Projection Connection
We interrupt, again, this programming to bring you the next set of Cardinals projections—ZiPS, come on down!
I follow ZiPS really enthusiastically every year because Dan Szymborski, their creator, seems to be working on them constantly. This year he's added a new way of formatting; they're ordered by the way a player's bat plays at each position, with players like Adam Kennedy appearing four times in different places. (Take a look at that average catcher the Cardinals apparently have—is Steven Hill the next Mike Jacobs, or what?)
There are some weird quirks to this year's ZiPS, as there are with any projections (in this case, D'Angelo Jimenez, who spent all of last year at Memphis, is predicted to be our middle infield savior), but what's more interesting to me are the narratives it continues from our look at the Bill James numbers.
Narrative One: Ryan Ludwick is actually really good. Nobody's going to predict him to continue to be the fourth best hitter in the National League, or slug .600 over 500 more at-bats, but so far both projections have predicted continued success for Thudwick.
| name | G | AB | H | 2B | 3B | HR | RBI | BB | K | AVG | OBP | SLG | OPS |
| James | 152 | 540 | 152 | 38 | 1 | 31 | 100 | 54 | 135 | .281 | .351 | .528 | 879 |
| ZiPS | 137 | 456 | 125 | 31 | 2 | 26 | 85 | 46 | 123 | .274 | .347 | .522 | 869 |
(Incidentally, the James projections lived up to their reputation for generosity; just about every player covered by both teams is noticeably better in their Jamesian form.) For a player like Ludwick, whose fluke season and late start to his big league career would seem to make projecting his future difficult, these are remarkably similar. And while it's a regression from 2008, I don't think the Cardinals would be at all concerned about taking it.
Narrative Two: Yadier Molina isn't a .300 hitter. I'll watch PECOTA and the other projectors—and our community projections, for that matter—with the most interest when they come around to Yadi, who has made quantum leaps in his offensive output two years running since his miserable 2006 campaign. Last year, his .300 batting average bringing his slugging percentage almost to .400, he had the Ozzie Smith skill-set: great defense and offense that, in its near-averageness, was a major asset at a position with low offensive standards. But even though he's 26 neither James nor ZiPS has any hope of it being a new talent level.
| name | G | AB | H | 2B | 3B | HR | RBI | BB | K | AVG | OBP | SLG | OPS |
| James | 126 | 473 | 129 | 22 | 0 | 8 | 60 | 37 | 40 | .273 | .329 | .370 | 699 |
| ZiPS | 122 | 414 | 111 | 20 | 0 | 7 | 45 | 33 | 39 | .268 | .324 | .359 | 683 |
This sort of regression knocks him just below where he was in his abbreviated 2007 campaign. (Incidentally, it's interesting how close offense-first Bryan Anderson is to Molina according to ZiPS; I like Anderson as a prospect, but unless he's got Ichirovian bat control he'll have to develop some power if he is to separate himself from all of the Yadi-types—great defense, balding, can hit a little—in the league.
Narrative Three: The Rotation is Anybody's Guess. If there were no such thing as injury concerns the Cardinals' ZiPS rotation would be wonderful. They sign Braden Looper and then—voila.
pitchers
GS
IP
K
BB
HR
ERA
Carpenter
11
62
48
15
6
3.48
Wainwright
26
165
110
49
12
3.71
Wellemeyer
24
144
105
61
18
4.31
Lohse
29
177
113
48
18
4.35
Looper
28
167
90
43
22
4.53
I like Szymborski's method for categorizing pitchers; here Carp and Wainwright are "top third" pitchers, and Wellemeyer, Lohse, and Looper are "middle third." If you don't have to worry about Carp's nerves rebelling this looks great! But if you do—
—well, it would help to name the pitchers who finish ahead of penciled-in #4 Joel Pineiro. Jess Todd; P.J. Walters; Mitch Boggs; Tyler Herron; Blake Hawksworth! Not much to say about this contract, anymore, except that even Jamie Moyer has to miss some bats.
64 comments
|
0 recs |
Do you like this story?
