Part of CSNPhilly.com


« First base, outfield glovework among nits to pick | Main | Phillies pitch past Marlins, lengthen lead in East »

Tuesday, September 14, 2010

Comments

From last thread: Just to inject some substance into the debate about UZR sample sizes, this is from the Fangraphs UZR Primer:

"In fact, the year-to-year correlation of OPS for full-time players, somewhat of a proxy for reliability, is almost .7. UZR, in contrast, depending on the position, has a year-to-year correlation of around .5. So a year of OPS data is roughly equivalent to a year and half to two years of UZR."

Only an idiot would call others "idiot" for using a season's worth of UZR to examine fielding performance.

Clout, are you asserting that the average of three unreliable numbers results in a reliable number? Yikes thats naive.

Is the start delayed? If so, appreciate any solid guesses as to probable duration. Need to call home and have teh recording extended before I get on this plane . . .

Spitz: Can't ANY statistically significant sample size be broken into smaller parts which, if taken alone, would be unreliable? If so, wouldn't it also follow that mutliple samples which are too small to be reliable can be combined into one larger sample which IS reliable?

The Good News: I have Phillies tickets for this coming Friday & Tuesday, which gives me a Roy/Roy Double Header of sorts.

The Bad & Ugly News:

"Kendrick will start only two more games this season, both against Washington." - Matt Gelb

Kendrick vs. the Natinals in '10: 3 GS, 12.71 ERA, 2.471 WHIP, Opponent Slash Line of .404/.475/.654/1.129, Kyle's worst numbers vs. any team he has faced this season.

Hugh: Should be starting soon.

Danke schoen

BAP - yes, any statistical sample can be broken down into smaller parts, but the average is only meaningul if enough of those smaller parts are used to compute it. In other words, a batter's OPS for a game is not meaningful. The same stat averaged over 3 games is also not meaningful. But over 162 games, it suddenly starts to mean something.

Think about it from a standard deviation or a standard error perspective. The standard deviation of any sample of three is very large. Why? Because a sample of three is meaningless.

Nats up 2-0 over ATL.

Nats up 2-0 in the 2nd.

Washington up 2-0 on a Dunn home run.

Love this ballpark...and this defense.

That was fast.

Polly!!!

No AAA journeyman one-hits this team twice.

Chase really chased that one...

Spitz: That's more or less what I was saying. If one season of UZR is too small a sample size to be reliable, I don't know about the validity of simply averaging 3 unrealiable numbers. But you could certainly take the overall average based on 3 seasons of data.

Fvcking A...Howard better be okay.

His name is Dan %$#@!

Dan Uggla is "the most prolific power-hitting 2nd baseman in history?" Are these guys for real?

****Dan Uggla is "the most prolific power-hitting 2nd baseman in history?" Are these guys for real?****

You caught that "gem" too?

Ridiculous.

Livan Hernandez, RBI double.

Jurrgens has had some rough outings lately. Anyone know who Atlanta plans to put up against our big 3?

Cole might be due for a bit of a clunker.

It's early but Cole looks off.

So BAP, are you saying that a 3 season average of UZR is a reliable measure or not? I am arguing that if you believe a single season UZR is meaningless, than a 3 season average of UZR is equally meaningless.

Hamels or KK...you decide.

Hamels gives up one run

This is what we get for praising our pitching staff for 3 days straight.

Hamels looks like total crap tonight. There's no "Fast" in his "Fastball". Maybe it'll just take him awhile to get warmed up ... *crossing fingers*

Hamels hopefully settles in and goes 7 or 8.

Not so bad.

Last fastball at 93. Nothing to sneeze at.

Spitz: It depends what you mean by a 3-season average. If you mean adding up his 3 unreliable single-season UZRs, & dividing by 3, then, yes, that would be unreliable. But I have a hard time seeing what would be unreliable about recalculating his UZR based on 3 years worth of data, even though any 1 year would be too small to be reliable in and of itself.

As for what Atlanta is likely to do next week...

I believe that if they stay in their same spots, the Phils would get Jurrjins, Minor and Hanson next week.

My guess, is that Jurrjins (who has been scuffling) moves to Sunday afternoon and Lowe is pushed back to Monday to start Game 1.

So, I think the likely matchups are:

MON- Hamels vs Lowe
TUES- Halladay vs Minor
WED- Oswalt vs Hanson

Does that Hit make Cole Hamels "the most prolific power-hitting Pitcher in history"?

Slugging Through Age 30:

Dan Uggla: .487
Chase Utley: .523
Rogers Hornsby: .570


If Mendez got to face the Phillies every time out, he might well end up as the most prolific starting pitcher in history.

Take it for what it's worth, but the statement about Uggla probably is connected to the fact that he just became the first second baseman in Major League history to have four consecutive 30 homer seasons.

