Notice: Trying to get property 'display_name' of non-object in /var/www/html/wp-content/plugins/wordpress-seo/src/generators/schema/article.php on line 52
keyboard_arrow_uptop

In Baseball Prospectus 2003, we introduced a new rate metric in lieu of Equivalent Average (EqA), which graced the pages of previous editions. This metric, Marginal Lineup Value Rate (MLVr), measures how much offense a player produces compared to an average player. Since the publication of BP 2003, one of the most common questions I’ve received concerns what the scale of MLVr is, or in other words, what a “good” MLVr is.

As a new and unfamiliar metric, MLVr lacks the built-in recognition factor that something like EqA had, which was designed to follow the familiar batting average scale. The tradeoff, however, is that the “units” of EqA don’t measure anything–one point of EqA doesn’t equate to one run, or a tenth of a run, or a fraction of a win, or anything else that’s tangible. Equivalent Average is essentially a dimensionless index that follows offense production, but does not, by itself, measure it. Instead it’s made so that the “installed base” of baseball fans can understand it.

MLVr takes the opposite tack, choosing to express results in terms of runs per game, (and more specifically, runs per game above or below a league average player), rather than a more familiar scale. This makes it more useful for quantitative analysis, at the expense of being more opaque to casual baseball fans.

An example of how MLVr can be used might be this: Suppose the Angels trade Garret Anderson and David Eckstein to the Mariners for Mike Cameron and Carlos Guillen. How much does this hurt or help the offense?

First, we look up each player’s MLVr (figures through May 6th):


Angels            Mariners
Anderson  0.350   Cameron 0.251
Eckstein -0.124   Guillen 0.110

The Angels are giving up players who together produce.226 runs per game more than two average players (.350 MLVr -.124 MLVr). The Angels receive two players who produce .361 runs per game more than two average players do (.251 MLVr + .110 MLVr). The Angels’ offense would expect to gain 0.135 runs per game (.361 – .226). If they averaged 5.2 runs per game before, they’d improve to 5.335 runs per game now, or about 22 runs over the course of a 162 game season (0.135 MLVr * 162 Games = 21.87 runs). It’s the fact that MLVr is expressed in runs per game that lets us work directly with the numbers themselves.

The advantages to the analyst aside, it is still hard to make the transition to knowing just by looking what a good MLVr is. The most basic rule of thumb is that a positive value means above average, and a negative value means below average, so checking the sign of MLVr gives you a very rough yardstick to measure by. But many of you have asked for a more handy cross-reference, so you can equate a 0.250 MLVr to a corresponding EqA or OPS value.

We can compare OPS and MLVr from a recent season to help us get a feel for the relationship between the two. I looked at all batters with 200+ PA in 2002, and looked at the distribution of OPS and MLVr, respectively. The table below summarizes the results as percentiles. For example, the 25th percentile value is the point at which 25% of players have a OPS or MLVr below that value. The median is equivalent to the 50th percentile and is the value where half of all players in the group are below, and half are above.


                          OPS      MLVr
 
Minimum                   0.497   -0.422
25th %-ile                0.688   -0.084
Median                    0.751    0.013
75th %-ile                0.813    0.121
Maximum                   1.381    0.994

Half of all regular players will fall within roughly -0.100 to +0.100 MLVr, or about a 650 OPS to 850 OPS.

We can also run a linear regression to help find a simple formula to convert MLVr to OPS and vice versa. Though these formulas aren’t using the exact values from the regression, they are close enough to make the tradeoff for simplicity worthwhile.


OPS ~= .750 + .6* MLVr

MLVr ~= (OPS-.750)*1.67

Armed with these new tools to help us relate this new tool to more familiar measures, perhaps we can start to develop some intuition about MLVr and incorporate it into our standard sabermetric bag of tricks.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe