Baseball Toaster Bronx Banter
Help
The Long Goodbye
2007-01-31 09:41
by Alex Belth
Note: The Bronx Banter blog has moved to bronxbanterblog.com.

I think most of us will agree that the addition of Mr. Minky represents the worst move the Yankees made this off-season. And while I believe that many think the team would be better off without the services of Bernie Williams, other fans don't want to see him go. What to do? Jon Heyman has the latest over at SI.com.

Elsewhere, Mike Mussina tells it like it is; David Pinto links to a story about Melky Cabrera; Steven Goldman compares the Yankee and Red Sox hitting, and according to the AP, the Yankees will host the All-Star Game in 2008.

Comments (80)
Show/Hide Comments 1-50
2007-01-31 09:54:35
1.   mehmattski
To jump on the theme suggested by the title:

Spending money wisely by refusing to ante up for Shea Hillenbrand types and instead improving the team's defense at 1B?

That's OK with me...

Trading away aging veterans in a clear sign that Cashman is finally in charge of the front office; making the team younger and more flexible in the long run?

That's OK with me...

Investing the Yankees vast resources not in overpriced 30-something free agents, but in scouting and developing amateur talent for free, to the point where Keith Law ranks the Yanks with the #5 overall farm system?

That's OK with me...
Politely letting our aging star know that his services are no longer required, but honoring him with Bernie Williams Day, a plaque in monument park, and ensuring that no one ever wears #51 for the Yankees ever again?

That's OK with me...

What? Don't tell me I'm the only Elliot Gould fan around here...

2007-01-31 09:54:46
2.   markp
I doubt PH goes north with the big team, but even at 5 IP per start he could be a big help in the Bronx come summer. I do think RJ going deep into games is going to be missed.
2007-01-31 09:56:04
3.   Shaun P
Sigh. I love Bernie, but its time to move on. I really hope this doesn't become a soap opera.
2007-01-31 10:04:04
4.   yaggi666
only if bernie learns to play first will he go north this year
2007-01-31 10:10:18
5.   Shaun P
Question for season-ticket holders - what kind of season-ticket package (if any) entitles you to be able to purchase a ticket for a special event, say the All-Star Game?

I would love to see the All-Star Game, and if buying a pack of 10 or 20 tickets or whatever guarantees me being able to buy an All-Star ticket (or two) for face value, I'd gladly pay upfront for the season ticket package rather than try to get ASG tickets through a '(wink wink) respectable broker'. Seeing a bunch of games during the Stadium's swan song also has a lot of appeal.

2007-01-31 10:20:29
6.   pistolpete
In other news, Gorilla Boy gets married.

http://tinyurl.com/39w4ez

In NYC...?

At a hot dog stand....?

2007-01-31 10:21:27
7.   Yankee Fan In Boston
mussina has climbed on board the pavano backer bandwagon!

"(H)e can do it, we know he can pitch, and we know he can get people out. ...(H)e'll be an asset."

i didn't read the rest of the article (or that particular quote) but it's good to see pavano fever is as contagious as the norovirus.

2007-01-31 10:22:20
8.   Yankee Fan In Boston
6 speaking of theo, today is national gorilla suit day.

http://tinyurl.com/cb7k9

2007-01-31 10:22:35
9.   Jim Dean
5 I was wondering the same thing. But I bet with all the tickets they have to set aside for other teams, I think only full season ticket holders get a shot. Partial plans may get a shot at the other events though.

I'm sure they'll have that worked out by next winter.

2007-01-31 10:33:13
10.   Jeteupthemiddle
I just want to say that the Mientkiewicz signing is not nearly as bad as a lot of people here seem to think.

He is a replacement level player with above average defense making $1.5M.

He doesn't strikeout all that often providing us with some contact at the bottom of the lineup, and his OPS+ has improved each year for the past 3 years.

Over 162 games, for his career he put up a line of .270/.359/.405....or roughly what Melky Cabrera put up last year.

I will gladly take those numbers from part of a platoon on any given day when the rest of our lineup looks the way it does.

2007-01-31 10:36:38
11.   mehmattski
10 But, what if Robinson Cano gets hit by a meteor and Derek Jeter contracts Hantavirus?????
2007-01-31 10:40:31
12.   Ben
Bernie's my favorite Yankee all time, but Bernie at first??? I can just see it. Big Papi at the plate, yanks one down the first base line, cut to: Bernie still in expectent crouch, motionless... blink, blink.

Leave Ferdinand in the field where he belongs!

2007-01-31 10:40:41
13.   Jeteupthemiddle
11 Well then we are clearly screwed because our back up catcher isn't above league average.
2007-01-31 10:45:24
14.   murphy
that's just it, 1 and 10, we've been shown here and on eleventeen other blogs that minty boy's defense prowess is mythical. he hasn't been any better than average in years.

sure, 1.5 mil is a small drop in the proverbially botomless yankee payroll bucket, but we could have gotten the same performance on both sides of the ball out of andy phillips. this is the waste of a roster spot.

2007-01-31 10:46:40
15.   Jim Dean
11 At least they have 64 RHRP to deal from when necessary.

13 And Cairo, he can play anywhere on the field! Maybe Torre will let him catch a game or three.

2007-01-31 10:59:28
16.   Yankee Fan In Boston
15 but those pitchers are gritty. and cairo...

eh... i can't even pretend to defend the cairo signing.

2007-01-31 11:00:01
17.   Jeteupthemiddle
14 I don't agree with that either. He may not be what he once was, but he is still above average and far above whatever else we had on the team and/or available to us.

