                        1996 Season Disk Notes
                        Written by Tom Tippett
                        ----------------------

Thank you for ordering the 1996 Season Disk for Diamond Mind Baseball.
This note provides some information on the stats and ratings in this
disk.  We also include our comments on the Gold Glove awards that were
handed out this year.


Version 7 update
----------------

Our version 7 update, which began shipping in July, 1997, includes many
new features, some of which require new information in the season disk
files for best results.  As a result, for all customers who purchase the
upgrade to version 7 and who previously purchased the 1996 Season Disk,
we are sending an updated copy of the 1996 Season Disk at no charge.

Here are the new types of information that are included in the version 7
update to the 1996 Season Disk:

- birthdates have been added for all players, making it easy to see each
player's age on the drafting screens (and in other places)

- we have compiled a complete list of real-life roster moves and game-by-
game starting lineups, to make replays with real-life teams and lineups
even more accurate

You do not have to install the new version 7 update to take advantage of
all of the new features, however.  When you use the original 1996 Season
Disk with version 7, the game converts all of the information (real-life
and computer-league) to the new format, and assigns ratings for the new
pitch-by-pitch system.  So, if you are in the middle of a 1996 project,
you can continue to use your original season disk with version 7.  The
only things you will be missing are the birthdates, transactions and
game-by-game lineups.  


The Players
-----------

Prior to 1995, we created player records for everyone with 30 atbats 
or 25 innings pitched.  Last year, and again this year, we have created 
player records for every player who appeared in the big leagues this 
season.

If a player appeared on more than one team in real life, we have created
a player record for each team (for people who do season replays using the
real rosters) plus a combined record (for use in draft leagues).  The
combined records appear in the free agent listings.


Parks and Weather Information
-----------------------------

Two parks recently changed their names, and we're using the new ones.
Joe Robbie Stadium (Florida) became Pro Players Stadium and Riverfront
Stadium (Cincinnati) is now known as Cinergy Field.  With the recent
trend toward selling naming rights to local corporations, we'll probably
be seeing a lot more of this in the next few years.

We've updated our park information to capture the weather as it was this
year.  One interesting discovery is that the wind was blowing out a
little more often this year than last.  This undoubtedly accounts for a
small part of the increase in homeruns, though we don't think it's a
big factor.


Real-life Statistics
--------------------

As always, your DMB Player Disk is the product of extensive research 
into player performance.  We begin with official batting, pitching and 
fielding statistics from The Baseball Workshop.  In addition, we use 
studies of detailed play-by-play information to rate players in areas 
such as running, throwing, bunting, and defensive range.

When I was growing up, I thought that the official stats were absolutely 
perfect, but after years of compiling and purchasing statistics from the 
leading statistical companies, I have learned that this is not always 
true.  I have the good fortune of living within a few minutes of Pete 
Palmer, who compiled the statistics for Total Baseball, and have been 
amazed at the large numbers of errors that he has uncovered in official 
statistics as part of his research.  More recently, it has become 
increasingly clear that published left/right statistics and fielding 
statistics are not always perfect.

In a typical year, there are over fifty situations in which a player 
bats from the opposite side of the plate than he normally does.  In
recent years, for example, some switch hitters batted the "wrong"
way against knuckleball pitchers.  Other switch hitters may bat the
wrong way when facing a pitcher they've really stuggled against.  And
players occasionally start or stop switch hitting during the season.
In 1995, Greg Harris pitched left-handed and right-handed in the same
game.  The system used by the Baseball Workshop handles all of these
cases, and we believe our left/right stats are at least as accurate as
others that have been published.

Sometimes it's not even clear which way a player bats or throws.  In
1996, for example, the STATS Handbook lists Yorkis Perez as a left-
handed batter.  So do some other sources.  But the Florida media
guide and other sources show him as a switch hitter.  We do our best 
to track all of these things down and get them right, but when Major 
League Baseball and STATS and Elias and Baseball Weekly and the media 
guides can't seem to agree, we can never be 100% sure.

We know that many of you are avid readers of the STATS Major League
Handbook and the STATS Player Profiles book.  We have found a number
of small errors in the data in these two books this year.  Many of 
them have been corrected in the STATS Online service that you can 
access through America Online.

In particular, you'll find that the left-right splits on our player
disk are different than those in the STATS Handbook for many NL
pitchers.  We're pretty certain that ours are near-perfect and that
the vast majority of the errors are in the STATS Handbook.  How
can we be sure?  Mainly because the STATS Handbook numbers don't
match the current STATS online numbers, and the updated STATS online
numbers match ours for many of these players.  There are still some 
players for which our numbers don't match the STATS online, but ours 
agree with the Elias Sports Bureau data for almost all of those 
players.

We have also found that there are differences between the fielding
statistics in the STATS books and other sources.  We are beginning 
to think that the STATS and Baseball Workshop numbers are more 
accurate than the MLB official statistics.  At this point, it looks 
like none of them is perfect, but they are all pretty close.


Bunting
-------

A player's sacrifice bunt total tells us how frequently he has
bunted successfully, but tells us nothing about any failures.  For
example, when a big-league player lays down a bunt, the runner or
runners move up about 82% of the time.  So a player who is
successful in 21 of 26 bunt attempts is not an excellent bunter 
despite his total of 21 successful sacrifices.

Further, if you include plate appearances in which the hitter fouled off 
at least one bunt attempt and then swung away, the big-league success 
rate drops to about 57%.  In other words, there are more bunting 
"failures" due to fouled bunts than to poorly-placed bunts.  You are 
missing a lot of important information if you try to measure bunting 
ability without considering how often a hitter fouls off his bunt 
attempts.

We examine each player's bunt attempts and determine how often he 
reached base on a single or error, caused the lead runner to be retired, 
sacrificed successfully, popped up, bunted into a double play, or fouled 
off one or more pitches before swinging away.  We look separately at 
sacrifices and attempts to bunt for a hit.  Our bunt ratings take all of 
these factors into consideration.


Running and throwing
--------------------

There are no official statistics for baserunning, though you may 
occasionally see some data on how often runners go from first to third 
on a single.  Our studies examine how often runners took extra bases 
on hits and fly balls, taking into account where the ball was hit (it's 
much easier to go to third on a ball hit to right), the number of outs 
(getting the jump with two outs makes a big difference), and whether 
the playing surface is grass or artificial turf (ground balls reach the 
outfielder more quickly on turf, so it's harder to take the extra 
base on many singles.)  

Outfield assist totals give us some insight into outfielder throwing, 
but they can also be misleading.  Some great throwing outfielders have 
their assist totals cut down by their opponent's unwillingness to run on 
them in the first place.  And not all assists are created equal.  Some 
assists result when a throw nails a lead runner, while others occur when 
the lead runners score but the batter is out at second advancing on the 
throw.

Our throwing ratings focus on the ability to prevent the lead runners 
from advancing using a combination of skills:  getting to the ball 
quickly, getting the throw off quickly, and throwing with strength and 
accuracy.  Some players with strong arms are not very accurate.  Some 
players with weak arms compensate by getting into throwing position more 
quickly.  As with all of our ratings, we measure results, not just how 
often a player looks good making a throw.


Measuring Defensive Range
-------------------------

Defensive range is one of the hardest elements of performance to 
measure.  But we have made some good progress in recent years, and
I'm very pleased to say that we think we've advanced the state
of the art this year.  Because this is one of my favorite subjects,
I'm going to spend a little more time on it than usual this year.

