Expos
Martin Dubuc
at the beginning of the year. Now, Expos seem to be one of the hottest team.
What suprises me is that there isn't much peoples who know the players of this
team. If you look at the all-star game voting, none of the players gets
higher than the fifth place, still some of them are the best in NL.
Just think about it: they lost 3 of their best starting pitchers (Langtson,
B. Smith and P. Perez) last winter and still have one of the best ERA of the
league. Rookie pitchers Mark Gardner and Bill Sampen are doing a great job.
Dennis Martinez (5-4, 2.50) and Kevin Gross (8-4, 3.38) continue to do
an excellent job as starting pitchers.
Who knows that Spike Owen is the best defensive shortstop of the NL?
Last year he had the best defensive average. So far this year, he has
played 60 games in a row without error. The longest game streak ever without
errors in both the AL and NL is 72 games! Owen has also a good batting average
too (above .280).
This year, Tim Wallach shows again his great defensive and offensive talents.
In my opinion, he is a more complete third baseman than Sabo.
Galarraga still has some problems with his batting, but is improving. He still
is the best defensive first baseman.
Dave Martinez has done a great job so far replacing rookie Marquis Grissom.
Maybe he will continue to play on a regular basis when Grissom comes back.
Expos have a lot of depth.
--
Martin Dubuc Local S-242, Pavillon Principal
du...@IRO.UMontreal.CA Universite de Montreal
Tel.: (514) 343-7599
David M Tate
>
>Who knows that Spike Owen is the best defensive shortstop of the NL?
>Last year he had the best defensive average.
Where did you get this info? I haven't yet found 1989 defensive average
statistics. Or do you just mean "fielding percentage"?
--
David M. Tate | "[Science Fiction] diverges from reality no
dt...@unix.cis.pitt.edu | more than economic theory, and is more useful."
|
"A Man for all Seasonings" | -- Walter E. Meyers
Martin Dubuc
>In article <1990Jun19.2...@IRO.UMontreal.CA> du...@ell.UUCP (Martin Dubuc) writes:
>>
>>Who knows that Spike Owen is the best defensive shortstop of the NL?
>>Last year he had the best defensive average.
>
>Where did you get this info? I haven't yet found 1989 defensive average
>statistics. Or do you just mean "fielding percentage"?
>
That is what I meant. It's not always easy to find english translation
to those french expressions commonly used in sport.
Shawn V. Hernan
>at the beginning of the year. Now, Expos seem to be one of the hottest team.
>What suprises me is that there isn't much peoples who know the players of this
>Galarraga still has some problems with his batting, but is improving. He still
>is the best defensive first baseman.
I must disagree. Sid Bream of the Bucs is the best defensive first baseman
in MLB.
But it doesn't matter. The Pirates are in Montreal this weekend for three
games. May the better *team* win.
Go Bucs!
Shawn
--
Shawn Valentine Hernan |Wizard-wanna-be | STOP
Computing and Information Services|Systems & Networks |the war on drugs!
University of Pittsburgh |vale...@unix.cis.pitt.edu| It is a
(412) 624-6425 |vale...@PITTVMS.BITNET | WITCHHUNT!
David M Tate
>In article <1990Jun19.2...@IRO.UMontreal.CA> du...@ell.UUCP (Martin Dubuc) writes:
>
>>Galarraga still has some problems with his batting, but is improving. He still
>>is the best defensive first baseman.
>
>
>I must disagree. Sid Bream of the Bucs is the best defensive first baseman
>in MLB.
I have to agree with Shawn here, and not just because he's at Pitt :-).
One of the biggest surprises about the 1988 Defensive Average numbers that
Pete DeCoursey published was at first base in the NL. Sid Bream had far and
away the best average overall, and the best ground ball average. He also
started double plays *twice* as often as the next best first baseman in the
league.
Galarraga, on the other hand, had the *worst* defensive averages, both overall
and against ground balls, of any NL regular first baseman. He led the league
in triples allowed, and was second in doubles allowed. The only thing he did
well was start double plays (he was second to Bream in DP frequency, 4.9% to
2.5%). And Galarraga won the gold glove!
Despite the fact that Galarraga had a monster year that year, DeCoursey makes
a good case that Bream was the more valuable player. This is because, when
you normalize to the same number of opportunities, you find that Bream would
have taken away 50 more singles, six more doubles, and one more triple than
Galarraga over the course of the season. Add these additional hits into Sid's
batting stats, and compare them to The Big Cat:
avg oba pct
Bream .358 .422 .516
Galarraga .302 .354 .540
I'd take Bream, there, any day.