Comments
Yadi
I think the key with Yadi is that he’s got pujols-esque work ethic. I know he’s worked with Pujols in previous offseasons to improve his batting, and that’s why I think it’ll continue to improve. He makes contact a lot, and it’s bound that they’ll fall for hits. Maybe it’’l take some luck to get over .300 but on the other side, I can see him adding some more pop too. Pujols is awesome, so imagine him teaching you, ahh, that would be nice.
Any although these prodictions don't keep him above .300,
these are still a ton better than the year he was hovering around .215 BA for most of the season.
defy, cards, defy. hey logic --- you suck.
I'd bet my weight in gold.....
That if healthy, Yadi hits higher than .275.
This is where projections fall short. As somebody that has actually watched Yadi over the years, you could really tell that things were starting to click for him last year. Will he hit over .300 like he did this year? Maybe not. Doesn’t mean he’ll fall back to the .260’s or .270’s though. I’d be pretty shocked if that happened.
"Stats are for losers," Muschamp said after last week's victory. "I like winning games."
.275+ With Great Defense
You’ve got to be happy with that from a catcher. You’ve got to go a long back to find a Cardinals catcher that hits as well as Yadi. Defensively, he may not be as good as Matheny was. Yadi is better defensively than most of the Cardinals catchers in recent history.
"The big possum walks late." - Harry Caray
What?!?
“Defensively, he may not be as good as Matheny was. Yadi is better defensively than most of the Cardinals catchers in recent history.”
What does this even mean? We don’t have to go too far back to find a good defensive catcher. Like you said Matheny was pretty good.
That being said, are we so sure Matheny was better defensively? I’ll watch a game just to see Yadi through back to first… its chilling.
They have very different skillsets
Matheny had much better glovework/blocking ball stuff. Molina has that absurd rocket arm. This is getting into sorcery-level talk, but it seemed like Matheny worked with the pitchers a little better. I’d say it’s pretty close to a wash.
They say that it's never too late, but you don't get any younger...
my personal no-evidence projection
.280 with heretofore unseen power. Fifteen home runs, maybe? He hits the ball hard and Brother Bengie eventually developed some pop.
ditto my no-evidence take
Two things affected him this season (previous VEB link): the off-season conditioning, and the concussion. Shifting weight around in the body will change center-of-gravity; I imagine that takes getting used to. For instinctive players like the Molinas, it may take a while to settle in.
I’ll be watching for that this off-season (is there room for him in the WBC?)— will he keep up with his conditioning, and is he really recovering from his last few hits at the plate.
"But listen, and understand: more Molinas are out there. They can't be bargained with. They can't be reasoned with. They don't feel pity, or remorse, or fear." - THT
also if our two favorite issues are resolved
I predict he’ll have a much better season controlling the running game too. See link above. MIs and bullpen, yesterday and forever please!
"But listen, and understand: more Molinas are out there. They can't be bargained with. They can't be reasoned with. They don't feel pity, or remorse, or fear." - THT
Thank goodness someone noted that projections can't predict every outcome ever.
Scouts everywhere were clutching their paychecks in fear.
I fail to understand why you are so threatened by projections and advanced statistics.
You don't have to be a jerk.....
Considering there are ALOT of other posts saying the same thing mine did.
I get it, you don’t trust your own eyes, so you “go by the numbers” on everything. That way you can always fall back on them, and you never have to go out on your own.
"Stats are for losers," Muschamp said after last week's victory. "I like winning games."
Yeah
I don’t think I’ve ever seen the ‘numbers aren’t ballsy’ argument before. There is obviously some credibility to the ‘youneverknow’ argument, the ‘numbers don’t capture everything’ argument, and the ‘a regression is only as good as the model used to generate it’ argument.
but this one is new to me.
They say that it's never too late, but you don't get any younger...
I guess he's going out on his own.
Kosuke Fukudome: $48 million .257 .359 .379
Skip Schumaker: $Free .302 .359 .406
Skippy needs a new publicist, but I heart Ben Zobrist
One hypothetical internet dollar says regression projections on the whole are more accurate than ANY one person's "insight"
Kosuke Fukudome: $48 million .257 .359 .379
Skip Schumaker: $Free .302 .359 .406
Skippy needs a new publicist, but I heart Ben Zobrist
this is friggin awesome
I’m glad I found this!
go rays
by Cards Fan in Chitown on Oct 28, 2008 12:38 PM EDT up reply actions
And you don't have to constantly repeat the mantra of how useless numbers are.
