A Comparison Between the Mets and the Yankees

1
Is Money Everything?
A Comparison Between the Mets and the Yankees
Many baseball fans criticize the New York Yankees
for buying the best players in Major League
Baseball and starting their own sort of regime.
Baseball fans often accuse George Steinbrenner,
the owner of the team, of simply wanting to
dominate all of baseball by throwing as much
money as the franchise can afford at the most
desirable players. Do this mean that money
controls the system and players are no longer
playing for JUST the love of the game? For
comparisons sake, lets use the New York Mets as
an example. The graph to the right is an
overlying histogram comparing the different
salaries for the two teams during the 2002
season. The blue lines are the Yankee salaries
and the red are the Mets. The mean salary average
for the Yankees was 4,818,792, while the Mets
averaged 3,639,754. So, because the Yankees make
more money (on average), does this mean they will
be a statistically better team? One way to try
to figure out an answer is by doing multiple
confidence intervals in the different aspects of
the two teams, i.e. pitching, home runs, RBIs.
VS.
Pitching
Clemens
Wells
Mussina
Pettitte
Hernandez
Leiter
Astacio
Trachsel
Estes
DAmico
Lets first look at pitching. What constitutes a
good pitcher? There are a variety of different
factors, but among them, for the purpose of this
project, we will examine ERA, home runs given up
and strike outs. I have selected the top five
pitchers from both teams and have created a
table, which is on the left. The null hypothesis
is that there is NO difference in stats between
the two teams and that they are equal in
standings. By doing a 95 confidence interval
for each of the categories, we can find out
whether or not there is statistical evidence in
favor of the null. Throughout these CIs, if 0
falls within the interval, then we have evidence
in favor of the null. Using Mini-Tab, I performed
a two-sample t-test for the two pitching squads
ERA. The computations gave me a CI of (-1.345,
0.517) with a P-value of .318. Because 0 DOES
fall within the CI, I have my first piece of
evidence to keep the null. Because the P-value is
so high is also another sign to keep accepting
the null. The same results apply for strike-outs.
Using the same two-sample test, the confidence
interval that Mini-tab gave me was (-38.1, 77.7)
with a P-value of .445, even stronger than the
first test. For the final test
I conducted, my results were (almost not
surprisingly) even more in favor of the null than
the first two. My 95 CI was (-14.22, 8.62) with
a P-value of .580. Therefore, amongst the
pitching staff, we have evidence in favor of the
null in that the Yankees do not simply buy up all
the best players.
Batting
VS.
 
 
The Yankee franchise has always been criticized
for finding the best batters and offering them
large salaries in order to come play for the
team. Does this mean that the Yankee batters
will
Williams
Giambi
Jeter
Posada
Soriano
Alomar
Vaughn
Piazza
Alfonzo
Cedeno
always do a better job simply because they are
being paid more? Looking at the batting
statistics for both teams, I decided that the two
most logical categories to test on would be RBIs
(runs batted in) and total number of hits for
each player. Observe the table to the right to
see the top five batters for each teams
statistics. Looking at the numbers for the Mets
(M) and Yankees (Y) at first indicate that there
might be
some evidence against the null hypothesis, but
only further testing will confirm this fact for
us. Lets go back to the Two-sample T-test and
confidence intervals. First, lets look at the
total number of hits. Using Mini-tab, the
confidence interval I got was (-79.7, -6.7) and a
P-value of 0.029. These results are very much
different than the results obtained from the
pitching statistics, as these new computations
provide evidence AGAINST the null hypothesis. To
confirm our finding, lets look at RBIs to see
if the pattern follows. Using Mini-tab again, I
got very similar results as the total hits test.
The confidence interval is (-65.2, -6.8) with a
P-Vale of 0.023. So once again, we have evidence
against the null hypothesis and in favor of the
alternative hypothesis, in that players might be
more attracted to a particular franchise, in this
case, the Yankees, due to offers of high
salaries.
Conclusions
By looking at the results from the various
T-tests that I have made, we can draw several
conclusions. It would seem that the pitching
squads are not as concerned with salaries as the
batters are. As you can see from the pitching
table above, many of the numbers are very similar
and there is not a huge difference in statistics
between the two teams. This is despite the fact
that the mean salary for the five Yankee pitchers
in 2002 was 7.5 million, while the mean salary
for the Mets was 5,958,000, a difference of
about 1.5 million. The Yankee franchise,
however, seems to be much more concerned with
getting better hitters, as you can see from the
striking difference in total hits and RBIs. The
mean salary for the five Yankee batters was
7,526,000 and the average for the Mets was 7.12
million. At first, one might question why there
is a larger difference for the pitching staff
rather than the batters, when clearly the Yankee
roster had better stats than the Mets. This
difference is due to a lurking variable in
Alfonso Sorianos salary, which is only 630,000,
while Williams, Giambi, Jeter and Posada each
make 12 million, 8 million, 13million, and 4
million, respectively. The reason for this
drastic difference in salary is because the 2002
season was only Sorianos second year playing
professional baseball. In summary, it would seem
that the Yankee franchise is very much concerned
with recruiting excellent batters, and those
batters are more than willing to play for them,
especially for a team that has won multiple world
series and is considered to be the best
franchise is Major League Baseball history.
Eric Raicovich Quantitative Reasoning Prof. B.
Hartlaub May 6, 2003