Tuesday, May 17, 2005

Introducing ... ERV Player Charts!

The crack R&D team here at The Nats Blog, Inc. (picture a cross between Oompa-Loompas and Beaker from "The Muppet Show") has done it again: ERV Player Charts! Each night, we will be unveiling a chart for each player in the starting lineup in the Nats. Tonight we lead off with the lead-off hitter, Brad Wilkerson:

The chart shows the cumulative Run Value and Win Value for Wilkerson's batting only, through tonight's game. As a reminder, RV reflects the runs generated/squandered by the player, independent of the score and inning of the game. WV adjusts the RV to reflect the score and inning. I'm not sure yet exactly what these charts show, but I think separation between the RV line and WV line indicates whether the player is producing or choking when the game is on the line, generally close and late situations.

[Edit: I've revised this chart based on the change to Win Value I made, described in this post.]

As you'll see over the next few days, the charts look very different for each player, and show a few surprises. Brad's seems to be a good depiction of his up and down season so far, though it must be stressed that he has stayed above average all year, which is an accomplishment. They remind me of stock charts, and in a sense they are the same -- they reflected the value that the player has added or subtracted from the Nats' fortunes during the season.


At 11:27 PM, Blogger tmk67 said...

"I'm not sure yet exactly what these charts show, but I think if the RV line is very far from the WV line, it means the player has not really been producing when the game is on the line, generally close and late situations."

I think the gap between the RV and WV line shows a degree of consistency in the player. Assume a player creates RV of 1 for every game. In some games, that run will matter, in others it will not. So his "RV" line will increase steadly while his "WV" line on the chart will waggle, and there would be increasing gap between the two. Indeed, by definition, the Cumulative WV Line would never catch up with the Cumulative RV Line for this player.

But I desparately want a guy who creates 1 RV a game, every game, on my team!

I would be less concerned about divergence in the Cumulative lines but would pay particular attention to a player whose lines are converge at high per-game WV levels -- in other words, the search for a player who performs in important situations. I'm not sure Cumulative lines would capture that -- maybe what you should plot is the per-game difference between RV and WV (x-axis) against different WV levels (y-axis). In other words, plot RV/WV difference against WV opportunities.

At 10:39 AM, Blogger SuperNoVa said...

This is great stuff, DM. Among other things, a better-than-average, consistent player should have a line that slopes upward at a consistent rate. As you can see, Brad has been kind of streaking during the season.

At 10:48 AM, Blogger DM said...

Backward K,

Good comments. I think you are right, but wait until you see some other player charts, which, as I said, are different and interesting.

SNV, same for you. I think you have described the chart for one "blue-chip" stock on the Nats.

At 1:21 PM, Blogger Chris Needham said...

Buck says that cryptically is referring to NJ!

At 2:07 PM, Blogger DM said...

No bet, Chris. But you might be pleasantly surprised by the ERV chart for Brian Schneider, which may provide support for your anti-Schneider :) position.


Post a Comment

<< Home