Dialed In at the Baseball Think Factory hasd Mientkiewicz as the 2nd best defensive 1B last season and saved 12 runs per 150 innings.

http://tinyurl.com/yblbqa

Within that same thread, Mitchel Lichtman, the creator of UZR agreed with the findings of Mientkiewicz, that he saved 12 runs per 150 innings.

http://tinyurl.com/2ko9l4

The Fielding Bible went back and looked at all the 2005 innings for every firstbaseman with at least 1000 innings to see how many "bad throws" a player saved.

Pujols was leaps and bounds better than anyone, but Mientkiewicz was 3rd with saving 20 "bad throws" in only 675 innings.

http://www.actasports.com/sow.php?id=85

For 2006, the Fielding Bible rated Mientkiewicz as the 2nd best firstbaseman behind only Teixeira.

TangoTiger also believes that Mientkiewicz is worth between $4M and $6M.

http://www.tangotiger.net/salary.html

2007-01-31 11:01:11
18.   Yankee Fan in Chicago
Surprised no one's commented on the Steve Goldman article.

Was Pecota smoking crack this offseason? It has Posada's Marginal Lineup Value at .036 and Varitek's at 0.47! While I don't doubt that Posada will decline from his excellent 2006 season, why does Pecota think that Tek will bounce back?

Not to mention a steep decline for Cano to the point where he's scarcely better than th boy wonder Pedroia. Apparently they see Cano's 2nd 1/2 last season as a fluke rather than a breakout, which, given his age, seems ludicrous.

There's more I could quibble with, but those are the most egregious.

2007-01-31 11:06:35
19.   Jim Dean
18 It was in the last thread. But most folks would rather be drunk on Kool-Aid.

I agree though that many of the projections seem strange. But I agree with his overall point is that it's closer than you'd think and 1B for the Yanks is certainly not helping things (no matter how drunk some folks are).

Bullpen goes one way, the rotation goes the other. It's going to be a very fun season.

2007-01-31 11:27:44
20.   jayd
The three first basemen is another problem keeping Bernie off the field. It is painful to see him being treated like a journeyman but until the Melky trade comes in, that's what perceptions are going to be.

Notice how the Yanks pulled the Melk out of winter ball so he could have an awesome spring.

The 3 First basemen, the fourth outfielder getting regular playing time on some rotation -- how does that work (Sorry Johnny, knew you were 5 for 6 yesterday but it's your turn to sit down). DH is going to be one busy position this year. Besides, Giambi hits better when he fields. I know it's not pretty, but I like to see the big guy happy.

I frankly just don't get alot of this. The upshot of it all is we put in Minky for Bernie? Say it ain't so, Cash.

In fact it's all so nonsensical, it only gives legs to JohanMyJohan arriving at the close of spring training.

I've seen Minky play outstanding defense, so that ain't myth or legend.

2007-01-31 11:42:23
21.   dianagramr
Meanwhile, Murcer continues his fight against cancer, and gets a visit from Andy ...

http://tinyurl.com/ysmghs

2007-01-31 11:42:38
22.   Shaun P
18 I said this in the last thread:

"PECOTA doesn't say Varitek will be better than Posada. When you look at projected defense and playing time, Posada's PA, VORP, and SuperVORP beat Varitek's across the board: 480/20.4/22.4 versus 405/15.5/17.4."

In fact, looking at the weighted mean PECOTA projections, except for MLVr and batting average, Posada's projection beats Varitek's across the board. Its sometimes close, and playing time is a factor, but its in Posada's favor.

Last thing - the BP folks themselves continually point out that PECOTA is often pessimistic on a number of players, and produces some strange projections for some guys. But, IMHO, in the aggregate, its a pretty good system (markp I know disagrees).

2007-01-31 11:45:36
23.   Yankee Fan In Boston
21 thanks for the link. i missed that today. good news.
2007-01-31 11:49:34
24.   Jim Dean
22 Let me ask you guys this:

Do we really need projection systems to figure out what guys will do?

Cann't we just rely on our heads and:
1) the player's age
2) their career stats
3) their stats from the previous three seasons

Doesn't that "formula" do pretty well? Sure, it won't give you numbers, but when there's going to be inaccuracies any ways (and I haven't seen any one with a projection system also report anything close to confidence intervals based on their past accuracy) why worry about the numbers?

I guess what I'm saying is it's at that point that I just look forward to the games.

2007-01-31 11:50:54
25.   standuptriple
9 I've been looking into in for SF this year. It seems like there's really only 2 ways to get them and most required a commitment already or a $500 deposit per ticket to get the lowdown. I imagine in a swan song season for The House coupled with already heavy demand will almosst certainly require a 40 game package. Start hitting your fellow YF's up for $ now. At the very least unattentable games probably won't be hard to unload for face or better.
2007-01-31 11:54:59
26.   Jim Dean
BTW: I hope I'm wrong and Ohlendorf is successful just so we can call him Dorf and add it to every part of our vocabulary.

"Man, Dorf was lights out yesterday."

"His sinker really dorfs batters."

"The young rotation is quite dorfy."

"Y'all are a bunch of drunken dorfs."

2007-01-31 11:58:02
27.   Yankee Fan In Boston
26 one more thing to get my hopes up. (as if igawa's interview and the return of igawa weren't enough...)
2007-01-31 11:59:06
28.   dianagramr
26

If he is homer-prone, we'll just have to call him Walldorf

2007-01-31 11:59:12
29.   Yankee Fan In Boston
27(replace the 2nd igawa with a pavano.)
2007-01-31 12:01:13
30.   standuptriple
I was kind of hoping it gets shortened like A-Rod's to "Oh-Dorf"
2007-01-31 12:03:05
31.   standuptriple
Let's hope its not D'oh!(rf), a-la Homer J. Simpson.
2007-01-31 12:12:36
32.   jakewoods
Its time we cut ties with Bernie. It should have happend after '04 when Beltran was free.
2007-01-31 12:26:57
33.   Shaun P
24 IIRC, that's what Tom Tango's Marcel forecasting system essentially does. Its very simplistic, I know that.

For me, I just like the numbers, 'cause it gets me thinking about the games.

28 Will whoever the catcher is then become Statler?

2007-01-31 12:34:37
34.   Yankee Fan in Chicago
22 My objection is not whether Pecota says/does not say that Tek will be better, but with the fact that it has them even in the same league.

His Mlvr last season was -.109. (Posada's was .154 on the plus side)

That not only seems a substantial difference, but as I noted I see no reason why they think that a 35 year old catcher will bounce back rather than continue his decline.

Likewise with Cano, a very young player who came back significantly bulked up after his midseason injury and proceeded to absolutely mash the ball. I know that the knock on him is that his stats are driven by his ba, but in the 2nd 1/2 of last year that wasn't true, he didn't just hit for avg, he slugged the eff out of the ball.

Yet they have Cano, a young player who's just learning the league and hasn't even peaked yet, declining from a .252 mlvr to a .083. I think that's ridiculous.

2007-01-31 12:56:00
35.   markp
A couple of questions for PECOTA fans:
1. What's the formula? How can you rely on it if you don't know what "it" is?
2. Have you ever compared PECOTA to any other method? The one given here (look at their age and past several seasons) has, from my observations, been far superior in predicting what a player's going to do than PECOTA.
2007-01-31 13:26:53
36.   vockins
35 Here's a link to a BBTF post concerning player evaluation for 2006:

http://tinyurl.com/ye7znp

I wouldn't call myself a PECOTA fan, but they certainly did something right last year, at least more right.

2007-01-31 13:29:18
37.   Sliced Bread
12 Ferdinand! Good one, Ben. I'd actually prefer the "blink blink" Bernie scenario you presented to watching Doug Out snuff another Yankees rally with his BELOW AVERAGE BAT (for a 1B).

14 Right on, Murph. Doug's a waste of space.

17 So what if Stinky saves 10-15 runs over an entire season with his magical glove, it's his broken back which concerns me?
The first time A-Rod launches a throw toward the box seats, will Stinky's surgically repaired back hold up?

And does Tango (assessing Stinky's worth $4-$6 million dollars) also project that he will grow 7 feet taller and become thrice as nimble before the season begins?

The only saving grace here is that Stinky comes cheap and will be easy to trade, or flush when the time comes.

2007-01-31 13:36:23
38.   ny2ca2dc
37 the fact that his is cheap and easy to trade/flush should not be underestimated. I think of him as a stopgap... If he's starting in the playoff, I'll be surprised and unhappy. But think of it this way, we could trade some of the live-arms for a 1B now, or we could wait a while and see how the team shakes out. Never know, Moose and Wang could get injured and those live-arms might need to be traded for a starter. The point is, the live-arm guys are a comodity, and be be traded for whatever's most necessary down the road a piece. I realize many are concerned about getting over a barrel'd at the deadline if we have a glaring hole, but that's just life - if we traded some of the chips for a 1B now, and needed a solid started at the deadline, we'd still be in a position of weekness, and would be shorter trading chips to boot.
2007-01-31 13:39:41
39.   Shaun P
35 You know the answer to number 1, and I honestly don't understand why that matters.

The basic factors that go into the PECOTAs are known. The basic methodology is known. Is PECOTA a black box? Sure, but I'm OK with that; I have an engineering background, so I'm used to black boxes. Much like my computer and my cell phone, I can't tell you every detail of every level of how they work - physics wasn't my best subject - but I rely on them nonetheless.

As for number two, no, I personally have not. But other people much smarter than me have, and its been found to be pretty equal.

In fact, a very interesting roundtable at the Hardball Times from August 2006 (Part 4 at http://tinyurl.com/2ztpdp) had leading sabermetricians and projection guys discussing the different systems. While its clear that similarity score-based systems like PECOTA have their limitations, and in particular probably cannot be refined as much as regression-based systems, noted sabermetrician Mitchel Lichtman (MGL) - creator of Ultimate Zone Rating and co-author of "The Book" said, and I quote:

"As far as I can tell, PECOTA is an excellent system overall."

It is perfect? Of course not. But its good enough for me.

BTW, if you have research that clearly shows that Jim's 'off the top of his head' system has been 'far superior' to PECOTA, I'd like to see it. I'd also like to know what this system thinks Scott Baker of the Twins is going to do this year; I'd like to keep him on my fantasy team, but PECOTA is not very optimistic on his performance this year.

2007-01-31 13:55:28
40.   Jim Dean
39 I'd like to see that resarch too, cause I could increase my bravado 100% and really make definitive statements.

Hmmm, Scott Baker ...computing...

Looks like he won't get alot of time in the Twins rotation or pen. And he's too susceptible to the long ball. If a better option could be had, I say dump him.

2007-01-31 13:58:47
41.   Shaun P
40 See, that's basically what PECOTA was saying (weighted mean projected ERA of 4.81). I was hoping I had a sleeper on my hands. C'est la vie.
2007-01-31 14:05:33
42.   Jim Dean
41 And my head is an even worse black box! Shoot, I don't even understand it.