In the beginning, there were the official fielding stats.  We had 
games, putouts, assists, errors, double plays, and fielding percentage.
But using these numbers to assess player skills is extremely difficult,
if not impossible.  The list of reasons is very long, but they all
boil down to a few things:

* they don't tell you how many chances to make plays were presented
to each fielder

* they don't tell you whether those chances were routine, normal, or
challenging

For these reasons, it's very difficult to measure fielding ability
using stats such as assists per game, putouts per game, total chances
per game, or fielding percentage.  It's sometimes possible to pick out
the very best and very worst fielders using these numbers, but it's
very hard to evaluate the majority of players.  Suppose a shortstop
led the league in assists (as Jay Bell did this year) or total chances
per game (Alex Gonzalez).  Is this because:

a) he played more innings than everyone else?

b) he played behind a pitching staff that doesn't strike out a lot
   of batters, so more balls are put in play?

c) his pitching staff induces a high percentage of ground balls?

d) his pitching staff has a lot of lefties, who therefore face more
   than the normal number of right-handed batters, and right-handed
   batters hit more balls to short than lefties do?

e) it just happened that more balls were hit to short when he was
   playing?

f) he was lucky to get an unusually high number of routine plays, such
   as lazy popups and soft line drives?

g) he's the best shortstop in baseball?

Baseball analysts, ourselves included, have made many attempts to
devise methods that deal with some of these other factors so
that we can isolate the contribution that the player is making.

The first and most obvious attempt deals with playing time.  Not all
games are equal.  Some players play almost every inning of the games
they play.  Others split the playing time with a platoon partner.
Late-inning defensive specialists often pick up a lot of games played
without playing a lot of innings.  So analysts started counting
defensive innings.  This was and is a very good thing, but it answers
only one of the questions posed above.

Using defensive innings to measure playing time, Rey Sanchez pops
ahead of Alex Gonzalez in the rankings, with a range factor (total
chances per nine defensive innings) of 5.40.  Bell is ninth.  So
can we annoint Sanchez as the best shortstop in baseball?  Not yet.

We still don't know how many chances he had to make plays.  In the
1996 National League, 75.8% of the plate appearances resulted in a
ball being put into play.  We arrived at that figure by subtracting
the walks, strikeouts, hit batsmen, and homeruns from the total of
plate appearances.  But Sanchez plays for Chicago, and his pitching
staff saw 78.7% of the batters put the ball in play.  And Bell, who
plays for Pittsburgh, saw only 72.3% of the balls put in play.

And not all putouts and assists are equal.  Catching a routine popup,
taking a throw from the catcher on a steal attempt, or making the pivot
on a double play don't tell you anything about range.  But snaring a line
drive or going into the hole to field a grounder does tell you a lot.

In the mid-1980s, I developed a new type of range factor that adjusted
for many of these factors.  It counted balls put in play while each
fielder was at his position.  It counted only those PO and A in which
a fielder's range was really being tested (such as creating a GB out
by making a throw or an unassisted putout, catching a liner, or catching
a fly ball).  It tracked balls put in play by left- and right-handed
batters separately, since players pull the ball more often than they
go the other way.  Independently, Gary Gillette came up with a similar
system for the Great American Baseball Stat Book.  This system helped
quite a bit, and I used it for several years, even though it didn't
overcome some of the remaining problems.

Some pitching staffs produce 12% more fly balls than the average team,
and some produce 12% fewer.  If you look at the regular CFers in this
year's STATS Handbook, 18 of the 24 players fell within +/- 12% of the
league average.  Without knowing what type of staff they were playing
behind, you have no way of knowing how much of their range factor comes
from ability and how much comes from having more or fewer opportunities.

In 1996, the average major league pitcher induced 1.30 ground balls for
every fly ball.  (That's according to the STATS Player Profiles book.)  
Chicago's staff was above that, at 1.44, and Pittsburgh was tied 
for second highest at 1.55.  Sanchez may indeed be a great shortstop, 
but how much of his league-leading range factor results from his 
ability and how much from playing behind a staff that puts the 
ball in play and keeps it on the ground?  Minnesota's staff was an 
extreme fly-ball staff, with a ratio of 0.89.  It's no surprise that 
every Twins outfielder is at or near the top in range factor.

