Monetary values for various skills can actually be found on fangraphs — another baseball site.

The problem is that On-base percentage and power are combined, though you could do the math to divide them using linear weights or the like (with quite a bit of effort). Also, it would cause some serious problems since OBP and SLG have a lot of overlap (generally speaking) since each home run also gives you points toward on-base.

Also, because of the difference in scales (defense is worth roughly half of offense, and base-running is even less important), the graph won’t line up even for the ‘average’ player.

1. I thought about labeling the last points, but there would have been too many overlapping series. I guess it would work for the two highlighted series in the last chart.

2. I also thought about aligning the tick marks with the labels. I didn’t want the first category lined up with the Y axis, so I didn’t do that (though hiding the Y axis line and tick marks and using horizontal gridlines would remove the awkwardness). I would have hidden the X axis tick marks, then added a dummy series, using crosses as markers, but decided enough was enough.

3. I’ve been monetizing this stuff via the charting utilities, and I have more of these to build. I’ve also started “my book” on several occasions, but it’s nowhere yet. One of these days….

There’s a really cool macro on the web called ‘Label Last Point’ that would work well with your reworked chart – it would do away with that nasty legend. Now, where did I see it again? Ahh, that’s right…it’s at http://peltiertech.com/WordPress/label-each-series-in-a-chart/ . Only problem is that sometimes the labels overlap.

Also, it’s very easy to add another series that shows the difference between the 2 highlighted series. Perhaps this makes it slightly easier to judge the relative skills at a glance…although you then have to rescale the axis to fit in both positive and negative difference ranges thanks to Adam Dunn’s fielding.

Pity that Excel does’nt allow you to easily line up the tick marks on the y axis with the data points. If you change the ‘Position Axis’ setting to ‘On tick marks’ then the points line up, but the first y axis label intersects the x axis, which looks really awkward.

Only way I can think to get around this was by first changing the “Interval between tick marks” setting to something large (anything above 4 in this case) so that you get no tick marks at all, and then adding 4 new dummy series (one for each point on the Y axis) and using data labels on each of these series to put a tick mark in the place where I want it.

I’ve added these to your example and posted my spreadsheet at http://cid-f380a394764ef31f.skydrive.live.com/self.aspx/.Public/Composite%20Player%20Evaluation%20with%20additional%20series.xls#resId/F380A394764EF31F!134

There’s a picture of my example at http://screencast.com/t/YQduptUS

Convention says to use bars for data when the points are important, lines when the trends are important. However, a bar chart withmore than two players would become difficult to read. As I mentioned in my response to Mike’s comment, the lines are important to indicate the data for a specific player over all categories.

Parallel coordinates analysis is widely used for much more complex data sets than I show here. If you want a more detailed explanation, please read Stephen Few’s article, Multivariate Analysis Using Parallel Coordinates.