36 Wowser! That's mighty good for PECOTA and position players. Still that's the aggregate and leaves room for alot of variability of individuals - like Cano and Jorge and Varitek.

But nothing seems to do great for pitchers. And those are all guys with established MLB stats.

I suppose even more reason why position players are more valuable :).

2007-01-31 14:08:39
43.   markp
Zone rating-I'd like to hear someone's explanation of how it isn't subjective. It's based on what the observer perceives the feilder's zone to be, his judgment of how hard the ball was hit, etc. Using the guy who developed that to praise PECOTA doesn't impress me.
PECOTA has been compared by several, disinterested parties and the results weren't close. There have been articles about PECOTA that disagree with what you're saying happened at Hardball Times. The use of phrases like "projection based systems" and "regression based systems" seems at odds with what every one of these formulas (including PECOTA) are: attempt to predict what a player is likely to do. The only distinctions are their accuracy and their transparency. PECOTA falls short on both counts.
2007-01-31 14:16:59
44.   dianagramr
And now for something completely different ...

John Kerry to question FCC on Extra Innings / DirecTV deal ...

http://tinyurl.com/3bpx45

2007-01-31 14:26:23
45.   Start Spreading the News
43 normally, i am good at googling but i can't find those articles you are talking about.

Can you post some links?

2007-01-31 14:27:14
46.   weeping for brunnhilde
I'm thrilled to hear the news about Melky.

Thrilled, I tell you.

He's going to be a monster.

No, I don't mean he'll hit 40 home runs and drive in 140, but he's going to be the guy I want to see up when the tying run is on second base in the eigth inning.

He's going to be the guy who goes up there each and every time with a solid, quality at-bat.

He's not going to be the one to strike out or pop up when wood needs to be put on the baseball.

I look for him to hit around .310 or even .320.

This kid is that good.

2007-01-31 14:30:18
47.   tommyl
44 Also, this idea that MLB is making more money by picking DirecTV is simply incorrect. Cable (through InDemand) offered $70M/year for non -exclusive broadcast rights. That means that MLB can still sell rights to DirecTV and the Dish for additional money. Surely they can get at least $30M a year from both of those companies combined. I think the sticking point is that cable companies want to put the MLB channel on their sports tiers as opposed to basic cable. MLB sucks.
2007-01-31 14:31:01
48.   weeping for brunnhilde
32 Speaking of Beltran, that was a terrible at-bat.

What a frustrating way to end a fine season.

2007-01-31 14:35:31
49.   Start Spreading the News
I found a Baseball Prospectus article:
http://www.baseballprospectus.com/article.php?articleid=2515

They compared PECOTA to the following:
1. BBHQ Ron Shandler's Baseball HQ
2. DMB Diamond Mind Baseball forecasts created by Tom Tippett
3. Primer Baseball Primer/ZiPS Projections
4. RotoTimes
5. RotoWire
6. Warren Scoresheet guru Ken Warren's projections

For 2004 data (and all caveats of using only one year apply here), they found:
PECOTA 77
DMB 62
Warren 55
BBHQ 54
Primer 43.5
Times 30.5
Wire 14

But I guess you would say that they are not a "disinterested" party.

Reading Wikipedia's entry on PECOTA, we find that for 2006 PECOTA beat Las Vegas (among others) in predicting team wins.

2007-01-31 15:00:56
50.   Bama Yankee
44 That's good news, I guess... As long as the "Swiftboat Veterans for DirecTV" don't get involved and Senator Kerry doesn't tell the FCC that he was actually for the DirecTV deal before he was against it. ;-)

Thanks for the link dianagramr. The following line was at least encouraging:
"Meanwhile, cable executives close to the negotiations said the industry has not been notified by baseball that it's out of the running to secure rights to the package."

Maybe there's hope yet...

Show/Hide Comments 51-100
2007-01-31 15:05:54
51.   Shaun P
43 Not Zone Rating, markp - Ultimate Zone Rating. Widely acknowledged as the best fielding system anyone had seen (this was back in '02 or '03, IIRC). MGL stopped doing it publicly because I believe he was using it for the Cards when he worked for them.

"The use of phrases like "projection based systems" and "regression based systems" seems at odds with what every one of these formulas (including PECOTA) are: attempt to predict what a player is likely to do. The only distinctions are their accuracy and their transparency. PECOTA falls short on both counts."

You first sentence makes no sense - every projection system that I know of is either similarity score-based, or regression-based. Those phrases simply describe, in general, how the system creates the projections it makes.

As for transparency, OK, PECOTA isn't transparent. You still haven't explained why thats a big deal. You can still test the results without knowing the method.

Speaking of which . . . as for accuracy, so far, folks have found links to studies showing PECOTA is pretty darn accurate.

But, if what you say, that:

"PECOTA has been compared by several, disinterested parties and the results weren't close. There have been articles about PECOTA that disagree with what you're saying happened at Hardball Times."

is true, then give us the links to those articles!

