Tuesday, July 03, 2007

Projecting the Second Half

As Independence Day nears, the Kansas City Royals have passed the midpoint of the season looking remarkably like a good baseball club the past month.

There will probably be many moves in the next month -- injured players returning, veterans getting traded, Buddy Bell giving Tony Pena Jr. a ten game shot at batting cleanup -- but for the moment, I want to focus on four numbers:

68, 72, 77 & 79

Sixty-Eight Wins
If you simply take the current winning percentage of the team and put it against 162 games, the final record will be 68-94.

Avoiding 100 losses is an accomplishment unto itself given the recent history of the franchise, but doesn't this team feel like it should win more than 68 games? After winning 62 games last year, a six game swing would be quite disappointing.

Seventy-Two Wins
I use the following formula to calculate "Pythagorean" wins:

(RS*2)/(RS*2)+(RA*2)

Statheads generally look at Runs Scored (RS) and Runs Allowed (RA) as a "truer" picture of how well a team has played.

People much smarter than me have suggested that a factor of 1.78 is more accurate, or even a factor that changes depending on the scoring environment of a particular season. Me? I'm happy to just square everything up and go with that.

What that gives us is a team that is actually good for about 72 wins on the season. Still not remarkable, but markedly better than the 62 of Ought-Six.

Seventy-Seven & Seventy-Nine Wins
The team has been playing much better lately, and I wanted to give them credit for that. Luckily, I have way too much free time today because the whole world is already on vacation for the 4th of July, so I decided to look at the pythag record for the last ten games and the last twenty five.

This chart shows the season so far in terms of real wins and the various calculations.

As discussed in an earlier post, the "10 Day Py" line shows those stretches when the team really plays well (and when they don't). In contrast, the "25 Day Py" line is perhaps a better predictor of future performance as it takes into account a larger sampling of games.

Now, 10 and 25 are purely arbitrary numbers. I still think this provides a good snapshot as to how the season has really gone for the team. Two really good stretches, culminating in one of the better months the team has had since 2003.

If the team finishes out the year winning 53-55% of their games (as they have for the last 10 and 25, respectively), they will finish the season with 77-79 wins.

Those totals won't bring home any hardware, but they'll come from a young team that will remain intact and improving for the next several years.

That's something to cheer about.


No comments: