Sir, I Say, Sir, You Have Offended Me

They scored 965 runs, 5.96 per game, easily the highest total in the American League.  They outscored their nearest competitors, the Rangers (who played in a bandbox) by 25 runs.  But the Yankees also dominated on the other side of the ball, giving up just 656 runs on the season, or 4.05 per game.  That was also good for first in the American League, but they prevented almost 70 runs less more than the next best team, the Boston Red Sox.

Let’s put that in some perspective.  The 2009 Yankees also led the AL in scoring, and did so by 30 runs.  However, their pitching staff allowed 753 runs (4.65/game), good for 6th.  The entire American League in 2009 allowed 4.75 runs per game.  Meanwhile, in 1998, the leagues scored 5.01 runs per game.  So, in a higher scoring environment, the Yankees of 1998 still managed to allow .6 runs per game less than they did last year.  Their team ERA+ was 116.  They allowed 24% fewer runs than the average AL team.  The 2009 team allowed 2% fewer runs than the average AL team.

Now, as Will points out, the 2010 Yankees have improved over the 2009 version, and last year’s Yankees won 103 games.  However, their Pythagorean record (based on how many runs they scored and allowed) was just 95-67 (which, coincidentally, was Boston’s exact record).  We are left to conclude that, while excellent, the Yankees significantly outperformed their stats, and were much closer in quality to the Red Sox.  To believe that the 2010 Yankees are better than the 1998 Yankees, you essentially have to believe that the team has improved by 13 games (the ’98 Yankees had a Pyth. record of 108-54).  Yes, Javier Vazquez is better than the Sergio Mitre, Chad Gaudin, Phil Hughes, and Chien-Ming Wang hydra.  However, Hughes is back in the rotation (and has had a 6.02 ERA as a starter over the last two seasons), replacing Joba.  Despite The Hut’s struggles, his 4.18 ERA as a starter over the same time period is far better than what Hughes has shown.

And as much as you might want to credit the Yankees for adding Johnson (who has a significant injury history, and is no guarantee to play 120 games) and Granderson, you have to subtract the contributions of Damon and Matsui.  Last year, according to Fangraphs, Matsui (2.7) and Damon (3.1) combined for a 5.8 WAR.  Johnson (2.5) and Granderson (3.4) combined for a 5.9.  And that was with Johnson playing 133 games.  Granderson’s likely to improve, I think, and Granderson/Johnson (again, provided Johnson’s healthy) figure to be better in 2010 than Matsui/Damon.  But it’s difficult to argue that the 2010 Yankees significantly improved their on field product over the 2009 Yankees with these moves.  Plus, you have to hope that Brett Gardner can duplicate the Gardner/Melky platoon from last year.

Yes, having ARod back for the first month of 2010 will help, but ultimately you’re talking about a swing of maybe a win or two.

Looking at the math, you’re essentially arguing that Vazquez + Hughes – Joba + Johnson + Granderson – Damon – Matsui – .5Melky + 1 month of ARod > 13 wins.  Plus, you have to deal with the additional injury problems that crop up this year with the aging core of Rodriguez, Jeter, Posada, Rivera, and Pettitte, and relying on Johnson.   The 2010 Yankees may outhit the 1998 Yankees, but taken as a whole the 1998 World Champs are better and deeper, and will likely be the best Yankee team you will see in your lifetime. 

Go ahead, prove me wrong.

13 thoughts on “Sir, I Say, Sir, You Have Offended Me

  1. 1 thing here, I see you used WAR numbers over a 1 year period which I find useful, but not as representative as if you used them over a 3 year period, for consistency and to help mask some blips.  In this case, Granderson/Johnson = 22.9 vs Damon/Matsui = 15.0.  7.9 is a lot more significant if you use that in that context.  So we are looking more at 2-3 wins right there.  Johnson will NOT be asked to play the field, where he has had to in the past in the NL.  I also think the defensive upgrade you get from Gardner vs the Melky/Gardner platoon and relatively little difference in the stick is also an upgrade.  The 1 month of A-Rod could be significant as well.  The man had 30 HR and 100RBi and missed a month.  If we were talking Paul Konerko here, and admittedly, Konerko is good, the difference would probably be negligible, but we are talking about A-Rod here.  I think given Hughes age and his progression with his pitches, we can only think he is going to be improved in the rotation this time around. 

    Don' t the Yanks play in a bandbox now when looking at the runs prevented stats?  What were the adjusted numbers?  I can't seem to locate them. 

  2. I worry about using W-L records in this instance. I know that's the ultimate measure of a team, but is it comparable across years? There's a distinct difference in team quality. There's a difference in schedule structure with the unbalanced schedule. It seems that while the 2010 Yankees might not win 109 games, that the 1998 Yankees would not necessarily win 108 games either if they played in this division. I don't find comparing W-L records here useful. The 2010 Yankees could still be better even if the record doesn't reflect that.

  3. "That was also good for first in the American League, but they prevented almost 70 runs less than the next best team, the Boston Red Sox."

    Pretty sure you meant to say, "…but they prevented almost 70 runs MORE than the next best team…"

    Or you meant to say, "…but they ALLOWED almost 70 runs FEWER than the next best team…" 

    The way you wrote it makes it sound like the Yankees were 2nd in run prevention.  For what it's worth.

  4. @Matthew

    Absolutely right.  Damn my grammar.


    @ Mark

    Good question.  I worry about W-L too, but the difference between the teams' W-L and Pyth. records are huge, and I think that's relevant.  How relevant is a legitimate question.  I think we'll be looking at that soon.

  5. It seems to me we are comparing apples and oranges: what the 1998 Yankees *did do* and what the 2010 Yankees *might do*.

    A more valid comparison, it seems to me, would be either a) after the season we can compare which we think is better, based on what actually happened, or b) we compare what we thought about the 1998 Yankees team as the season just started. What were the indications that that team would be a 114 win juggernaut?

  6. @ eddy:  3 year samples make sense when you look at stats like UZR to regulate the chances a player has to prove himself.  However in WAR the playing time is accounted for, and injury history like pulling a calf muscle has been factored in and is unlikely to happen again, at least in a predictable manner.

  7. Great it did not like my email address so I have to write all this again lol.

    If I am not mistaken, and I have not looked at it in awhile, WAR is a mixture of the following components:

    Woba = obp and slg * 2 divided by a league average factor of 3.

    for defense it uses UZR coupled with the Tango Defensive Positional Adjustments. 

    Than it incorporates VORP.

    I am not sure where WAR is adjusted for injury history and anything else significant that would make it any better than any other 1 year predictor.  Can anyone comment on this?  I think maybe instead of cumulative an average WAR over a period of time might make more sense but I am not sure. 

  8. Eddie,

    I misspoke a bit, I'm trying to say that WAR takes into account the number of games played over a season, so if someone has a freak injury, including those numbers in any kind of predictive way would ruin the measure.  I think that's what you meant to correct when you start averaging over years, but if that's what you want to do there is probably a rate statistic that will project with more predictable results.  However, for someone like Nick Johnson, your idea may have some value…

  9. Mr. Common Man –

    I hesitate to call you The Common Man, because most guys with "The" in their names are people I don't want to mess with: Attila The Hun for example, or Vlad The Impaler, or Mott The Hoople.

    I read your piece and Will's piece, and came here to agree with you, based mostly on my sense that the current Yankees are too old a group to match the historic greatness of the 1998 team.  Then I did some research.

    The 1998 Yankee roster was an older group than the 2010 version: 30.4 years old on average, as compared to 28.3 for the current model.  True, some of the age on the 1998 group came from bench players like Chili Davis, Tim Raines and (yes) Joe Girardi.  But Paul O'Neill was 35, D. Strawberry (DH) was 36, Orlando Hernandez was 32, and the Davids (Cone and Wells) were both 35. 

    In contrast, the current Yankees starting outfielders are all under 30, as is the right side of the infield.  Pettite and Vasquez are, on average, the same age as Cone and Wells were in 1998.  Mariano is older than he used to be, but the principal guys backing him up (Robertson and Chamberlain) are under 25. 