The next logical step is to use a system that counts actual 
opportunities to make plays.  We weren't able to do that until 1989,
because until then, nobody tracked the location of every batted ball.
The good folks at STATS were the first to do it, and they were quick
to develop the zone rating to take advantage of this new information.

The zone rating should have been a tremendous breakthrough, but STATS
made some serious errors in designing this statistic.  Unfortunately,
we used the zone rating to rate our fielders for a couple of years
before we learned about the problems with it.  

STATS says the "zone rating measures all the balls hit in the area
where a fielder can reasonably be expected to record an out, then
counts the percentage of outs actually made."  This is a step in
the right direction.  Instead of having to infer the number of
opportunities to make plays from defensive innings, percentages
of balls in play, left-right composition of the pitching staff,
and the staff ground-ball/fly-ball ratio, we can actually count the
balls hit to each fielder while they are in the game. 

The first problem is that they don't count all the balls.  For
example, nobody is charged with an opportunity when a GB is hit
down the 3B line, between SS/3B, up the middle, between 1B/2B, or
down the 1B line.  In fact, the average shortstop zone rating was
.935, meaning that they only count the balls that over 90% of the
shortstops get to.  So the plays that go into the zone ratings are 
the ones where the ball is hit more or less at the fielder.  The 
ones that are left out are the ones that only the best fielders 
get to.  The net result is a system that rewards good hands over 
real range.

The second problem occurs when an infielder starts a double play.  
STATS credits him with two outs and one opportunity.  Bip Roberts, for
example, has a 1.013 zone rating for 1995, meaning that he created an
out 101% of the time.  This approach gives a significant boost to 
infielders who play behind pitchers who put lots of runners on base 
and/or with a pivot partner who turns the DP well.

Once again, let me say that the idea behind the STATS zone rating is
basically sound.  Done properly, it would be an improvement over the
adjusted range factors I was using in the 1980s.  But these two
problems are enough to make it worthless for evaluating player range.

We moved on to another type of zone rating that is called Defensive
Average.  It was developed by Pete DeCoursey and Sherri Nichols and
uses play-by-play data from the Baseball Workshop.  Defensive average
uses the same principle of counting batted balls hit into a fielder's
zone and counting the number of plays made.  But it covers the whole
field and doesn't mix apples and oranges by double-counting double
plays.  

We felt we got much better results from defensive average (DA) than
from zone ratings.  But DA isn't perfect either.  One of the perplexing
problems is how to assign responsibility for balls hit between fielders.
In both the STATS and DA systems, the player making the play gets one
opportunity and one play.  But things get tricky when the ball falls
in for a hit.

In DA, each player gets half an opportunity and no play when the play
results in a hit.  That means that someone playing next to a weak fielder
tends to look worse than he is, because if the other guy had made the
play, there would be no opportunity, but if the ball falls in, it costs
him a half of one.  In past years, as we've done the ratings, we've been
aware of this limitation and have tried to make intelligent adjustments
to compensate for it.  But we've always wanted to see if we could do
better.

For 1996, we've developed a new approach to computing zone ratings that
does a better job in three areas -- it counts all batted balls, it
handles the plays between fielders better, and it takes into account
the difficulty of the play.

Other systems, like the STATS zone rating and DA, ignore certain
types of batted balls.  Bunts are excluded from both systems, so we get
no help with catchers.  And a significant type of play is left out for
pitchers, third basemen and first basemen.  Popups are excluded from
both systems, in the belief that all popups are routine and don't
therefore measure range.  But some popups, like the not-so-high ones
that an infielder must go back on, are plays that only the best
fielders will make.  The system we've developed counts all batted
ball types and covers the entire field.