2007-01-31 15:43:43
52.   markp
Who? Zips? Diamond Mind?? (You do know Diamond Mind is a computer baseball game, right?)
I'm talking about the real projections, like using Bill James formulas, or even taking the past three seasons K/9 OPS+ ERA+, etc. and using that.
I already said I don't have the links. They were posted here last summer.
Range factor isn't subjective. UZR is. Until we have the technology to take the observer out of the equation, it has to be subjective. Some people think it's great. Others say rating fielding is still in its infancy and anything with a subjective element is unreliable. I agree with the latter. Some of the ratings made by the UZR people are pretty hard to credit. Again, if you like it, use it.
Look, if you want to use PECOTA, knock your socks off.
BTW every formula since the first Baseball Abstract was used not only to judge what happened the year before, but to project what was going to happen in the player's immediate future (the following season.) I'm sorry you didn't understand that, but there never was a difference between "progression" and "regression" in sabermatics. It's always been about what a player's real value was, and that means last year (regressive, I presume) and next year (progressive?)
2007-01-31 16:08:18
53.   yankz
After Jeter and Cano beat their projections by about 2 billion OPS points and the Red Sox sucked it up, I stopped believing in projections.
2007-01-31 16:09:09
54.   RIYank
Speaking of Beltran 48, he's among Melky's "comparables" at Pecota; so I see over at YFvSF, anyway, and I can't check because I'm not a subscriber. Two of the other top four comparables: Roberto Alomar and Rickie Ashburn. (They don't mention the fourth, but the context makes it clear that it's an unimpressive player.)
2007-01-31 16:29:51
55.   Start Spreading the News
52 If you said that you didn't have the links, you didn't say it in this thread of discussion.

I also think that the confusion here is a result of two seemingly opposite terms (projection vs regression) that actually work toward the same goal.

Here is my attempt to define terms so we all are working with a common terminology.
The projection system (term first used by Jim Dean in this thread) refers to any system that tries to project the performance of a player.

Regression based systems are systems that use regression analysis to project a player's performance.

Thus a projection system may use regression equations to predict future performance. For the sake of our discussion, let us say that regression based systems are a subset of all the projection bases systems people use to analyze/predict a player.

So the idea why PECOTA is better than regression analysis systems is given here from wikipedia:
"Unlike performance forecasts that commonly assume a single pattern of change during a player's career, PECOTA employs several models that take into account not just a player's performance in the previous three years but also his age, speed, handedness, and body type (basically, body mass index). Furthermore, instead of focusing on making point estimates of a player's future performance (such as batting average, home runs, and strike-outs), PECOTA relies on the historical performance of the given player's historical "comparables" to produce a probability distribution of the given player's predicted performance during the next five years."

So if 95% of fast,tall, lanky players who hit 250 homers with a .280 AVG declined by 30% in between the age of 35 and 36, PECOTA will make such an adjustment to an active baseball player who fits that description.

It makes other adjustments as well (park factor, NL/AL, etc...)

The problem with PECOTA is with the outliers. Since it uses historical data to adjust an active player's stat, how does it adjust someone like Barry Bonds who has no historical comparable. So for those type players, PECOTA will struggle.

2007-01-31 16:33:36
56.   OldYanksFan
S.Goldman predicts: Rodriguez's 500th HR will come at Yankee Stadium on August 30. It's a day game against the Red Sox."

Alex... Cliff... can you guys set up a 'office pool' as to guessing what this date will be? The winner can be excused from having to debate Jim for 2008.

2007-01-31 16:39:37
57.   weeping for brunnhilde
54 Thanks for the heads up. Now if I only knew what "comparable" meant.

I'm thinking it means something other than what it sounds like.

:)

(I'm a stat dinosaur and remember well the days when there were no "holds" and "saves" really meant something, so you'll have to forgive me.)

:)

2007-01-31 16:58:38
58.   mikeplugh
I'm a BP fan, and I'm a "Premium Subscriber", one of the few places I'm willing to shell out the dough to be an "insider". I still have questions about PECOTA when it comes to certain factors like injuries that are not fully disclosed, small sample sizes for young players, and such. If Varitek, for example, has a certain physical ailment, like degeration of ligaments or something in his left shoulder and elbow (all completely hypothetical) and the doctors and trainers haven't said anything to the public other than, "he's experienced some wear and tear", then PECOTA can get thrown out the window.

I know that may be an exception to the rule, but I believe we know far less about the actual condition of aging veterans than we realize. Medical info is vague and while the longer trend of similar catchers at the same age can be projected onto Varitek's current situation, any individual's physical characteristics will potentially skew the data to a point beyond viability. This is particularly true of catchers, I think.

I have absolutely NO evidence of this, but my common sense tells me that I'm not too far off base. I chose Varitek as an example, because his past performance and the similarity comparisons may tell us what to expect next year, but my eyes tell me that there's something far more speedy about this player's decline than meets the numbers. Just an opinion.

The small sample size for young players is also something I question. No one would have projected Cano as anything more than a fair Major Leaguer based on his production at the lower levels. He looked good, and some scouts liked him better than others, but the BA people thought less of him, and the Yankees farm in general, than what we've seen so far.

When he came up, it was kind of a "give it a whirl" move by the front office to fill a hole. No one thought he'd be compared to Rod Carew (for whatever that's worth) at any point in his 23rd year on Earth. If you look at what he did last year and project it out to 162 games, it was very similar to Donny Baseball's "age 23" season with the Yankees. The 2nd half power surge bodes well too. Most young players don't find their power stroke until they are around 25, give or take a year. Cano MAY HAVE found his in the 2nd half of 2006, and should he continue to progress he'll absolutely blow PECOTA out of the water. He may have blossomed late, and I find it very hard to imagine from watching the guy for 2 years that he's peaked, or that he'll slip significantly. I think his 2nd half performance last year is just as likely a future outcome for Cano in his prime, and any depreciated output based on PECOTA. If I'm right, we can tip our cap to BP and know in our hearts that they missed one.