BAP, you would get the same result either way. The average of three yearly averages = a three year average.

Scott: Obviously, that was the basis for their assertion. But it's still pretty far out there. I mean, Uggla's a fine power hitter & there haven't been a whole lot of power-hitting 2nd basemen. But, until & unless he has about 5 more similar seasons, the words "Dan Uggla" and "baseball history" do not belong in the same sentence.

Remember Uggla's misplays in the 2008 all-star game? The one where Lidge threw like 100 pitches in the pen and the game was almost called? That's the only tie to baseball history Uggla has thus far.

Nats up 4 to 0 in top of 4th

Spitz: Not true. For example, let's say Player A has the following numbers over 3 years:

Year 1: 100 AB, 30 hits, BA .300
Year 2: 500 ABs, 120 hits, BA .240
Year 3: 500 ABs, 120 hits, BA .240

The average of his 3 single season batting averages is .260 ((.300 + .240 + .240)/3)

In actuality, his batting average over that 3-year span is .not .260, but .245 (270/1100).

BAP -- The Fangraphs quote I posted is pretty clear that there's nothing inherently unreliable about a full season of UZR. There's nothing "magical" about 162 games that gives BA, OPS, or any other offensive stat, either. Is 110 games reliable for offensive stats? At all? I think so.

Remember when Jayson Werth was freakishly good at getting 2-strike hits?

I guess last night's hit was a fluke for Werth.

Twice now where he's crapped the bed again.

GTown, I think Livan Hernandez has usurped Hamels "the most prolific power-hitting Pitcher in history"

This is really frustrating to watch. This pitcher sucks and we're just letting him slip away inning after inning.

sifl: I take no position on whether or not a full season of UZR is reliable for what it purports to measure. I was just commenting on the notion that, if 1 season of UZR is junk, 3 seasons must be junk too.

Personally, I think UZR is junk over any time period, for reasons that have been disussed here ad nauseum.

BAP, good point. I was assuming the yearly averages would be roughly the same sample size.

These Marlins' announcers make Wheels, McCarthy & Sarge look like Hall of Fame broadcasters.

"And while it warms my heart any time a Howard throw does not involve the left fielder..."

Aksmith, that line had me laughing pretty hard.

And for the record, I think it's a mistake that the Phillies aren't swapping Halladay and Hamels in the rotation for the ATL series. (If they did, they'd Doc/Hamels/Oswalt, in that order, in both ATL series) If things hold out as they are, and the season goes down to the last game (which it most likely will), and the Phillies make the playoffs, they will need to pitch Halladay on short rest to start the NLDS. As of now, Hamels would be lined up to pitch game one.

We demand extensive softball experience for our ballgirls...too funny.

Enough UZR talk - we need Chooch to motivate some RUiZ talk.

Phils have a good opportunity to pick up some unexpected ground on both ATL and CIN tonight - they need to capitalize.

TMac, how is the Lakewood/Greenville game on the Toyota MAJOR LEAGUE Scoreboard?

Maybe Mendez will hurt himself running to 1B again. At least Cole appears to have settled in.

The Phillies ought to go back to demanding a certain amount of attractiveness -- like, any -- in their Ballgirls.

This game feels like a loss. Cole's control is off and, for reasons completely mysterious to me, the Phillies can't hit this AAA pitcher.

Is this the guy that threw out Howard?

Cole has settled in...what an Ace.

Fatalotti - do you have some sort of crystal ball that says that ATL and Philly will be seperated by 1 or 0 games 17 games from now?

The Phils have a team of ballgirls that play softball games, hence the softball requirement.

I wonder how accidental that was.

Don't look now, but Hamels is close to his first 200 strikeout season. Also, accodring to B-R.com, Roy Oswalt is on pace for 204 Ks this season.

Have we ever had a season with 3 pitchers who had 200 strikeouts?

The Phils have a team of ballgirls that play softball games, hence the unattractiveness.

***The Phils have a team of ballgirls that play softball games, hence the softball requirement. ***

I had no idea...that's cool.


Valdez is quite predictable at times.

Uggla looks pretty not lousy tonight.

You can't hope to stop Wilson Valdez ('s double plays), you can only hope to contain Wilson Valdez ('s double plays).

Interesting trivia question I heard today:

Only 4 active RHP have had at least 4 seasons w/ 200+ Strike Outs. Name them.

You know how they sometimes have those skills contests during the All Star break? Maybe next year they can have a skills contest for grounding into DPs. Valdez would be the heaviest favorite since the Bird/Jordan/Magic Olympic Dream Team.

Fatalotti, are you sure? I thought the last series was Fri-Sun, with the NLDS beginning Wed. If that's so, Doc would get 5 days between starts.