Unfortunately, it isn't clear that Bream has managed to hang on to his great
range and skill through serious knee surgery. He doesn't *look* very mobile
any more, that's for sure. We'll have to wait and see how the numbers turn
out.
(An interesting aside: Pete points out that, of the 8 first basemen in the NL
who were above average in DP frequency, 7 were left-handed. Conversely, of
the 7 who were below average, 6 were right-handed. What is the cost of having
a right-handed first baseman? Looks to Pete (and to me) like it's a *lot* of
double plays... )
Dave Masten
>Pete DeCoursey published was at first base in the NL. Sid Bream had far and
>away the best average overall, and the best ground ball average. He also
>started double plays *twice* as often as the next best first baseman in the
>league.
Dave, you seem pretty sold on these defensive averages. Perhaps you can share
some of the reasons for this. A few questions:
1) Do they represent fraction of balls fielded and turned into outs or fraction
of balls hit to a certain player zone (whether touched or not?) and turned
into outs? If it represents (outs/fielded) then it penalizes guys who get
to a ball deep in the hole but have no chance to get the runner. If it is
(outs/ball in zone) wouldn't it strongly affected by the pitcher? A hard
shot a few feet past a player who had no chance would adversely affect his
DA. These plays would not avg out between different teams as long as you
accept that pitching counts for something! A comparison of adjacent fielders
on the same team may give some indication of this (bad DA for both 1b and 2b
may indicate pitchers on that team let lefties rifle shots thru the hole).
2) You probably don't have the numbers, but are there consistencies from year
year? I would intuitively expect players to be more consistent in fielding
than hitting. Since we tend to see the same names at the top of BA, etc
every year, I would expect to see the same names (with the similar fldg avgs)
at the top of every year. If not I wouldn't believe the stat represents
a player's defensive contribution (barring explanation such as injury).
3) How do the numbers compare with expectations (subjective ratings of the
fielder's ability as you or anyone you trust sees it). I guess any stat
(even good ones) will come up with some surprises. But if there is no
correlation with observation, I'd doubt it (ex: Gee, I thought Ozzie was
good, but he was 10th in the league. And boy, look how good Jefferies is
at second!).
Just a few thoughts.
Dave Masten
David M Tate
>
>Dave, you seem pretty sold on these defensive averages. Perhaps you can share
>some of the reasons for this. A few questions:
>1) Do they represent fraction of balls fielded and turned into outs or fraction
> of balls hit to a certain player zone (whether touched or not?) and turned
> into outs?
It's balls hit to a particular zone. Some balls, in fact, may be charged to
more than one player (ground ball through the hole at short, charged to both
third baseman and shortstop, for example).
Your point about how this could depend on pitching is noted, and since that
seems to me to be the biggest potential problem, I'll come back to it in a
minute.
>2) You probably don't have the numbers, but are there consistencies from year
> year? I would intuitively expect players to be more consistent in fielding
> than hitting. Since we tend to see the same names at the top of BA, etc
> every year, I would expect to see the same names (with the similar fldg avgs)
> at the top of every year. If not I wouldn't believe the stat represents
> a player's defensive contribution (barring explanation such as injury).
This was exactly my thinking on the subject, which is why I'm waiting on pins
and needles to see the numbers from last year. Sherri Nichols says that Pete
has had the raw data for a couple of months now, so they should be coming out
soon.
Of course, two years is not a huge sample, but if we see no relationship from
one year to the next, that will be a giveaway that something's amiss.
>3) How do the numbers compare with expectations (subjective ratings of the
> fielder's ability as you or anyone you trust sees it). I guess any stat
> (even good ones) will come up with some surprises. But if there is no
> correlation with observation, I'd doubt it (ex: Gee, I thought Ozzie was
> good, but he was 10th in the league. And boy, look how good Jefferies is
> at second!).
Here are the 1988 league leaders in Defensive Average for appropriate infield
positions (outfield has some problems, which I'll get to):
Position American League National League
1B Fred McGriff Sid Bream
2B Willie Randolph Bill Doran (*)
SS Tony Fernandez Ozzie Smith
3B Kelly Gruber Tim Wallach
(*) Jose Oquendo had far and away the best DA, but only played 70 games at
second. You be the judge whether that's enough to count or not...
