What is the most accurate SABR statistic?

In article <1114016901.834225.215970.RemoveThis@l41g2000cwc.googlegroups.com>,
psychsound <PsychSound.RemoveThis@aol.com> wrote:
>Of all the newfangled SABRmetric statistics devised over the past 25
>years, what is the most accurate? And which is the easiest to
>calculate and understand?
>
>My sense is that Runs Created is the easiest to calculate and
>understand since the formula adds and divides numbers easily available
>to us and the statistic itself roughly resembles runs scored or RBI,
>allowing for quick judgments about a player, i.e., 95 Runs Created is
>good, 54 Runs Created is bad.

Given equal playing time. But 54 runs created could be a great season
by a late callup.

And 95 runs created could be poorish for a full time 1B in Colorado.

>On-Base plus Slugging would seem to be
>the easiest to calculate. Any thoughts on this?

Clay Davenport designed EQA so that it has a similar range to batting
average. Means that even if most people don't understand how to
calculate it, they have a good feel for what a good season is.

And EQA has another important thing going for it. It's more
accurate than runs created. (Runs created has one of the larger
standard errors among the good metrics of offensive performance.
In no small part because it uses bad weights for some of the
lesser variables. Runs created isn't substantially more accurate
than OPS and OPS is a heck of a lot easier to calculate)

Runs created in its basic forms has a honking big flaw. Because
its multiplicitave it has the rather odd effect of asserting
that a HR hit by (say) Frank Thomas is more valuable than one
hit by (say) Joe Carter.

Doubt this? Take any of Joe Carter's seasons. Calculate the runs created.
Then add a 1-1 with a HR and recalculate the RC. Repeat with any
of Frank Thomas'. To pick one year at not random, in 1995 that extra
HR is worth around 1.43 for Carter and 2.02 runs for Thomas.)

Dave Tate worked out a way around this -- what he called marginal
lineup value. Calculate team runs created. Then calculate team
runs created with the player's stats removed. Credit the player
with the difference.

In recent years Bill James has adopted Dave Tate's approach -- without
giving any credit to Dave. (Though to be fair, James probably came
up with the idea on his own. He'd been aware of the problems with
runs created for decades. Which is why he gave prominent space to
Paul Johnson's approach -- which once you simplify the equation turns
out to be a linear weights approach)

As for the best sabrmetric work, my vote would go to Michael Wolverton's
work on starting pitching. SNWL (support neutral wins and losses) and
company. Doug Drinen's work with relief pitching was great too, but
he's too busy with www.pro-football-reference.com to continue doing them
any longer.


--
RNJ
 >> Stay informed about: What is the most accurate SABR statistic? 

Realto Margarino wrote:

>
> OPS does not measure a baseball skill. It is meaningless.
>
> cordially, as always,
>
> rm

Yes, Hits is the most accurate statistic.
 >> Stay informed about: What is the most accurate SABR statistic? 

On Wed, 20 Apr 2005 19:16:43 -0500, John Gregory
<ashbury DeleteThis @mirage.skypoint.com> wrote:

>If Hits be a skill, 'tis a strange one, that goes up and
>down depending on the skill of the day's opposing pitcher.
>
>I like Plate Appearances. It measures the skill of being
>recognized by the manager the most times as being the
>favorite of the fans.

But by maynard logic it's a variable skill too, one that only shows
itself 3 to 5 times a game.

Bob Roman
 >> Stay informed about: What is the most accurate SABR statistic? 

"Eric Ramon" <Eramon1.DeleteThis@aol.com> wrote in message
news:1114041388.609922.91660@z14g2000cwz.googlegroups.com...
>
> Realto Margarino wrote:
>
> >
> > OPS does not measure a baseball skill. It is meaningless.
> >
> > cordially, as always,
> >
> > rm
>
> Yes, Hits is the most accurate statistic.

:: rolls eyes ::

Haven't you been paying attention? Rings, you bozo, r-i-n-g-s. What Luis
Sojo has five more of than Ted Williams and Barry Bonds combined.
 >> Stay informed about: What is the most accurate SABR statistic? 