It handles plays between fielders better because it calculates the
percentage of those plays that are made by the average fielder at
each position.  Instead of arbitrarily assigning half an opportunity
to each fielder, it assigns responsibility more fairly.  If a line
drive between short and third is handled 20% of the time by the 3B,
25% of the time by the SS, and is a hit the rest of the time, these
are the percentages used to assign responsibility.  Each player
gets credit for the number of plays made compared to the league
average for his position given the mix of batted balls he faced.

This approach also takes into account the difficulty of the play.
Other systems charge the player with one opportunity for every ball
hit into the zone, regardless of type (grounder versus line drive)
or location (at the fielder or in the gap).  If a popup to short is
caught 99.8% of the time by the shortstop, and the shortstop makes
the play, he gets credit for 1 play and .998 opportunities, or .002
over the league average.  If a grounder in the hole is turned into
an out 27.6% of the time by the shortstop, and the shortstop makes
the play, he gets credit for 1.0 - .276 = .724 over the league
average.  This way, a fielder isn't penalized if he happens to face
a tougher array of chances, and he gets extra credit for making the
tough plays.  Of course, it also costs him more if he fails to make
the easy plays, and that seems fair, too.

Before I move on to talk about some of the results of using this
new system, let me take a moment to discuss some remaining challenges
in the seemingly never-ending pursuit of a better fielding metric.

None of these systems account for park effects.  Coors Field affects
outfielder ratings since the ball doesn't stay in the air as long at
that altitude.  Some balls that can be caught in Tiger Stadium will
hit high on the wall in Fenway.  Some infields are so choppy (Mile 
High in '93/'94) that 60% more infielder errors are made there.  This, 
in my opinion, is where the next innovation in fielding ratings needs to 
be made.

And we may never be able to separate the contributions of the pitcher
and the fielder.  If a ball drops in, is it because the pitching was
bad or because the fielder failed to make the play?  Our new system
helps a little here.  If a bad pitching staff showers its defense with
lots of line drives, that's ok.  Our system measures fielders on how
often they field those line drives and doesn't penalize them for facing
more than the average number of line drives.  But I still think good
pitching will make fielders look good, and we keep the quality of the
pitching staff in mind as we assign our ratings.

My goal with this article is to help you understand what you can and
can't rely on when you're looking at fielding data in the newspapers,
in books, and on the Internet.  These are the ideas I hope you'll come
away with:

* range factors should be ignored.  They're accurate in some cases,
but there are so many ways they can be hopelessly biased that you
have to be really, really careful in how you evaluate them.  Simple 
rankings, such as the ones in the STATS books, are highly misleading.  
Adjusted range factors are better, but still mostly useless if you don't 
mentally adjust them for the GB/FB ratio and other possible distortions.

* the STATS zone ratings for infielders should also be ignored.  They, 
too, are accurate in some cases, but hopelessly biased in others.  The 
DA figures are better, because they're based on the same ideas but lack 
the fatal flaws in the STATS system.

* we think our new system is better than the STATS zone ratings and DA,
but we're not saying it's perfect, and frankly, it may never be perfect.
But it's infinitely better than relying on the media to tell us who can
play.  (More on this later.)


How We Assign Range Ratings
---------------------------

It should be clear to you by now that we do a lot of statistical
analysis in the process of assigning our range ratings.  Before
we get into this year's results, let's take a moment to discuss
other sources of information that we *could* use.

We could rely on the opinions of sportswriters and members of the
broadcast media.  This is problematic for several reasons:

* when they say "he's the best second baseman in baseball," it's not
always clear what they mean.  They could mean he's the best overall
player at his position (including hitting, running, etc.).  They
could mean he has great hands.  They could mean he turns the DP well
or that he has great range.  They probably mean all of these things
to some degree, but an overall evaluation doesn't help us.  We have
separate ratings for separate skills, and we need opinions about
each skill.

