal sets the market at somewhere between $3.6M and $4M (unadjusted for inflation, but all numbers will be unadjusted) per projected WAR depending on which of the above scenarios you see as more likely.  Translating those numbers over to the Pujols deal yields an 8 year deal between $180M and $197M, so I’d say 8 yrs $190M would be about “right”.  If we back up to a 6 year deal that would work out to $145M to $160M over 6 years, although the $/win for Mauer was driven up slightly by the length of deal.  If we limit the Mauer deal to 6 yrs then the price for Pujols translates to $140M to $145M over 6 years.  Most of these parameters work out to somewhere in the $25M/year range, so that “feels” about right.

I’m warm to the idea of putting McClellan in the rotation, and I’ll explain why I like it. First of all, McClellan does seem to have the repertoire of a starting pitcher. Here’s a look at some of his Pitch F/x data proving he has enough weapons to succeed as a starter. The data comes courtesy of TexasLeaguers.com.

Type Count Selection Velocity (mph) Vertical (in) Horizontal (in) Spin Angle (deg) Spin Rate (rpm)
FF 483 44.40% 91.4 7.2 -7.79 227 2,136
CU 278 25.50% 75.8 -8.02 6.53 40 1,688
SL 153 14.00% 87.7 3.43 0.72 172 755
SI 95 8.70% 90.8 6.54 -9.15 235 2,220
FC 60 5.50% 88.5 5.3 0.19 180 1,090
CH 20 1.80% 84.6 5.54 -7.61 233 1,769

So what does this prove, exactly? Well, first of all it proves that I like making tables even though I stink at formatting them, that much you already knew. Getting on point…in order to succeed as a big league starter, there are some ingredients you must have, unless you’re a freak. Those ingredients are at least one “plus” pitch, two average pitches and average command. Looking at this chart, Mac has the pitches. And we’ve all seen him pitch dozens of times, I think our eyes tell us he has the goods.  (Some quick clarification  - sinker/fastball, same thing. Bad Pitch F/x algorithm! Bad! Same goes for his cutter/slider).

Anywho, his two-seam fastball is average. He doesn’t generate tons of sink, but the pitch has good “tail”. His cutter/slider and his curveball can both be very good pitches at times.

Let’s look at his results -

Type Strike Swing Whiff Foul In Play
FF 65.20% 45.30% 3.50% 21.50% 20.30%
CU 48.60% 32.00% 10.40% 10.40% 11.20%
SL 56.20% 45.10% 9.80% 17.00% 18.30%
SI 61.10% 41.10% 6.30% 14.70% 20.00%
FC 66.70% 50.00% 5.00% 20.00% 25.00%
CH 45.00% 35.00% 10.00% 5.00% 20.00%

A fair share of whiffs on the curve and slider. His command is fair enough, although he did walk a few too many hitters last year.

So I hope I’ve established that he has the pitches to start, how exactly would he do? Sean Smith did a study on pitchers from 1953 up to 2008, and found that when switching from starting to relief and vice-versa, a pitcher’s walk rate would stay static, while their hits went +/- 5%, their homers went +/- 15%, and their strikeouts went +/- 16%. Let’s apply those numbers to McClellan and see what we come up with. First, here’s his 50th percentile CHONE projection as a reliever:

McClellan 63 5 25 2 48 3.99

Now let’s see what we come up with for 150 innings for McClellan as a starter, take it for what it’s worth -

McClellan 150 14 61 5 102 4.42

150 innings, 4.42 FIP is a 1.6 WAR pitcher, which is a little over 3 times higher than what his projected WAR would be coming out of the bullpen with an average leverage index of 1.3; in other words him pitching as the primary set-up man. That’s Nick Blackburn/Jon Garland territory, which is serviceable. Let’s put it this way – if the Cardinals had Nick Blackburn, would you prefer they started him or put him in the bullpen? You’d want them to start him, of course.

But wait, to who does Mac’s innings as a set-up man fall to? Yo-yo and unproven arms like Motte, Hawksworth and Boggs pitching in the 8th inning is a scary proposition. Let’s just say for now that McClellan’s innings would fall to Jason Motte. His CHONE projection calls for a 4.4 FIP. The bad news is no non-LOOGY reliever is very likely to do better. And one of those pitchers would be taking Motte’s spot, and so forth. If you give his innings to Motte, that’s a loss of 0.4 WAR, and an increase in sales of garden tools…I mean angry mob supplies in the greater St. Louis area.

Let’s not also forget Jaime Garcia. He’s not projected to fare as well (4.69 FIP), but as a 5th starter, that’s fine and it’s feasible he plays better than projected. Garcia also is a talented arm. With the way the bullpen is set-up now, given the chaining, putting McClellan in the bullpen or in the rotation ends up being closer to a wash than I would have imagined, and that’s assuming he’d succeed according to the numbers I laid out. So is it worth it?

Speaking from a long-term perspective, I’d say heck yes. It would be more beneficial for the Cardinals to have a nice, cost-controlled pitcher in their rotation than one in their bullpen. If the Cardinals think Mac is even close to being the real deal, they need to upgrade the bullpen with someone available for cheap like Kiko Calero or (I can’t believe I’m about to type this) Chan Ho Park. It would be a worthy investment to develop a nice, young starter while saving your team some unnecessary angst in the late innings.  And hey, if he bombs as a starter, there’s no harm in having some extra depth in the bullpen, anyways.

Sign Kiko. Give Mac a long look.

CHONE projects that the Cardinals are on course to win big an a weak division.