JPM III wrote:
> psychsound wrote in
> news:1114016901.834225.215970@l41g2000cwc.googlegroups.com:
> >
> > Of all the newfangled SABRmetric statistics devised over the past
25
> > years, what is the most accurate? And which is the easiest to
> > calculate and understand?
> >
> > My sense is that Runs Created is the easiest to calculate and
> > understand since the formula adds and divides numbers easily
available
> > to us and the statistic itself roughly resembles runs scored or
RBI,
> > allowing for quick judgments about a player, i.e., 95 Runs Created
is
> > good, 54 Runs Created is bad. On-Base plus Slugging would seem to
be
> > the easiest to calculate. Any thoughts on this?
>
> OPS is by far the most accurate, and it's easy calculable by adding
readily
> available statistics (OBP and SLG).
>
> I think a modified OPS would be even more accurate if you could work
in
> stolen bases or a speed factor somehow, but as it stands OPS is by
far most
> accurate.

By far the most accurate? As I've heard it explained, OPS is slightly
less accurate, as a predictor of runs scored, than most of the other
"total offense" stats; even on-base TIMES slugging is more accurate.
OPS's selling point has been its simplicity.
 >> Stay informed about: What is the most accurate SABR statistic? 

psychsound wrote in
news:1114016901.834225.215970@l41g2000cwc.googlegroups.com:
>
> Of all the newfangled SABRmetric statistics devised over the past 25
> years, what is the most accurate? And which is the easiest to
> calculate and understand?
>
> My sense is that Runs Created is the easiest to calculate and
> understand since the formula adds and divides numbers easily available
> to us and the statistic itself roughly resembles runs scored or RBI,
> allowing for quick judgments about a player, i.e., 95 Runs Created is
> good, 54 Runs Created is bad. On-Base plus Slugging would seem to be
> the easiest to calculate. Any thoughts on this?

OPS is by far the most accurate, and it's easy calculable by adding readily
available statistics (OBP and SLG).

I think a modified OPS would be even more accurate if you could work in
stolen bases or a speed factor somehow, but as it stands OPS is by far most
accurate.
 >> Stay informed about: What is the most accurate SABR statistic? 

Realto Margarino wrote in news:42675b12@news.usenetzone.com:
>
> Steve Grant <ACE1242 RemoveThis @comcast.net> trolled:
>
> > Haven't you been paying attention? Rings, you bozo, r-i-n-g-s.
> > What Luis Sojo has five more of than Ted Williams and Barry Bonds
> > combined.
>
> Sojo had the more successful career.

No he didn't. Sojo didn't win those rings. His teams did. He was just a
small part of it.

He was not successful as a player. He just happened to play for teams that
made up for his physical ineptness.
 >> Stay informed about: What is the most accurate SABR statistic? 

psychsound wrote in
news:1114016901.834225.215970@l41g2000cwc.googlegroups.com:
>
> Of all the newfangled SABRmetric statistics devised over the past 25
> years, what is the most accurate? And which is the easiest to
> calculate and understand?
>
> My sense is that Runs Created is the easiest to calculate and
> understand since the formula adds and divides numbers easily available
> to us and the statistic itself roughly resembles runs scored or RBI,
> allowing for quick judgments about a player, i.e., 95 Runs Created is
> good, 54 Runs Created is bad. On-Base plus Slugging would seem to be
> the easiest to calculate. Any thoughts on this?

Also, for career value, I like using the sum total of Total Bases, Bases on
Balls (including HBP), and Stolen Bases. Because whereas OPS measures
offensive effectiveness per each plate appearance (how good a player is with
the bat), TB+BB+SB measures how effective a player was over the long term.
And only players good enough to stick around for long careers will be near
the top of that list.
 >> Stay informed about: What is the most accurate SABR statistic? 

"+Alan Hicks+" <alan DeleteThis @lizella.net> wrote in message
news:426ab162@news.usenetzone.com...
> JPM III <jpmccord DeleteThis @hotmail.com> trolled:
> > Realto Margarino wrote in news:42675b12@news.usenetzone.com:
> >>
> >> Steve Grant <ACE1242 DeleteThis @comcast.net> trolled:
> >>
> >> > Haven't you been paying attention? Rings, you bozo, r-i-n-g-s.
> >> > What Luis Sojo has five more of than Ted Williams and Barry Bonds
> >> > combined.
> >>
> >> Sojo had the more successful career.
> >
> > No he didn't. Sojo didn't win those rings. His teams did. He was just a
> > small part of it.
>
> He was just as big a participant as all the other members of his
> team.

I told you guys not to underestimate how stupid maynard can be.
 >> Stay informed about: What is the most accurate SABR statistic? 