* they don't talk about all the players.  We have 1100+ players to
rate each year, and only a small number of them are frequently
discussed in the media

* they may be biased toward the teams they cover or the players who
play for teams in media-intensive cities

* they don't see many of the players, so their opinions are often
second- or third-hand

* they may be slow to pick up on a change in performance.  We've been
doing this for ten years now, and we've found that the media is often
18 months to two years behind the numbers.  For example, if the numbers
show that a player is losing a step, we'll start to hear about it two
years later.  In the 1980s, we routinely gave Eric Davis Ex in center
but dropped it to Av after the numbers showed that he'd been slowed by
his knee injuries.  Two years later, the media started to say the same
thing.  We started giving Barry Larkin an Ex rating two years before
the media figured out that were other shortstops besides Ozzie Smith.

* they are in the entertainment business.  It's in their interest to
make someone out to be the greatest ever so people will want to watch
or listen or read about the games they cover.

We could rely on the opinions we hear from other players, managers, and
team executives, but they don't see all the players either, and their
remarks are influenced by the needs of the team.  It's to their
advantage to talk about players in certain ways, whether it's to hype
someone for marketing purposes, or to talk them down in a salary
squabble.  We need something less prone to bias.

We could use the opinions of professional baseball scouts.  This is better
than using the media because scouts are trained to see things that other
people don't see.  But it's difficult to find a collection of scouts who
have seen every player and can make their evaluations available to people
outside the organizations they work for.  And, of course, scouting is
not an exact science either.

We could base our judgments on how often someone shows up on SportsCenter
or This Week In Baseball.  But the photogenic play isn't always the best
play.  The exact same fly ball might produce a routine play for a great
fielder, a diving catch for the average fielder, or a single for the poor
fielder.  The diving catch is the one that makes the highlight films.

We could give a lot of consideration to Gold Glove awards.  But we need
to be careful here, too.  We can't award a player an Ex range rating
just because he won the Gold Glove.  There are two main reasons for this.
First, the Gold Glove is given for overall fielding performance, including
range, throwing, and avoiding errors.  Our fielders have separate ratings
for these three factors.  It should not be surprising that a fielder may
occasionally win a Gold Glove by virtue of excellence in throwing and
avoiding errors, while having only average or above-average range.  Second,
the voters do not have access to the information we compile in our fielding
studies, and we believe they make mistakes.

So, how DO we approach the ratings process?  We start with the best
statistical analysis we can find or produce ourselves.  This provides
complete coverage of all players without the biases associated with
other types of player evaluation.  But we also watch ESPN.  We watch
a lot of games off satellite.  We read all the baseball magazines and
books.  And when the numbers don't match the player's reputation, we
look very, very hard for the reasons why.  We talk to people who
really know baseball and have seen the player quite a bit.  And we'll
sometimes nudge a rating up or down if we feel there are reasons why
the numbers might not be telling the truth.

As we've said earlier in this document, we began using a new type of
fielding analysis this year.  (We really need a name for this.)  We
feel very strongly that it is a significant improvement over what we've
been doing in the past.  That means two things.  First, we have more
confidence than ever in our ratings this year.  Second, we made some
ratings mistakes in the past because the older methods weren't as good.
The following paragraphs will discuss some of the players that have
attracted your attention in the past and some that might prove to be
controversial this year.  We'll do this in the context of our annual
review of the Gold Glove awards.

* Ivan Rodriguez and Charles Johnson, catchers.  No arguments with the
voters on these two choices.  Both scored well in fielding bunts in our
analysis this year, and nobody throws better than these two.

* J.T. Snow, 1B.  Snow has a great reputation, but has never even
showed up as average on any statistical measure.  He's been quite
a bit below average for all of his young career.  His fielding average
is good but not great.  I don't get it.

* Mark Grace, 1B.  An excellent choice.  We've never rated him worse
than Vg and he's earned the Ex rating just about every year.  And he
was tied for the league lead in fielding average in 1996.

* Roberto Alomar, 2B.  Second in the AL in assists this year, tied for
third in fielding average, and has a reputation as one of the best of
all time.  But in eight full seasons, we've never rated him higher than
Vg.  Why?  Zone ratings, defensive average, and our new system have
consistently placed him in the middle of the pack.  He's played in San
Diego on grass, in Toronto on turf, and now in Baltimore on grass.  He's
been on good teams and bad ones and in both leagues.  And the numbers
say the same thing every year.  He makes an average number of plays given
the opportunities presented to him.

Now, we've seen the highlight films, too.  He looks terrific when he's
taking that ball in shallow RF and spinning to throw to second.  He's
a tremendous athlete with great body control.  But our new system has
uncovered something interesting.  For each of the past five years, he's
had very good results going to his left.  But he's well below average
on balls up the middle and in the middle of the pack on balls hit
more or less at the second base position.

* Craig Biggio, 2B.  Led the NL in assists for the third year in a row,
so there's no doubt that he's making a lot of plays.  But that's mainly
because he plays a lot (162 games this year) and does so behind a staff
that gets grounders.  He is getting better with experience at the
position (he started as a catcher) and is right around the Vg/Av boundary
now.  We rated him Fr last year, but our new system shows that to be a
mistake.  He should have been Av.  (By the way, his 1996 STATS zone
rating is awful.)

* Robin Ventura, 3B.  Tied for second in the AL in fielding average and
made his share of plays.  According to our analysis, he was outstanding
in 1992-3 and still very good the past three years.  But there are few
contenders in the AL.  Wade Boggs has made himself into a very good
fielder, and we think he was Ventura's equal this year.  Scott Brosius
is an up-and-comer who recently started to get some media attention
for his defense, but he played only 109 games at third this year.  Ventura
is a good pick.  Boggs would have been fine, too.

* Ken Caminiti, 3B.  Has an amazing arm and made a few of the best plays
we've ever seen at third.  Played through an injury for much of the season,
so he gets a lot of points for heart.  But his fielding percentage was
below the league average, and at age 33, he no longer has the range he 
showed a few years ago.  Our vote would have gone to Shane Andrews, who 
led the majors in range (by our methods) without making a lot of errors.

* Jay Buhner, OF.  Yes, he has a strong arm.  Yes, he made only three
errors.  But his range is very limited.  Was this year.  Has been for
the past five years.  But who else is there?  Most of the good RF are
in the other league right now.  Michael Tucker is better, but he
didn't play much.  Shawn Green is better, but he's not a great 
outfielder.  What I don't understand is why they didn't go for another
CF.  There's no obligation to pick a RF.  Johnny Damon would have been
my third outfielder if I had a vote.

* Ken Griffey, OF.  A wonderful pick.  You may be surprised to hear
me say this, since we've been trying to shoot down his candidacy for
the past five years.  But we were only partly right and more than a
little wrong about this guy.  Our new system does a better job of
isolating performance on balls hit between fielders.  If we'd had this
tool the past five years, we'd have rated him Av/Av/Vg/Ex/Ex instead
of Av/Av/Av/Vg/Ex.  His ratings must have been pulled down by having
Mr. Buhner to his left all that time.  We still believe he hasn't
always been the best CF in the league (at least not while Lance
Johnson and Devon White were still in their prime and in the AL),
but certainly is now, and by a good margin.  Not only is he one of
the great hitters of our time, but we figure he took 41 hits away
with his defense this year.

* Kenny Lofton, OF.  We're more than a little puzzled by his performance
in recent years.  He still steals a ton of bases, so with the exception
of last year (when he was slowed for a while with a leg injury), it
appears that his legs still have a lot of jump in them.  Our results
show him with above-average range, but nothing near the reputation
that he has.  His numbers don't look that different from those of
Rich Becker and Brady Anderson.  Perhaps playing next to Albert Belle
is hurting him.  But there's a real shortage of good-fielding outfielders
who played a lot in 1996, so Lofton remains a good choice for the Gold
Glove.