Personally, I don't find any of these surprising except McGriff and Randolph.
Don Mattingly was fourth, behind Greg Brock and Pete O'Brien, and tied with
Alvin Davis. Harold Reynolds was well below league average. In fact, Billy
Ripken was uniformly better than Reynolds, in range, extra base hits allowed,
double plays initiated, etc.
>Just a few thoughts.
>
>Dave Masten
And good ones they are, Dave.
Let's get back to the question of whether Defensive Averages are meaningful.
There are two major effects which they do not account for: differences between
pitching staffs, and park effects. Park effects include the difference
between turf and grass, size of the park, quality of surface, etc.
Limiting the discussion to infielders for a moment, I don't think pitching
staffs will be that much of a factor. A pitcher who's getting rocked doesn't
give up screaming ground balls, he gives up screaming line drives, and extra-
base hits into the corners/gaps, and so on. These don't get charged to the
infielders at all, and my guess is that they pretty much even out over the
course of the season. Basically, balls are only hit on the ground by mistake,
so any ground ball is as likely to be easy as hard.
On the other hand, astroturf vs. grass makes a big difference. Pete gave
separate home/road stats for NL shortstops in 1988, and there were marked
differences, usually strongly correlated with the home-field surface of the
player in question. The best DA on the road was Garry Templeton, who edged
out Ozzie Smith by 3 points. Smith's DA was 25 points higher at home; Tempy's
was 18 points higher on the road. On the other hand, Kevin Elster went from
.711 at home (4th in the league) to .604 on the road (13th), implying that
perhaps he has sure hands but slow feet. Shawon Dunston was average on the
road (.653), and atrocious in Wrigley (.612). Only Andres Thomas, among the
regulars, was worse.
For outfielders, though, I think that park effects and pitching staffs will
have to be factored in somehow. I'm thinking about it, but the fact is that
I'm much more interested in evaluating infielders than outfielders. We all
know why outfielders are there: to *hit*. Infielders earn their bread and
butter with defense, and it would be nice to be able to quantify just how
much bread and how much butter they've earned.
Finally, I'll reiterate Pete's point: DA doesn't measure ability, it measures
performance. This is what they actually *did*. We don't try to adjust
batting averages to reflect the fact that some pitches are harder to hit than
others; we don't count a double in Fenway as less than two bases. There's no
reason to do that with defense, either.
Sherri Nichols
>some of the reasons for this. A few questions:
Since I'm the one who sold Dave on defensive average, and wrote the program
that computes them, I'll take a stab at answering these.
>1) Do they represent fraction of balls fielded and turned into outs or fraction
> of balls hit to a certain player zone (whether touched or not?) and turned
> into outs? If it represents (outs/fielded) then it penalizes guys who get
> to a ball deep in the hole but have no chance to get the runner. If it is
Defensive average = Balls turned into outs / balls hit to a certain zone
> (outs/ball in zone) wouldn't it strongly affected by the pitcher? A hard
> shot a few feet past a player who had no chance would adversely affect his
> DA. These plays would not avg out between different teams as long as you
> accept that pitching counts for something!
It's very hard to separate that sort of stuff out. Do some pitching staffs
give up a significantly higher percentage of hard hit balls through the
infield than others? It's not obvious to me that this is necessarily the
case. I have no way of measuring this directly; the Project Scoresheet
data only gives location, not "how hard". I believe the Stats data does,
but I'm pretty suspicious of information that subjective. The Project data
does attempt to distinguish between line drives and fly balls, and I've
discovered that no two scorers make the same distinction; i.e. your line
drive is my fly ball.
A comparison of adjacent fielders
> on the same team may give some indication of this (bad DA for both 1b and 2b
> may indicate pitchers on that team let lefties rifle shots thru the hole).
The data Pete publishes includes numbers by team for 1b-2b combined, 2b-ss
combined, 3b-ss combined, and entire infields and outfields. I don't
remember noticing any effects like you describe.
There's a real problem in trying to draw conclusions about pitching staffs
from defensive numbers, and about defense from pitching numbers, and that's
that we don't have a clue as to how to separate the two. Is the defense
bad because the pitching is bad, or is the pitching bad because the defense
is bad?