It’s exhausting reading the same thing from you every time. I really don’t care if you think I’m a jerk. I really don’t care if you think I don’t trust my eyes. You haven’t been around long enough to know whether or not I make predictions outside of the numbers so feel free to judge and I’ll feel free to ignore your opinion of me that isn’t based in reality.
If you think a certain projection is wrong, how about discussing it w/ an actual reasoning. This “projections fall short” thing is misguided on your part. Projections are far more accurate than any individual person. There are simply too many ingrained biases for an individual person to overcome. Swallow your pride and accept as what they’re intended to be: a valuable compliment to scouting that helps to neutralize and point out areas where an individual may be wrong.
Umm.....
If you think a certain projection is wrong, how about discussing it w/ an actual reasoning.
Didn’t I do just that by saying:
As somebody that has actually watched Yadi over the years, you could really tell that things were starting to click for him last year.
Those projections are taking into account some downright terrible years at the plate for Yadi. Anybody that watches STL games knows he’s made huge strides the last 2+ seasons. So using a season from 5 years ago to project his 2009 makes ZERO sense. ZERO. He isn’t that player.
For you to jump on THIS post of mine is humerous. I’ve made much more in depth complaints in the past. This one was rather tame, in comparison. You don’t like what I have to say, don’t read it, or don’t respond. But I will keep posting my feelings.
"Stats are for losers," Muschamp said after last week's victory. "I like winning games."
by SoonerfanTU on Oct 27, 2008 11:54 PM EDT up reply actions
Go ahead and keep posting your feelings. When they sound like a giant “Waaah”, I’ll do my best to gloss over them. You don’t have to like what I say but you do need to offer some kind of coherent reasoning behind your statements.
What does clicking mean? What makes you think he’s finally getting it? Why should we just disregard the bad seasons? Do you disregard Rasmus’ bad season? That seems like a bias to me.
I substantially disagree that multiple years don’t make sense. You assume that Yadier can only get better. Young players hit stumbling blocks all the time — why should Yadier be any different. Yes he was good in 2008 but I, personally, don’t have that hard of a time imagining him going back to the 2006 ways where every week was a new batting stance. I also don’t have a hard time imagining his slow footedness being a real hindrance to him having a good season. That’s cost him before and if he’s not lacing the ball on a consistent basis, it’s liable to happen again.
I'm not trying to be a hardass stat nazi
although I suspect that’s the opinion you’ve formed of me. It just seems like an awful lot of posters want to wishcast things they don’t like/understand away because they feel they know better as an individual. Disregarding the validity of that sentiment, if every posts devolves into a "I think, you think" with no logic to back it up, there’s not a lot of discourse to be had. (Game threads are exempt from this. They’re a completely different animal.)
If people just want a place to record their opinions and not have to back them up, buy a diary. Otherwise there should be some rationale explained that can be discussed without everyone getting all fussy about it. I always try to wear my big boy undies to VEB — I’ll let people that I see face to face deal with the temperamental crap.
the irony here is
that I’ve used stats (specifically, BABIP peripherals) to attempt to show that evidence exists to support Sooner’s “feelings” with respect to Yadi a number of times, going all the way back to 2006.
the enemy's gate is down.
my question is
what the hell is a sooner?
go rays
by Cards Fan in Chitown on Oct 28, 2008 12:42 PM EDT up reply actions
Someone who stole land when they opened up Oklahoma to white settlers
They say that it's never too late, but you don't get any younger...
wow, what a pessimistic set of projections.
Does Zips really believe that every single starting player on our team this year outperformed their true talent, by an average of over 50 points of OPS each+?
+(non-weighted for PA because of time concerns)
Zips 09 2008 "regression" yadi 0.692 0.741 0.049 Apu 1.039 1.105 0.066 AK 0.647 0.693 0.046 Miles 0.682 0.753 0.071 Glaus 0.811 0.855 0.044 Iz2 0.612 0.628 0.016 Skip 0.732 0.765 0.033 Luddy 0.869 0.966 0.097 Barton 0.725 0.746 0.021 Mather 0.733 0.78 0.047 ankiel 0.814 0.843 0.029 0.0519
the enemy's gate is down.