I'll go on the record today to say that Robinson Cano will hit like Don Mattingly in his prime, more than once in his career. I don't know if he has 40 home run power, but I believe he has 30 home run power. His spot in the lineup will prevent him from driving in 125-130 runs, and he'll need to work hard to reach 200 hits from the bottom of the order, but consider that he hit 165 times in 122 games last year. Get him the extra 120 or so at bats that he'd have in a full season and he'd put up close to 200. In fact, with 120 more at bats last year at the pace he set, he'd have about 205 hits from the bottom part of the order. PECOTA be damned.

2007-01-31 17:01:50
59.   RIYank
57 It does mean what it sounds like. They find the players with the most similar statistical profiles at the same age/stage of career. Of course, the statistical profiles are a lot fancier than what you would use, I'm sure, but the concept is transparent.

Why do saves mean less now than they used to?

2007-01-31 17:08:32
60.   weeping for brunnhilde
59 Thanks, RI.

As to saves?

Well, back when I was a lad relievers (not closers) would come in for more than three outs, first of all.

Second of all, they'd actually come in with the express purpose to get the starter out of a jam, thus "saving" the game.

I'm sure there's all sorts of statistical data to illustrate that somehow it's a better to have a specialist get those last three outs as they do today, but what can I say, that's not how they did it in my day, by gum.

;)

2007-01-31 17:13:08
61.   RIYank
Ah.
No, my guess is that the data illustrate that the closer, who is your best relief pitcher, should be brought in to get you out of a jam, whether it's the sixth inning or the ninth. Managers don't use them properly, on the whole. If we have a tie game in the seventh and Pettitte starts us off by giving up a single and a walk, I would much rather see Mo than Bruney. But we know it will never happen.
2007-01-31 17:40:28
62.   mikeplugh
61 That point was explained nicely in the BP book "Baseball Between the Numbers: Why Everything You Know about the Game Is Wrong." There's an entire chapter, written by Keith Woolner called "Are Teams Letting their Closers Go to Waste". Pick up a copy and give it a spin. It's a good thought-provoking read.
2007-01-31 17:41:45
63.   mikeplugh
62 BTW....Woolner is the guy who invented VORP.
2007-01-31 17:43:23
64.   Jim Dean
56 And if I win I get to debate myself everyday.

58 Good morning! Here, here on Cano. I can't see how he could regress. Further, I can't see how Melky can't improve significantly with his plate discipline. Two more fun things to watch.

Finally, PECOTA misses what it means to make it in New York. Whatever part of the game is mental is missed by all projections, and making it in the Bronx must be a huge lift to the Psyche, esp since it's so rare. More common is the difficulties faced by guys who come to New York and we've seen that with multiple HOFers. Call it the NY effect. Some guys it affects negatively and others rise with the pressure.

Cano and Melky have proven that they can make it, and at very young ages. There's no underestimating that.

2007-01-31 17:48:52
65.   weeping for brunnhilde
64 Absolutely, Jim.

Frankly, I don't see how anyone could even dream about letting Melky go having seen his first game last summer, where he fought and scratched and drew a walk out of Schilling in what was a very big spot.

Was it the same game, or just the same series, where he kind of botched that fly ball?

Anyway, you don't need to see the guy's statistics, just watch him play and you know he's the real thing.

2007-01-31 17:58:05
66.   mikeplugh
I've become quite a stathead, as much as I can comprehend most of it. I like the idea that past performance, similarity scores, in game outcome probability, and the rest can predict what will happen in the future. It works.

The part of me that will always be rooted on the other side of the fence is my trust in my own eyes and intellect. It may fail me more often than not against the numbers, but I feel that there are certain points of analysis that human intellect and experience can contribute that fill the gaps between all the rational data that makes up metrics.

Guys like Cano, to me, are those cases that require a bit of "eyeing" to fully understand. If I'd never seen him, and I'd never watched what he can do, I'd believe all the PECOTA projections and I'd probably think he was a "nice little middle infielder with occasional pop". I may use that kind of thinking for guys that I've never seen in the San Diego farm system or something, and I'll go to my grave trusting it enough to make some educated guesses on things I have no first hand knowledge of. That's the gift of modern baseball analysis. The failing is the intuitive part of the process that allows even old school "belly full of guts" type scouting to work.

If I had to put a number on my feelings, I'd say that metrics are about 75% of the story to me, and the other 25% is my eyes and intuition. I feel comfortable with that. Other may go 90/10, 50/50, or 23/77 for what it's worth. Interesting to hear what other people put their numbers at....

2007-01-31 18:52:31
67.   Shaun P
52 If these links were posted here last summer, then they ought to be easy enough to find, right?

I just Googled "PECOTA 2006" and restricted the search to bronxbanter.baseballtoaster.com

I did find a post from you, markp, on April 14, 2006: "33. markp
Pecota is a joke. There have been a few nice articles on the internet detailing its many faults (the biggest being its inaccuracy.)"

No links there. In fact, no links anywhere that show PECOTA's 'many faults'.

I just don't understand what your beef with PECOTA is. Yes, I get it, its not transparent - which is true but why is that important since you can evaluate the method by the results? And, according to you, its not accurate - but so far you have not been able to show any evidence supporting that statement.

If PECOTA is so bad, and there are articles on the Internet proving its faults, then I would like to see them so I can read them and try to understand where they are coming from, and perhaps temper my expectations and reliance on PECOTA.

In the meantime, PECOTA is what it is - not perfect, not the end-all-be-all, but all the evidence I've seen says its pretty good for what its supposed to do. Until I see solid evidence to the contrary, I have no reason to think otherwise.