NL Central Wins Losses
St. Louis Cardinals 91 71
Milwaukee Brewers 81 81
Cincinnati Reds 81 81
Chicago Cubs 79 83
Pittsburgh Pirates 74 88
Houston Astros 73 89

Sean’s comments -

Thanks to resigning Matt Holliday, the St Louis Pujols should have a comfortable season in the NL Central. Most of the teams in that division don’t spend enough money to compete. Then there are the Cubs and Astros, who have spent on some real albatrosses. I like the young talent in Cincinnati, but they probably aren’t ready to challenge Prince Albert just yet.

Considering the current talent, and barring any moves on the part of the competition, it looks like a walk in the park. A lot could change. The Reds and Brewers have the prospects to make moves, if they’re competing I’d look for them to be in the thick of a blockbuster. The Cub’s farm system is also much improved, but they have so little in the way of payroll flexibility with so many commitments.

CAIRO’s projected standings are updated, too. For what it’s worth, the Cubs are the favorite for the wild card. I think people, myself included, have overlooked the Cubs a bit. Aramis Ramirez will be healthy again, Geovany Soto should bounce back, Alfonso Soriano can’t possibly be worse than he was in 2009 (can he?) and their rotation should be very good. Their bullpen looks a bit shaky to me. Still, they are returning much of the same talent that won 97 games just a couple of seasons ago.

NL Central W L RF RA Div WC StD W Std RF Std RA
STL 90.5 71.5 717 638 56.5 9.0 84-97 677-757 602-675
CHN 86.2 75.8 729 679 24.5 17.3 80-92 697-760 645-713
CIN 84.2 77.8 731 705 12.0 15.3 78-90 696-766 669-741
MIL 80.9 81.1 736 738 6.5 5.0 74-88 697-774 697-779
PIT 72.6 89.4 671 745 0.0 0.0 65-80 635-706 703-787
HOU 66.1 95.9 650 796 0.5 0.0 60-72 618-682 756-836

W: Average wins over 100 iterations
L: Average losses over 100 iterations
RF: Average runs for/scored over 100 iterations
RA: Average runs allowed over 100 iterations
Div: # of times team won division over 100 iterations
WC: # of times team won wild card over 100 iterations
Std W, RF, RA: W, RF and RA within one standard deviation in either direction

It’s time to look at another “too-early” projected NL Central standings, this time we’re looking at CAIRO, named after the lovable utility infielder you and I are all too familiar with. While more changes are going to be made before now and the season, CAIRO was the best crystal ball when it came to picking the standings last season.

TM W L RS RA Div WC W+/- RS+/- RA+/-
Cardinals 92 70 762 676 42.50% 9.70% 1.1 32 36
Reds 86 77 755 721 22.20% 9.00% 7.5 82 -2
Cubs 84 78 749 726 17.30% 9.20% 0.7 42 54
Brewers 81 81 762 769 13.70% 7.50% 1.3 -23 -49
Pirates 70 92 730 839 1.90% 1.70% 8.2 94 71
Astros 69 93 663 771 2.50% 2.30% -4.9 20 1

W: Projected 2010 wins
L: Projected 2010 losses
RS: Projected 2010 runs scored
RA: Projected 2010 runs allowed
Div: Division win percentage
WC: Wild card win percentage
W+/-: 2010 projected wins minus 2009 actual wins
RS+/-: 2010 projected runs scored minus 2009 actual runs scored (positive means they are projected to score more)
RA+/-: 2010 projected runs allowed minus 2009 actual runs allowed (negative means they are projected to allow fewer)

Beware the Red Menace? I don’t have access to CAIRO’s individual projections, but we can make some educated guesses about why it projects the Reds to improve by 82 runs on offense next year.

  • Joey Votto will be awesome again.
  • Jay Bruce should improve. He was a -2 offensively last year, CHONE projects him for a +18 in 2010.
  • Willy Taveras was 27 runs below average last year with the bat, most of his playing time will be given to Drew Stubbs, who can’t possibly be worse.
  • Chris Dickerson should get most of the playing time in LF, and not Laynce Nix.
  • Scott Rolen is a nice improvement over Edwin Encarnacion.

Well shut my mouth. Neyer’s got a take on this as well.

As for the Cardinals, the offense will be better with Holliday (duh), but the pitching will likely regress to the mean.  Yep, it’s just hard to improve upon a season where your 1-2 starters finish 2nd and 3rd in the Cy Young voting. And Penny in 2010 isn’t likely to match Pineiro’s 2009, but neither is Pineiro for that matter. Sorry, I’m telling you stuff you already know.

As for the Cubs, they’re hanging in there. Soriano and Soto can’t possibly be worse, Ramirez will be healthy, Byrd in 2010 ought to be better than Bradley was in ’09. On the other hand, Randy Wells should fall back to earth some, and Rich Harden’s spot in the rotation is going to be replaced by Tom Gorzellany or *gasp* Carlos Silva.

Barring a disaster, this season *should be* a walk in the park for the Cardinals.

Season’s over, we win.

Little work has been done on the depth charts, I’m told. And there are some free agents to be had out there still. But for what it’s worth, here’s the NL Central.

89	73	802	724	St. Louis Cardinals
82	80	780	772	Cincinnati Reds
77	85	779	823	Chicago Cubs
75	87	766	829	Houston Astros
75	87	827	895	Milwaukee Brewers
70	92	754	873	Pittsburgh Pirates

Some of the other divisions look pretty strange to me – like the NL East, for instance, but this jives OK with my pathetically unsophisticated Win Talent WAR estimates OK enough. The only real shocker to me is the Brewers, who I think is a better team on paper than PECOTA does, apparently.

The ZiPS are out too. Here’s the ZiPS, wOBA-fied for hitters and FIP-ified for the pitchers:

Continue reading »

© 2011 Gas House Graphs Suffusion theme by Sayontan Sinha