Astros vs. the Contenders using WAR

[subnuclear]

I drank a bunch of coffee this morning and decided to sit here and figure what it would take to get the Astros into contention. I really like this list of WAR by position provided by Baseball Reference (scroll down the page a bit). WAR is not a perfect stat, but the teams with the top WAR are also the teams in contention and it tries to take into account defense.

In any case, I rated the positions for each team into four groups: Exceptional (Top 3), Probably OK (4-7), Bleh (8-11), and Exceptionally Bad (Bottom 3). I listed the position WAR in parentheses so you can see the relative value.

Los Angeles Angels of Anaheim

Exceptional - C (0.6),  SS (1.7), CF (5.3), RF (2.3)
Probably OK - 1B (0.3), 2B (3.0),
Bleh - SP (2.2), RP (0.2), 3B (-1.7), LF (0.0), DH (-0.6)
Exceptionally Bad -

Oakland Athletics

Exceptional - RP (2.4), C (1.7), 3B (4.4), RF (2.1)
Probably OK - SP (2.4), 1B (-0.3), LF (1.0), DH (-0.3)
Bleh - 2B (-1.4), SS (-1.0)
Exceptionally Bad - CF (0.0)

Seattle Mariners

Exceptional - SP (4.7), RP (3.5), 2B (3.6), 3B (3.4), SS (0.9)
Probably OK -
Bleh - LF (-1.0), RF (-1.0)
Exceptionally Bad - C (-1.4), 1B (-1.9), CF (-1.7), DH (-2.4)

Detroit Tigers

Exceptional - SP (7.1), 1B (2.5), DH (2.7)
Probably OK - 2B (2.6), LF (1.1)
Bleh - C (-0.6), CF (0.1), RF (-0.9)
Exceptionally Bad - RP (-1.7), 3B (-3.6), SS (-1.9)

Kansas City Royals

Exceptional - LF (4.0), CF (2.7),
Probably OK - SP (3.4), RP (1.6), C (0.5), SS (0.1), RF (0.0)
Bleh - 1B (-1.7), 2B (-1.3), 3B (-1.9)
Exceptionally Bad - DH (-2.6)

Baltimore Orioles

Exceptional - 1B (1.7), SS (1.4), LF (3.1), CF (2.5), DH (1.7)
Probably OK - RP (0.2), C (-0.1), 3B (1.2), RF (0.2)
Bleh - SP (-0.5), 2B (-1.0),
Exceptionally Bad -

Houston Astros

Exceptional - 2B (3.6)
Probably OK - SP (3.7), DH (-0.7)
Bleh - C (-0.2), SS (0.1), LF (-1.3), CF (0.4), RF (0.7)
Exceptionally Bad - RP (-5.7), 1B (-3.4), 3B (-3.5)

[subnuclear]

The top teams (LAA, DET, BAL) all have have team WARs of ~11. The borderline teams (SEA, KCR) have team WARs around ~5.0. Astros have a team WAR of -6.8.  The Astros will have to improve their team WAR by around 11 to get in contention.

Since Springer and Grossman got some of their rookie jitters out of their system, I think they will gain 2 WAR in the OF with the players they have. Relief pitching has the worst WAR of any position in the league, so it can gain 5 WAR just by somehow becoming average (I'll just say 3 WAR improvement). I think we all see talent in Singleton, but 1B needs some help and could improve by 2 WAR by becoming average. Good 3B can be hard to find, so I'm not sure we'll see improvement there any time soon. Also, I think WAR doesn't capture what Dominguez does extremely well (soft hands, great throws to 1st).

So, that's 7-10 WAR improvement by my hand waving. Thinking they could nip the heels of KCRs and SEAs of the world is not crazy, but they are still a few really good players away from division titles.

[NeilT]

The top teams (LAA, DET, BAL) all have have team WARs of ~11. The borderline teams (SEA, OAK) have team WARs around ~5.0. Astros have a team WAR of -6.8.  The Astros will have to improve their team WAR by around 11 to get in contention.

Since Springer and Grossman got some of their rookie jitters out of their system, I think they will gain 2 WAR in the OF with the players they have. Relief pitching has the worst WAR of any position in the league, so it can gain 5 WAR just by somehow becoming average (I'll just say 3 WAR improvement). I think we all see talent in Singleton, but 1B needs some help and could improve by 2 WAR by becoming average. Good 3B can be hard to find, so I'm not sure we'll see improvement there any time soon. Also, I think WAR doesn't capture what Dominguez does extremely well (soft hands, great throws to 1st).

So, that's 7-10 WAR improvement by my hand waving. Thinking they could nip the heels of KCRs and SEAs of the world is not crazy, but they are still a few really good players away from division titles.

-6.8?  Give peace a chance!

[Uncle Charlie]

Astros have a team WAR of -6.8.

As a bit of context and full disclosure, I used to be really into stats but have not delved into WAR yet (though I did get into Win Shares when that was becoming popular in some circles).  However, this is an illustration of why I do not understand WAR at its face.  WAR is explained at a high level as the incremental wins one could expect by replacing a player with a guy who was not on the 25-man roster (I.e. A good AAA player).  Therefore, I cannot understand how this team would have collectively lost 7 games MORE than a team of 25 Good AAA players - do I misunderstand the concept?

Another issue that I have with WAR is that a majority (I don't not know the exact % nor would hazard a guess) of the people who talk about or even reference wat could explain it simply enough for people to "get it".  Of the people who can do that I would guess that most of those people couldn't calculate it either.  Therefore, I think too many folks toss it around and utilize it out of context.  There are certainly problems with WAR in evaluating players.  I don't know what they are of course, but anything that can take something this complex and break it down to a single number has problems and must be used in the right context.

As an expample, I'm a commercial / economics guy.  Some of the key statistics we use are rate of return (IRR), net present value (npv) and profitability index (PI).  We also use some proprietary ones, but I won't get into those.  The point being is that each of those statistics have value but only in the right context.  Some of our proprietary stuff is like WAR in that it combines things, but until you understand how it works it can be dangerous because it has it's contextual weaknesses as well.

Sorry to get on my soap box....I'll get down now...I feel (a little) better.

[subnuclear]

Therefore, I cannot understand how this team would have collectively lost 7 games MORE than a team of 25 Good AAA players - do I misunderstand the concept?

The WAR model predicts 25 good AAA players would have a .294 winning percentage over a season. The Astros thankfully did better that, but WAR is not linked directly to wins, but is an accumulation of individual plays normalized to wins. If those bad plays had been equally distributed through the season they should have had a much lower winning percentage. Instead, those plays were likely concentrated in a non-small set of blowout games some of which I had the pleasure of watching.

This exercise was interesting to me because I wanted to understand the ratio of really exceptionally players on good teams to replacement level players. How many terrible players can a playoff team have? (The Astros outfield in 2005 was not good for instance). How many really good players will the Astros need to compete? I think WAR is ok for sorting that out because I can't keep track of a large number of players except through stats.