* Marquis Grissom, OF.  Had a STATS zone rating fifty points below the
league average.  We think he's lost something in the last two years, but
we continue to show him as well above average in range.  Among everyday
players, though, we think Lance Johnson was better this year.  But
Grissom was second, and we can see why he was chosen.  There's an
excellent crop of RF this year (headed by Jordan, Sosa, and Sanders),
but until they change the rules and select players from each outfield
position, it's hard to argue against choosing a very good center
fielder over an excellent right fielder.

* Barry Bonds, OF.  Among the LF who played 140+ games, Bonds is best.
There are other good ones who played a little less (Luis Gonzalez,
Rickey Henderson, Ellis Burks), but it's hard to argue with this choice.
Our analysis shows that Bonds isn't quite the Ex fielder was in his
prime, but he's still a solid Vg who's closer to Ex than Av.  We bumped
him down to Av the last two years, but our new system shows that to be
a mistake.  He should have been Vg the last two years.  

* Steve Finley, OF.  Not quite the Jay Buhner pick of the NL, but not
easy to defend, either.  Finley has average range.  Lance Johnson is
better.  So is Devon White.  And Ray Lankford.  And I'd rather pick one
of the top RF than go with a middle-of-the-pack CF.  Finley didn't even
show up well in traditional stats such as fielding average or assists.

Well, these are the Gold Glove winners, but there are a few other players
we'd like to comment on while we're at it.  Each of these players has
a reputation for being a very good or excellent fielder but doesn't seem
to make as many plays as you might think.

Gregg Jefferies -- we were surprised when he showed up near the top of 
our charts this year.  You may recall that he started his career as a 
2B/3B, so he has some athletic ability.  It appears that we underrated 
him last year, as our new system shows that he has improved each year
since he made the switch to 1B in 1993.

Hal Morris -- we rated him Ex last year and Av this year, and it appears
that we were wrong last year.  He got that rating because the Defensive 
Average and Zone Rating systems both made him look very good, but our new 
system shows that he's average now and always has been.  It also shows
that Boone and Larkin deserve more of the credit for the good Cinci 
infield defense than Morris does.

Delino DeShields -- his game seems to have fallen apart in many ways.  
He's stopped hitting and his defense has suffered, too.  He still runs
well.  

Matt Williams -- a shoulder injury forced him to switch to 1B for a 
number of games this year, and this may have affected his ability to 
play third for a while.  He didn't stand out from the pack like he
has in recent years.

Ray Ordonez -- entered the league with more hype than I can remember
in thirty years of following baseball, but by mid-season a lot of 
folks were talking up Edgar Renteria as the best young defensive whiz
in the league.  Although Ordonez made some spectacular plays early in
the year, there's absolutely no evidence that he made more plays than 
the average shortstop this year.  He's right on the league average in
all three systems.   

Cal Ripken -- early indications are that he's going to be a terrific
third baseman.  The O's are wise to make the move now, because his
range is slipping dangerously close to Fr territory.

Tim Salmon -- an interesting case.  According to defensive average 
and the zone rating, he has looked terrific some years and only a little 
above average in others.  We don't like seeing this type of volatility,
because it makes us wonder whether there's something quite random going
on.  According to our new system, he's been average all along but slipped 
noticeably this year.  


Summary
-------

All in all, we feel the best way to rate players is using informed 
judgment.  We believe it is essential to be informed, so we use very 
sophisticated analytical techniques and the best information available.  
But judgment is essential, too.  Many players do not get enough playing 
time in a single season to enable us to reach conclusions based on 
numbers alone.  Sometimes the way a player is used can produce a biased 
view of his abilities.  So we always look at the player's results in the 
context of his team, career (including minor league records for younger 
players) and an overall understanding of baseball.  We hope you are 
pleased with the results.