    I won't argue that some key cogs in the current Yankee machine are a bit worse for wear, and that the team would struggle if they had to rely too heavily on the Marcus Winns and the Ramiro Penas.  But that's true for every team in baseball, pretty much every year.  If Jeter had been hurt in 1998, he would have been replaced by Dale Sveum or Luis Sojo. 

    Even the 1998 team's pitching looks less wondrous on closer look.  Remember that the number 4 starter on the 1998 team was Hideki Irabu.  You have to be careful lavishing too much praise on a pitching staff where Hideki Irabu takes the mound every fifth day.  Irabu had a decent year in 1998, meaning that this was his career year.  Andy Pettite had an off-year in 1998.  Rookie 28 (er, I mean 32) year old Orlando "El Duque" Hernandez pitched well, but in only 21 starts.  So the dominant starting pitching really came from the 35 year olds, Wells and Cone.  Wells and Cone were very good pitchers with very good careers, but I think we should hesitate before putting them in the same league with Koufax and Drysdale, or Spahn and Sain.

    I know it's not scientific or sabermetric, but if I exclude the guys who appear on both the 1998 and 2010 rosters, 2010 has the advantage: Teixeira over Martinez, Cano over Knoblauch, A-Rod over Brosius, Gardner over Curtis, Sabathia over Wells and Vazquez over Irabu.  I'll give the 1998 team Williams over Granderson (Williams had a helluva year in 1998), O'Neill over Swisher, Cone over Burnett (that one is close), and El Duque over Hughes.  That's 6-4 in favor of 2010.  Then you then need to consider the 1998 versus the 2010 versions of Posada, Jeter, Pettitte and Rivera.  You'd only need two of these four guys to put up numbers like 1998 (or 2009), and you'd end up with a 2010 team that compares pretty well with 1998.

    This is not the conclusion I expected to reach when I started writing this post.  I still agree with you that Will has a lot of nerve comparing the 0-1 2010 Yankees to a team that won 125 games 12 years ago.  My feeling is that Will's numbers somehow fail to capture how well that 1998 team played on the field, but hell if I can find any fault with the numbers he's using.  I fall back on my old-school gut feeling that this team is not as good as the 1998 version, and gut feelings don't cut it around these young stat-heads.

    So I vote for Will.

    I thought about posting under Will's argument, but he's getting more posts, and I believe in keeping things fair and balanced.

  10. The 1998 team was better defensively and had a better bullpen and bench.  I don't have the RISP #s in front of me, but it sure felt like the team was better in the clutch, took extra bases more often, etc…in other words, figured out how to score runs without a bunch of 30-40 home run guys.  It was also an expansion year I believe, so that helped.  And while Cano over Knockelbock sounds good, check how many runs Chuck scored that year.  Sure, guys hitting behind him helped, but he got himself into position to score those runs too.  When you look at what a productive tablesetter he was until his arm fell off, the analysis is less obvious.

  11. .02 cents from a "late to the argument" Padre Fan:

    That 1998 Padre team was REALLY, REALLY, GOOD!

    When the Yankees were able to slice us up  like a knife through warm butter it was a pretty good indicator of that Yankee squad's greatness. Swept. It pains me to acknowledge this all…….as though my psychiatrist is IIATMS.

  12. For all the talk of the new yankee stadium being a “bandbox”, the yanks scored a whole 5 (five) runs more at home than on the road.  Compare that to the Red Sox splits.  Yes, there are days when fly balls carry nicely to right field and I suspect that some of these balls are getting an extra 5-8 or so feet on some days, but what the stadium gives it takes away in other respects.  The Yankee pitchers figured it out in the second half and you had a lot fewer looks of beffudlement from Pettite and Burnett as fewer fly balls were reaching the seats.

  13. Isn't there a huge difference I haven't seen mentioned?  The AL in 1998 had only 1 other team with more than 90 wins (Boston had  92).  Not that we can guess how strong the AL will be this year, but it seems likely the Yanks will have close competition this year (much like last year).