Is ZIPS notoriously low?
Others have said James is usually more optimistic with his projections. Is ZIPS just pessimistic?
if not pessimistic
far more likely to underweight recent major league performance than you and I, just eyeballing the stats, would be.
caveat on the PAs and IPs
Nobody should get worked up about the relatively small number of PAs for Ludwick or IPs for the rotation. These are, sort of, middle-of-the-road projections that include some likelihood that they’ll get hurt on the first day of the season and never play again. The rate numbers are the ones that should be paid attention to. If Ludwick stays healthy, and hits at the rates the ZIPS projections suggest, he’ll end up w/ more than 30 bombs and the other counting stats will take care of themselves.
Rotation innings...
The thing that sticks out to me is how not one of our projected starters is expected (by Zips) to throw 200+ innings.
Which would mean that (unless we pick up another starter) we’re bound to see a significant amount of innings from Boggs/Todd/Walters/etc.
for what it's worth
ZiPS, as far as I can tell, projects innings only inasmuch as it can see innings there. So a guy like Wellemeyer or Looper, whose innings were limited by some circumstance other than injuries, be it playing time or role, might be shortchanged there.
Assuming ZIPS is regression type analysis?
Does anyone know is ZIPS is just doing a multiple regression, and it is purely based on empirical data. Or does the forecast include intangible factors, where an analysist would quantify some qualitative trait such as “player is starting to decline”, “the previous numbers were injury related so add a factor to bump up figures”, etc…
In differentiating ZIPS from James, I am assuming that their methodology is similar (regression) and they just use different exogenous variables? Just a guess.
born Dodger blue, now dyed Cardinals red
No qualitatives
“Player starting to decline” would show up in the statistics, injury factors are not considered whatsoever…Mike Hampton had a projection this year. Dunno about ZiPS but I know PECOTA takes things into account like height/weight, signing bonus/draft slot, but mostly it’s just regressing how players of a similar type age.
Kosuke Fukudome: $48 million .257 .359 .379
Skip Schumaker: $Free .302 .359 .406
Skippy needs a new publicist, but I heart Ben Zobrist
In multivariate regression analysis
I think generally that, rather than trying to add in qualitative factors, they try to isolate skewing variables and simply get rid of them or make them into binary answers. Adding fudge factors is another way to create garbage output.
by Red in Chicago on Oct 27, 2008 3:55 PM EDT up reply actions
ZiPS doesn't project playing time with anything other than throwing crap on the wall and looking at the smudges, Dan himself has said it like 100,000 times over at BTF
The rates are the only thing it actually projects and it pro-rates them for arbitrary playing time measures. There are almost always more innings/hitter ABs than there are in a season for all the players projected…
Kosuke Fukudome: $48 million .257 .359 .379
Skip Schumaker: $Free .302 .359 .406
Skippy needs a new publicist, but I heart Ben Zobrist
thought looper, pineiro, thompson and boggs were better in 08
than looper, wells, reyes, maroth and thompson were in 07.
if the top 3 in 09 make 92 starts, 70 starts are going somewhere.
it’s currently carp, pineiro, thompson and boggs.
dear santa, plz don’t forget johnies 4th SP.
why
are we predicting such low number of starts for Carpenter?
Yadi swings and hits a high fly ball... Endy Chavez goes back, to the track, to the wall... ITS A GUNNER!! Yadi gives St. Louis the lead in the top of the ninth!
[Total Guess Alert!]
Carp didn’t exactly start a lot of games after 2006. I’m guessing these projections take most recent years into account more than less recent ones, and that it doesn’t really have a “rational” decision-making process for predictin number of starts. As posters have mentioned above, the rate stats are usually more relevant that the counting stats. The projection probably doesn’t know how healthy the player is/was and I assume it would be very tedious for a human being to go through each player and adjust for the peculiarities of his situation. Then again I realy have no idea and am just spitting out what makes sense to me…
ZiPS doesn't care much for playing time
it can only predict it in the bluntest fashion—based on what the player’s done the last few seasons.
it's just going on the last few years
since he’s not been playing much, it just says that he will get hurt again pretty much
go rays
by Cards Fan in Chitown on Oct 27, 2008 5:34 PM EDT up reply actions
Some could ask
why would the Cardinals predict a higher number of starts for Carpenter than that?