BTW, that was Exxon's 19th GIDP in just 311 PAs.

What phillie is named Sanchez. Because uggla just threw him out according to tmac

Spitz, no crystal ball, but I think it's far more likely that those 3 games in ATL at the end of the season mean something (at least the first 2) than that they won't. Even if Atlanta is out of it, the Phillies will have to get really lucky to have locked up the 1 seed before that series.

Plus, wouldn't it be good to get Hallady lined up now so that he can go into game 1 of the NLDS, never having thrown on short rest?

Cole has amazing stuff.


Just sayin'.

Ok, I think I misunderstood. How do the Phils propose to pitch the last series, if not Doc, Cole, Oswalt?

195th K on the year.

GBrett, the last series against ATL begins on 10/1. The NLDS begins on 10/6, with 4 days between the two dates. The pitcher who gets the ball in game 1 of the ATL series is lined up to pitch game 1 of the NLDS (should the Phils make it) on normal rest.

If they go Hamels, Halladay, Oswalt in this upcoming Atlanta series, it would be Hamels, Halladay, Oswalt in the final ATL series. If that's the case, either Cole gets game 1 of the NLDS on regular rest, or Doc gets the ball in game 1 of the NLDS On 3 days rest.*

*This assumes that no pitcher pitches on short rest during the regular season.

By the way, Hamels is striking out everyone.

196th K on the year.

75 pitches through 4....lucky to have him go 6 tonight :(

Hamels, after that 1st inning, has simply had unhittable stuff tonight. His fastball and changeup are both working and he's just mowing these guys down.

His pitch count is rising though so he might only go 6.

Fatalotti, given the last month, I think I'd rather have Cole pitch game 1 of NLDS...especially if the Phils have the option of taking the 5 games in 8 days and not having anyone go on short rest the entire series!

cole is dealing but it looks like we'll only have him through 6, unless he gets a real quick inning. how is this mendez guy shutting us down again? i can accept last week, since it was the first time facing him -- they should be teeing off on this guy today

I'd go Halladay, Hamels, Oswalt in any playoff series. Split the RH arms with probably the best LHP in the NL right now.

Shane Victorino for president....er are Hawaiians considered citizens? haha JK. but seriously, he should be our full time lead off guy....anyone got his numbers from the 1 spot this year?

Gtown: Halladay & Vasquez are 2. Still thinking (or, more accurately, looking names up on the Internet). I'm thinking at least 1 of the other 2 is some older guy who is pitching relief now or whose best years are behind him.

75 pitches, 14 swinging strikes for Cole.

The baseball gods just love 5 hour downpours when you try to plan your rotation 3 weeks in advance. And the way the top 3 are pitching right now, It doesn't matter what order they pitch in the playoffs.

Steve: 300 AB .283 / .345 / .480 / .825 11 HR, 41 RBI, 19 SB, 23 BB, 39 K

"I'm thinking at least 1 of the other 2 is some older guy who is pitching relief now or whose best years are behind him."

Jose Contrares?

****.er are Hawaiians considered citizens****

Only if they're really from Kenya...JK

~voted for him~

My expectations for this inning are dropping rapidly with that flyout. I can easily see K, K here.

What they most likely will do, at least according to Zolecki:

I messed up, they can go Hamels, Doc, Oswalt in this upcoming series, and then switch Halladay and Hamels in the finals series:

9/14, FLA: Hamels, 5
9/15, FLA: Halladay, 4
9/16, OFF
9/17, WSH: Oswalt, 4
9/18, WSH: Kendrick/Worley, 6
9/19, WSH: Blanton, 5
9/20, ATL: Hamels, 5
9/21, ATL: Halladay, 5
9/22, ATL: Oswalt: 4
9/23, OFF
9/24, NYM: Blanton, 4
9/25, NYM: Hamels, 4
9/26, NYM: Halladay, 4
9/27, WSH: Oswalt, 4

My guesses from there on out:
9/28, WSH: Kendrick/Worley, 9
9/29, WSH: Blanton, 4
9/30, OFF
10/1, ATL: Hamels, 5 or Halladay, 4
10/2, ATL: Halladay, 5 or Hamels, 6
10/3, ATL: Oswalt, 5

Who knows what they'll do, but they are going to have to flip Halladay, Hamels at some point, if they want Doc going in game 1 of the NLDS. Why not now?

well I think that solves it then, Gtown. I definitely think Shane should lead off for good.

"And the way the top 3 are pitching right now, It doesn't matter what order they pitch in the playoffs."

Exactly.

b_a_p: Your thinking is correct. Hint - Two of the Four are teammates.

The comments to this entry are closed.

EST. 2005

Top Stories

HardballTalk

Rotoworld News

Follow on Twitter

Follow on Facebook

Contact Weitzel

CSG