+Alan Hicks+ wrote in news:426ab162@news.usenetzone.com:
>
> JPM III <jpmccord DeleteThis @hotmail.com> trolled:
> > Realto Margarino wrote in news:42675b12@news.usenetzone.com:
> > >
> > > Steve Grant <ACE1242 DeleteThis @comcast.net> trolled:
> > >
> > > > Haven't you been paying attention? Rings, you bozo, r-i-n-g-s.
> > > > What Luis Sojo has five more of than Ted Williams and Barry
> > > > Bonds combined.
> > >
> > > Sojo had the more successful career.
> >
> > No he didn't. Sojo didn't win those rings. His teams did. He was
> > just a small part of it.
>
> He was just as big a participant as all the other members of his
> team.

Yeah but in this case, size doesn't matter.


> > He was not successful as a player. He just happened to play for
> > teams that made up for his physical ineptness.
>
> He played for a winning team. Therefore, he had a successful year.
> In fact, it is impossible to have a more successful year than
> playing on the WS winner.

Wrong. You have no real concept of success. A player who plays for a World
Series champion but looks absolutely horrible doing it and is never hired to
play baseball again as a result has quite an unsuccessful year.
 >> Stay informed about: What is the most accurate SABR statistic? 

+Alan Hicks+ wrote in news:426aaf8f@news.usenetzone.com:
>
> JPM III <jpmccord RemoveThis @hotmail.com> trolled:
> > psychsound wrote in
> > news:1114016901.834225.215970@l41g2000cwc.googlegroups.com:
> > >
> > > Of all the newfangled SABRmetric statistics devised over the past
> > > 25 years, what is the most accurate? And which is the easiest to
> > > calculate and understand?
> > >
> > > My sense is that Runs Created is the easiest to calculate and
> > > understand since the formula adds and divides numbers easily
> > > available to us and the statistic itself roughly resembles runs
> > > scored or RBI, allowing for quick judgments about a player, i.e.,
> > > 95 Runs Created is good, 54 Runs Created is bad. On-Base plus
> > > Slugging would seem to be
>
> 95 is good at what? What is 54 bad at?
>
> What are you talking about?
>
> > > the easiest to calculate. Any thoughts on this?
> >
> > OPS is by far the most accurate, and it's easy calculable by
> > adding readily available statistics (OBP and SLG).
>
> It's like adding apples and oranges. That's why real stat fans
> won't use it.
>
> > I think a modified OPS would be even more accurate if you could
> > work in stolen bases or a speed factor somehow, but as it stands
> > OPS is by far most accurate.
>
> Sorry, Rhonda, but OPS is worthless. The number does not measure a
> skill.

Correct. It does not measure a skill. It measures a player's effectiveness
as a hitter.
 >> Stay informed about: What is the most accurate SABR statistic? 

Seapig wrote in news:1114291325.197388.100240@o13g2000cwo.googlegroups.com:
>
> JPM III wrote:
> > psychsound wrote in
> > news:1114016901.834225.215970@l41g2000cwc.googlegroups.com:
> > >
> > > Of all the newfangled SABRmetric statistics devised over the past
> > > 25 years, what is the most accurate? And which is the easiest to
> > > calculate and understand?
> > >
> > > My sense is that Runs Created is the easiest to calculate and
> > > understand since the formula adds and divides numbers easily
> > > available to us and the statistic itself roughly resembles runs
> > > scored or RBI, allowing for quick judgments about a player, i.e.,
> > > 95 Runs Created is good, 54 Runs Created is bad. On-Base plus
> > > Slugging would seem to be the easiest to calculate. Any thoughts
> > > on this?
> >
> > OPS is by far the most accurate, and it's easy calculable by adding
> > readily available statistics (OBP and SLG).
> >
> > I think a modified OPS would be even more accurate if you could
> > work in stolen bases or a speed factor somehow, but as it stands
> > OPS is by far most accurate.
>
> By far the most accurate? As I've heard it explained, OPS is slightly
> less accurate, as a predictor of runs scored, than most of the other
> "total offense" stats; even on-base TIMES slugging is more accurate.
> OPS's selling point has been its simplicity.

Haven't heard that one (OBP*SLG). I'll look into it.
 >> Stay informed about: What is the most accurate SABR statistic? 

+Alan Hicks+ wrote:
> JPM III <jpmccord.RemoveThis@hotmail.com> trolled:
>
>
>>>Sorry, Rhonda, but OPS is worthless. The number does not measure
>>>a skill.
>>
>>Correct. It does not measure a skill. It measures a player's
>>effectiveness as a hitter.
>
>
> The only thing that measures effectiveness in baseball is wins and
> losses. Nothing else counts.
>
> cordially, as always,
>
> rm

Oh, god, he's morphed again. What a wanker.
 >> Stay informed about: What is the most accurate SABR statistic?