66 I'm definitely a 90% numbers/10% eyes guy. There are plenty of things that I can't see that nonetheless happen. But, no projection system is capable of taking human factors into account, like succeeding in NY, or dealing with personal issues, or Joe Torre's irrational reliance on 'guts' relievers and intentional avoidance of others. Stuff like that I leave to my eyes, to temper the numbers, and remind me that you need both to understand the game.

2007-01-31 19:38:48
68.   Jeteupthemiddle
37 The point is that he is still a good defensive first baseman, despite popular belief on this thread.

As for offense, I believe he is the definition of a replacement player. Which means he will still have value because he will perform offensively exactly as you could expect Random Triple A Man but he will also provide you with run saving defense.

To top it all off, he is cheap. Cheaper than any other option available on the market.

If Eric Duncan all of a sudden shows he is unbelieveably amazing and batting .400 in August and ready to be called up, it is very easy to drop Mientkiewicz.

I think people here are over critical of this signing. For $1.5M it is a decent signing. Certainly not a GOOD singing, but it is a far cry from terrible.

2007-01-31 20:11:52
69.   markp
Were you reading Bronx Banter last summer? I don't recall seeing your handle until very recently.

Here's a couple of discussions showing the high regard BP has among the SABR world.
http://www.baseballthinkfactory.org/files/newsstand/beat/Projections/
http://www.baseballthinkfactory.org/files/newsstand/discussion/ny_times_schwarz6/

The number of links that show up when you google PECOTA and a word like "inaccuracy" or "wrong" is similar to when you type almost any pair (or trio) of words-thousands. I did get these two and saw quite a few others, but filtering through them is too much work. I suggest going to various fan sites and noting the disdain that teams sabermaticians have for it. I also found a post I'd made at NNYFans concerning the predictions for Jose Contreras, Jeff Weaver, etc by BP a few years ag0. The difference between what they'd predicted for the Yankees and Red Sox pitching staffs and what really happened were comical. (I saw a similar post at a Chisox website laughing about the 2005 BP predictions for the their staff.

If you really want to find evidence pro or con, you do the legwork. I (and others) have already done our homework. And I know in three months there'll be someone else saying "show me the link", and when I refuse, saying there must not be one.

2007-01-31 20:27:38
70.   OldYanksFan
I'm not a big StatsHead, but by the very nature of statistics, even if PECOTA is wrong X% of the time, doesn't mean it's not a good, or maybe even the best system.

The odds of rolling snakeseyes on a pair of dice is 1:36. However, give a dice 36 rolls, and you might get snakeeyes twice, or not at all. Roll the dice 360 times. Did you get snakeeyes 10 times? Double sixes 10 times?

Statistical 'Odds' are not a prediction, simply a best guess. Maybe the problem is calling PECOTA a 'forecaster' as opposed to a 'best guesser'. Add in the dozens, if not hundreds, of human factors that effect any one ballplayer's year, there is simply no way to accurately predict anything. You can just, based on the odds, make a best guess.

2007-01-31 20:32:39
71.   dianagramr
"PECOTA is my homeboy"
2007-01-31 21:00:40
72.   Start Spreading the News
69 Can the Pecota predict wrongly for players? ABSOLUTELY. It is just the other tools have been shown to be wrong more often than PECOTA is.

Besides the links you provide don't really help your argument that PECOTA sucks. The first link is to an MLB blog where the guy says what you do: Pecota sucks because it is a black box. That is not really a valid argument. You can still measure the tool based on how well it predicts. We have provided links in this thread that compared PECOTA to other tools.

The second link is to a NY TImes article that praises PECOTA. From the article: "No organization uses Pecota in a vacuum, instead incorporating it with other projection techniques like traditional scouting reports. But a decade after Bill Pecota retired, his name is growing more prominent each winter.

''Everyone in baseball is in the guessing business,'' the Mariners executive Dan Evans said. ''This makes it a little bit less of a guess.''"

That is from the article that you provided as proof against PECOTA. Yet it summarizes why people use PECOTA. Have you ever looked at a PECOTA prediction? Since I subscribe to the BP site, i can look at it. But sadly it is not free. The PECOTA page for a player gives different levels of predcition. For example, Derek Jeter:
Percentile -- Batting Avg
90% -- .347
75% -- .336
60% -- .328
50% -- .321
40% -- .315
25% -- .304
10% -- .288
Weighted Mean -- .322

So looking at that Jeter, if he performs in the 90th percentile of his ability (according to PECOTA), would bat .347. In the worst case scenario, Jeter bats .288. PECOTA provides such predictions with the IMPLICIT understanding that its predictions fall within a range. Typically writers use the weighted mean when saying that PECOTA predicts that Jeter will bat .322 next year. That was just batting average. PECOTA provides such predictions for many other stats: OBP, SLG, VORP, Defense, etc...

Additional PECOTA also provides the % likelihood that a player will have a breakout year (Jeter 6%), only improve (35%), decline a little (Jeter 25%), and decline a lot (Jeter 7%).

So there is more to this PECOTA thing that I think you are giving it credit for.

2007-01-31 21:18:15
73.   Russell W
Projections are like an educated guess. Guess is the key word and they can, and often, are off.
2007-02-01 03:25:38
74.   Jim Dean
66 I'm 90/10 too. The stats are the stats. But I don't think I need fancy projection systems to tell me much. I look at the numbers, take into account things like plate discipline, position, and age and that's that.

What's been bugging me about all the prospect lists is that the weight they give to "talent" or "potential". To mean that seems more like a recipe for disaster (Henson, Van Poppel) than if they just stuck to the numbers - stats, age, level (though for defense I think you need good scouts). So a guy like Clippard consistently gets rated under guys like Betances and Chamberlain. Weird, because Clippard has shown he knows (and can learn) how to pitch, is young, and is advanced.

Consider (IP HR9 BB9 K9):
Age 19 A 149 0.72 1.93 8.76
Age 20 A+ 147 0.73 2.08 10.32
Age 21 AA 166 0.76 2.98 9.47

What's not to like? That he doesn't throw a blazing fastball? If anything, it's even more impressive to me that he can K so many without it. Sure, he's having a little bit of trouble BB guys, but he's going to be 22 in AAA and is still building arm strength!

By contrast, Betances and Chamberlain have blazing fastballs and great reputations and have yet to prove anything (though Chamberlain looked very good in winter ball). I'll take Clippard, thanks, even as folks say he'll be a back of the rotation starter. The last I looked guys like that were getting 55 mil for 5 years. And I think Tyler will be much better than that.

The other guys don't have much of a track record besides radar gun readings and "reputation". Weird.

However, I also understand that that formula (stats, age, level - SAL) is good enough for us fans but not good enough for organizations because of the difficulities in comparing NCAA, High School, and international teams. Scouts get you going but ultimately the proving ground is the field of play. And I think the minor league system does a really good filtering - better than any other sport, I'd say, and indeed most professions.

That's why I'd much rather see teams trade the kids with "potential" for more established players (record in at least A+ ball). Sure you could be moving first-rate talent, but with pitchers it goes so wrong so often it seems like in the long run you'll make out just fine, especially with postion players.

2007-02-01 03:37:35
75.   Jim Dean
68 Instead of looking at last year - look at the last three years. Sure, he may end up as an average player. But the chances are better he could be well-below average. That's the problem.

And sure, he's cheap and they could easily get rid of him. But there are things they could do in the Spring that they won't do in June. For instance, if they had Craig Wilson last Spring there's little doubt in my mind he would have earned the starting 1B slot all year and heck he may have gotten some time behind the plate and in the OF. Instead, he had little room for error in June.

The cost of Mghfdjh in February is much greater than the roster slot and the money - it's the lost opportunity for the whole season.

2007-02-01 04:04:58
76.   Shaun P
71 Amen to that, Diana!

74 Jim, you make a lot of good points in there. From what I understand, the hardest level changes players make in the minors are from A to AA, and then AA to AAA. And, historically, the number of pitchers who dominate AA and AAA and go on to have decent major league careers without a 93+ MPH fastball or a 'plus plus' pitch is very small. The guys who have a blazing fastball and show some ability to actually pitch - like Betances and Chamberlain - are always hyped because those tend to be the type of guys who do make it. And chick dig those big numbers on the radar gun.

I think Clippard will be one of the exceptions because, as you said, the dude clearly knows how to pitch. But I understand why prospect types may be underrating him - they are hedging their bets. They've probably seen guys like Clippard before, and been fooled before, and don't want to be fooled again. Especially in a business where its easy (and potentially costly) to be wrong, I understand the reluctance to stick their necks out. They play it safe, and I can't say I blame them.

2007-02-01 04:28:21
77.   Jim Dean
76 See, but then they rate highly guys like Henson and Van Poppel. You look back and they did little to warrant it.

That's why I cringe with all the hype for Tabata.

I have little doubt that he "looks" good. Now let's see him prove it in A+ ball.

Montero doesn't exist yet to me.

2007-02-01 06:01:30
78.   Jim Dean
I can always leave it to Goldman to make me think in the morning.

I hadn't thought about it but what happens in this year's post-season means some interesting and disgusting possibilities for that All-Star game in the Bronx for 2008.

Suppose the Sox make the Series then Francona gets to pick and manage the AL team in Yankee Stadium.

More interesting perhaps, suppose in the best case scenario, that the Yanks win it this year and Torre decides to retire to manage his garden gnomes in Malibu.

Do the Yankees and MLB bring him back for one last send-off at the break? Would he name Bernie to the team as the 40 year old CF? Maybe Mattingly to play 1B?

2007-02-01 09:58:56
79.   Jeteupthemiddle
75 I doubt very much that anyone will read this thread again, but I do want to respond to that and do not wish to clutter another thread.

Over the last 3 years, his OPS+ has improved each year.

Over the last 3 years, his numbers are .253/.335/.386 for a .721 OPS....certainly not great, but certainly a replacement player...which is all I am saying Mientkiewicz is. Offensively, he is a replacement player.

Defensively, however, he is above average. I posted the links previously to show that.

For $1.5M, Mientkiewicz is an overall, net positive for the Yankees.

The fact that he is part of a platoon is even better.

2007-02-01 10:11:17
80.   Jim Dean
79 You're using the term "replacement" very loosely there. If he puts up .253/.335/.386 he's easily the worst 1B in MLB. So what exactly is that a replacement for?

And you're further reading too much into the OPS+ "improving" each year. He's been a part-time player each of those years and the variability is typical with fewer AB's.

Further, he's consistently hit worse against RHP - presumably his role in a platoon.

Sorry, man, the guy is an utter waste of a roster spot, and they should have brought in a kid to win the job outright. If we're lucky, mid-season will bring Teixeira, Sexson, Helton, or Nick the Stick.

Comment status: comments have been closed. Baseball Toaster is now out of business.