by Hardcore Legend on Oct 27, 2008 6:03 PM EDT up reply actions
In today's mailbag
and on the Depth Chart Carpenter is not in the starting five. His health is the main concern coming into the 09 season and rotation so you could be right. Yet again, I do believe if Carpenter is close to 100% health at ST, he will have as many starts as possible…
Yadi swings and hits a high fly ball... Endy Chavez goes back, to the track, to the wall... ITS A GUNNER!! Yadi gives St. Louis the lead in the top of the ninth!
no kidding
unless that someone came from an alternate universe where Carpenter started more than 4 games total in the last two seasons.
I will be surprised if Carpenter makes a single start next season.
* sarcasm might be involved in this comment
unfortunately,
+1
Space.
It's a problem we face.
So we never go anywhere.
We just stay in one place.
he didnt do much to explain his interpretations
Looks like I picked the wrong week to quit sniffing glue.
Totalzone is a psuedo play-by-play metric based off of retrosheet data
It looks at where the ball was hit to according to retrosheet and assigns zones based on that. Basically is a simpler version of UZR and is similar to SFR that Dan Fox developed over at Baseball Prospectus.
hey thanks
defensive stats are so frustrating because they always come with several exceptions, and there really isn’t one universally accepted metric that does a sufficient job of incorporating all the info.
so robot, you should nominate a system for sole use on veb. that would help a lot.
Looks like I picked the wrong week to quit sniffing glue.
does anybody know
of a site that gives home and away splits on defensive metrics? I have been reevaluating my long-held opinion that Orlando Hudson is overrated, at least in part. I wonder if the desert air, or whatever it is that makes Chase Field such a great place to hit, not only helps the hitters, but skews the zone ratings as the harder, drier balls get through the infield faster?
If you look at ARI’s infielders, they all seem to have pretty poor zone ratings. Prior to coming to ARI Hudson was kind of meh offensive player with a good glove. Since arriving in the desert his offense has improved, but his defense has declined.
For the record, I think he is overrated offensively, but he might be a better defender than I think. It would be nice to see his zone ratings or +/- when playing in more normal humidity.
Those Pilgrims ain't lookin' so proud now...
I don't think that
exists, but I do know that MGL has explicitly talked about Arizona having a fast infield and that UZR applies a park factor, so your hunch is correct. I don’t think Dewan uses a park factor.
I'm not sure that I'd pick a single one
because the error bars are much larger on defensive metrics. If you want a great writeup on defensive systems, read this post by Justin Inaz. The scatterplot matrix is great and should show you why the DTs are not very reliable.
PMR and UZR are my favorites. RZR is nice because it’s updated daily but it’s not as accurate as the other two. TotalZone isn’t in there but it’s probably on the same tier as RZR and SFR.
Whoops
This is the one that has the explanation of the systems:
http://jinaz-reds.blogspot.com/2007/10/player-value-part-3a-fielding.html
the variability
of the various systems is whats getting me. by looking at the various systems i can come to an almost democratic conclusion about who is a better fielder at one position or another, what i can’t do is determine an actual value in terms of, say, runs or win shares when compared to traditional offensive stats.
Looks like I picked the wrong week to quit sniffing glue.
several of the systems use
plays above average as a baseline. You can usually convert from plays to runs by multiplying plays*.8 - there’s some discussion to be had here based on position but I’ll set that aside. That should allow you do develop a baseline between UZR and say PMR or the +/ system. Come to think of it, UZR might be the only one that reports it’s findings in terms of runs. . .the others all use plays, I think.
Dial's system uses runs
and it’s pretty solid. And speaking of Dial and runs/play, he’s got some specific numbers here: http://www.baseballthinkfactory.org/files/dialed_in/discussion/dr_strangeglove_or_how_i_learned_to_stop_worrying_and_love_zone_rating1/
WS weather
WS Philly weather
light rain, 46 F, NW wind at 15mph, gusts to 23mph
"But listen, and understand: more Molinas are out there. They can't be bargained with. They can't be reasoned with. They don't feel pity, or remorse, or fear." - THT

by 