>2) You probably don't have the numbers, but are there consistencies from year
> year? I would intuitively expect players to be more consistent in fielding
> than hitting. Since we tend to see the same names at the top of BA, etc
> every year, I would expect to see the same names (with the similar fldg avgs)
> at the top of every year. If not I wouldn't believe the stat represents
> a player's defensive contribution (barring explanation such as injury).
You do see consistency from year to year.
>
>3) How do the numbers compare with expectations (subjective ratings of the
> fielder's ability as you or anyone you trust sees it). I guess any stat
> (even good ones) will come up with some surprises. But if there is no
> correlation with observation, I'd doubt it (ex: Gee, I thought Ozzie was
> good, but he was 10th in the league. And boy, look how good Jefferies is
> at second!).
There are some surprises (or it wouldn't be interesting, would it?), but it
also matches up somewhat with expectations. Both Ozzies (Smith and
Guillen) are consistently near the top of the league for shortstops.
Howard Johnson is terrible at third base; Tim Wallach and Terry Pendleton
are near the top (don't have enough data yet on Matt Williams to draw many
conclusions, but my recollection is that he did well in limited time last
year). Gary Templeton is probably the biggest surprise to me; he's still
above average at short. Sandberg is good at second, as is converted
shortstop Oquendo; Jefferies doesn't show up well at second.
The
only problem with matching up with expectations is that for a lot of
people, good defense equals not making errors. There are certain players
who have good hands but limited range who have the reputation of being good
defensive players; these players tend not to do well in defensive average.
Keith Hernandez near the end is a good example. I'm always suspicious of
errorless streaks for that very reason; often those players simply aren't
getting to a lot of balls. Ryne Sandberg is an exception; I'm suspicious
about Spike Owen given his past performance, but I'll check the numbers.
Sherri Nichols
snic...@adobe.com
Sherri Nichols
>separate home/road stats for NL shortstops in 1988, and there were marked
>differences, usually strongly correlated with the home-field surface of the
>player in question. The best DA on the road was Garry Templeton, who edged
>out Ozzie Smith by 3 points. Smith's DA was 25 points higher at home; Tempy's
>was 18 points higher on the road. On the other hand, Kevin Elster went from
>.711 at home (4th in the league) to .604 on the road (13th), implying that
>perhaps he has sure hands but slow feet. Shawon Dunston was average on the
>road (.653), and atrocious in Wrigley (.612). Only Andres Thomas, among the
>regulars, was worse.
Actually, I think it's more complicated than just a simple turf/grass
difference. I think there's quite a variation among turf parks, and a
variation among grass parks, and it can change from year to year. There's
got to be a significant difference between a park with turf that's 10 years
old and a park that just installed the latest, greatest turf. Likewise,
not all grass fields are equal either. In one of his books, Earl Weaver
talks about how they kept the grass at Memorial Stadium really short
because the O's had a good defensive infield, so a faster infield was to
their advantage. Other teams have been known to let the grass grow in the
infield (and outfield; I'm surprised outfielders haven't got lost in that
jungle in Wrigley) to help cover up defensive insufficiencies.
>
>For outfielders, though, I think that park effects and pitching staffs will
>have to be factored in somehow. I'm thinking about it, but the fact is that
>I'm much more interested in evaluating infielders than outfielders. We all
>know why outfielders are there: to *hit*. Infielders earn their bread and
>butter with defense, and it would be nice to be able to quantify just how
>much bread and how much butter they've earned.
The overlap between outfielders is also bigger than it is for infielders.
There aren't that many ground balls that could easily be fielded by more
than one infielder, but there are lots of outfield balls that could easily
be caught by more than one outfielder. It seems to vary from team to team
who takes precedence in such cases. There's also some significant fraction
of outfield play that is concerned with preventing extra bases: cutting off
hits in the gap, all the aspects of the throwing game. Defensive average
doesn't measure this at all. I'd like to know more about outfield defense;
I have a theory that it's actually more important than we think; that, in
particular, an outfield that covers a lot of ground is more important than
an infield that covers a lot of ground.
>
>Finally, I'll reiterate Pete's point: DA doesn't measure ability, it measures
>performance. This is what they actually *did*. We don't try to adjust
>batting averages to reflect the fact that some pitches are harder to hit than
>others; we don't count a double in Fenway as less than two bases. There's no
>reason to do that with defense, either.
Good point.
Sherri Nichols
snic...@adobe.com