There's been some discussion of late about the Phillies dependence upon the long ball as their primary method of run production.  The recent run of 15 straight home runs coming as solo-shots (as in....nobody on base) coupled with a depressed offensive output for the team has made this discussion even more heated recently.  There are two primary schools of thought, with competing themes.

In order to win baseball games, you have to score runs.  As Harvey Haddix found out long ago...being perfect, even extending into extra innings, can still lead to a loss if the guys at the dish are as inept as your opponents at making contact and putting balls where the opponents are not.  Therefore, hitting home runs is good.  A good power hitter is in scoring position the minute he steps to the plate.  Common sense says that the more home runs you hit, the more potent your offense, the better your chance to win games.   There really isn't any arguing going on with this point. 

The discussion at hand however is going on the premise that if a team seems to rely on the long ball too much for their scoring, they might do better cease trying to hit the ball into next week, focus on making contact, getting on base, and being able to score by more reliable means.  My take on this has been that if you run into a homer drought, you're dead.  Also..in post-season you tend to face good pitchers on better teams.  Hitting dingers off the middle relievers from marginal teams is likely easier than doing it against a staff ace from a playoff team (but then again...so is getting a hit of any kind..)  Small ball isn't necessarily more exciting, we all remember McGwire and Sosa, and we all likely know that Rickey thefted 130 bases one year....but that's not how the Phillies are built nowadays.  So the question at hand is...do teams that have a higher or lower percentage of the runs come via the home run have better success in the regular season and in the playoffs?

One of my co-workers (Chris Dial...recognized for his defensive analysis and a regular contributor at www.baseballthinkfactory.org was kind enough run a query in his stat software and give me the data for all teams since 1993 with regards to their runs scored, and the percent of runs scored via the homer.  In my attempts to make some sense of the data, I discarded 1993-1995 as it was pre-divisional play, and also included a couple strike shortened season (makes correlating percentages to win totals a little difficult).

So what did I find?

Let's get to the data.

Here's each year, and the MLB average for the percent of runs scored via the homer:

1996       35.0%

1997       33.8%   

1998       34.3%

1999       35.9%

2000       36.7%

2001       36.6%

2002       35.7%

2003       35.7%

2004       36.8%

2005       35.9%

2006       36.0%

2007       34.1%

2008       34.3%

The numbers hold pretty steady, roughly 34-37% over a 13 year period.  My litmus test of precision is the term ‘%RSD' which effectively normalized standard deviations.  Those 13 values yield a %RSD of 2.97, which I'll call ‘tight' with a little room for improvement, but for the purposes of what I'm doing here...good enough to warrant discounting year to year variations and fluctuations in how teams are doing.

The first correlation I'll go after is runs scored to wins.  Here's the plot:

via lh6.ggpht.com

While this looks scatter shot, you will notice there's a slight tilt to the dots, as evidenced by the trendline.  The rationale here is essentially, home runs = runs; runs = wins.  One other thing to think about though is that the teams above that 100 win mark seem to be centered around 30-45%.

Now to determine how teams did in the playoffs, I gave them a ‘score' based on their performance. 

Zero = missed playoffs altogether.

One = made the playoffs, but lost in first round.

Two = made it to the LCS, but lost.

Three = made it to the World Series, but lost.

Four = won the World Series.

How does the homer percentage correlate to post-season performance?  Lets see....

via lh3.ggpht.com

Before you dig into this, remember that a ‘dot' on the 4 line, should also be dropped to the other 3 lines below it as well, as this is a population graph, so attaining a score of ‘4' means you also made it to the LCS, and the playoffs as well.

What do we notice?

A bit of a parabolic effect.  Teams that have made the playoffs pretty much cover the same range as the teams that don't,  but there's only a couple blips down there at the 25% range.  Looks to be that if you want to make it to the playoffs and have any shot of advancing, you best get about 28-45% of your runs via the homer. 

The data seems to support both sides of the argument.  Home runs are good for your offense, but overly relying on them might get you there, but you won't hang around for long.  As of Sep. 8th, the Phils were scoring 46% of their runs by the homer.  If they were to win or even make it to the World Series with that mark, it would be the highest value of any team to do so in the past 14 years.