NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/rec/sport/baseball/38449

Internet Message Format | 1993-01-27 | 5.4 KB

Path: sparky!uunet!comp.vuw.ac.nz!canterbury.ac.nz!otago.ac.nz!guilford
From: [e-mail]
Newsgroups: rec.sport.baseball
Subject: Re: Cecil signs :-( :-( :-(
Message-ID: <[e-mail]>
Date: 28 Jan 93 14:38:59 +1300
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Organization: University of Otago, Dunedin, New Zealand
Lines: 102


In article <[e-mail]>, [e-mail] (Dr. Dave) blurts out:

> Thus spake [e-mail] (Edward [Ted] Fischer):

>>In article <[e-mail]> [e-mail] (Dr. Dave) writes:

>>  Billy Hatcher and Dan Gladden seem
>>clearly useful, as without them (and their clones) around, major
>>league baseball would only be able to field about 15 teams.

> On the other hand, without them (and their clones), the teams that win
> pennants wouldn't win nearly as easily.  The fact that they may be the
> best available for a given team at a given time does not change the fact
> that they are making it *very* difficult to compete with the good teams.
> Which is all I'm talking about.  For purposes of comparing AGAINST THE OTHER
> TEAMS IN THE LEAGUE (which is what pennants are all about), league average
> is the correct baseline.  W/L record is a zero-sum game among a finite
> collection of teams, where every team gains advantage only by improving
> relative to the quality of the opposition OVERALL, not by improving relative
> to the ability level of the worst players in the league.

1. Okay, Dr. Dave is defining baseball value as wins over .500, rather
than just wins.

2. Any baseline will do in making comparisons among, say, first basemen.
Consider five baselines: worthless, O'brien, replacement, average, Fielder.
The distance between O'brien and Fielder is the same, of course, regardless
of the baseline used to measure their value.

3. Of course, the interesting nugget in this thread is, how to make
comparisions BETWEEN (or *between*, take your pick) positions.  So
does Dr. Dave's valuation method help us answer this question?

4. At first glance, maybe.  Suppose Cecil Fielder and Omar Vizquel are
both one win above the average level (I'm assuming you measure distance
from average in terms of wins).  There is some intuitive appeal to
saying that Cecil and Omar are equally valuable -- the presence of
both on an otherwise average team will make that team 82-80 instead
of 81-81.

5. On the other hand, is that relevant?  The presence of both on
a hypothetical team consisting of nine players -- one for each
position -- who are each one win above average will not change that
team's won-loss record -- it will be 90-72 regardless of whether they
have Cecil or Omar.

6. And now suppose that the replacement level at SS is 2 wins below
average and the replacement level at 1B is 1 win below average.
A replacement level team will get 3 wins by acquiring Omar, 2 by
acquiring Cecil.

7. Furthermore, suppose the Mariners and Tigers are contemplating a trade,
CF for OV, where the M's would play a replacement level SS and the Tigers
a replacement level 1B.  Clearly the M's would lose one win by making
this trade -- they wouldn't make it.

It seems more likely that the M's would be able to trade Omar for a 1B
that was 2 wins above average.

(Market imperfections?  I dunno)

7. Now suppose Frank Thomas becomes twice as good a hitter as he used
to be.  Clearly this drives up the average level at 1B.  Does Cecil
Fielder become less valuable?  Well, how does this impact the Tigers
chances of winning ballgames?  Clearly, it's harder to beat the White
Sox now, so Cecil may seem less valuable....but isn't Lou Whitaker
and Mickey Tettleton just as less valuable?  I'm confused as to how
Frank's improvement changes Cecil's value, especially in making
comparisons with other positions.  Suppose Cecil is now only 0.7 wins
above average.  This doesn't seem to affect the equation in any
potential trades, so I don't see how it affects his value.

(Actually, I do think it affects his value, and the value of all of the
other 1B -- just subtly, and the change in value doesn't seem to me to
be the same as the distance from the average).

> If you choose to field a team in a given league but not to try to win, you
> of course will have a different notion of what a player's use value is.

But why is "try to win" synonymous with .500, with the average level?
It's possible to "win" with a 79-83 record, with below average players.
At the same time it's nearly impossible to win with a 83-79 record.
So this celebrated role for the average level seems to me somewhat
artificial in principle....

>>>Thought experiment:  suppose everybody else's starting 9 suddenly got much
>>>much better, but yours didn't, and the talent not currently in the majors
>>>stayed the same.  Clearly, your players are suddenly much less valuable than
>>>they used to be, because they will win many fewer games than they used to.
>>>But replacement level HAS NOT CHANGED; average level has.

Perhaps it would be true, if there were only 28 infinitely-lived players
for each position, that use value = distance from average.  But I'm
confused.  Why average.  And anyway, this isn't the way it is, careers
are short and some teams have lots of talent at a position and others
have little -- replacement level IS of interest.

And besides, why are use value and market value different?  I still don't
follow....

Bill Guilford