The Value of a Walk in OBP Leagues
Last week I took a look at an alternative scoring setting that is becoming more common in fantasy baseball, the quality start. This week, I will examine what I consider to be the offensive sibling as far as changes from traditional formats, on-base percentage as a replacement for batting average.
The Gainers
| Name | AVG | OBP | OBP-AVG |
| Joc Pederson | -2.0 | 0.6 | 2.6 |
| Carlos Santana | -1.2 | 0.9 | 2.1 |
| Joey Votto | 1.8 | 3.9 | 2.1 |
| Jose Bautista | -0.5 | 1.5 | 2.0 |
| Yasmani Grandal | -1.1 | 0.8 | 1.9 |
| Chris Carter | -2.4 | -0.6 | 1.8 |
| Bryce Harper | 2.4 | 3.9 | 1.5 |
| Lucas Duda | -0.8 | 0.7 | 1.5 |
| Mike Napoli | -1.5 | -0.1 | 1.4 |
| Alex Rodriguez | -0.5 | 0.8 | 1.4 |
| Curtis Granderson | -0.2 | 1.1 | 1.3 |
| Anthony Rizzo | 0.5 | 1.8 | 1.3 |
| Mark Teixeira | -0.4 | 0.9 | 1.2 |
| Alex Gordon | 0.2 | 1.5 | 1.2 |
| Steven Souza Jr. | -1.5 | -0.3 | 1.2 |
| Andrew McCutchen | 1.0 | 2.2 | 1.2 |
| Miguel Montero | -0.6 | 0.5 | 1.1 |
| Adam LaRoche | -2.1 | -1.0 | 1.1 |
| Paul Goldschmidt | 2.1 | 3.2 | 1.1 |
| Dexter Fowler | -0.5 | 0.6 | 1.1 |
| Chris Davis | -0.1 | 1.0 | 1.1 |
| Luis Valbuena | -1.5 | -0.5 | 1.0 |
| Brian McCann | -1.2 | -0.2 | 1.0 |
| Shin-Soo Choo | 0.4 | 1.4 | 1.0 |
| Russell Martin | -0.9 | 0.1 | 1.0 |
| Chris Coghlan | -0.5 | 0.4 | 1.0 |
The Drainers
| Name | AVG | OBP | OBP-AVG |
| Chris Colabello | 2.1 | 1.2 | -0.9 |
| Juan Lagares | -0.2 | -1.1 | -0.9 |
| Matt Duffy | 1.1 | 0.2 | -0.9 |
| Eugenio Suarez | 0.6 | -0.4 | -0.9 |
| Starlin Castro | 0.0 | -0.9 | -0.9 |
| David Murphy | 0.7 | -0.3 | -0.9 |
| Nolan Arenado | 0.8 | -0.1 | -0.9 |
| Yoenis Cespedes | 1.0 | 0.0 | -0.9 |
| Gerardo Parra | 1.0 | 0.0 | -0.9 |
| Scooter Gennett | 0.0 | -1.0 | -0.9 |
| A.J. Pierzynski | 1.3 | 0.3 | -0.9 |
| Yasmany Tomas | 0.3 | -0.6 | -0.9 |
| Adeiny Hechavarria | 0.6 | -0.4 | -0.9 |
| Erick Aybar | 0.2 | -0.8 | -0.9 |
| Nick Hundley | 1.3 | 0.3 | -1.0 |
| Francisco Lindor | 1.8 | 0.8 | -1.0 |
| Jose Altuve | 1.8 | 0.8 | -1.0 |
| Brandon Phillips | 1.1 | 0.0 | -1.0 |
| Jean Segura | -0.3 | -1.4 | -1.1 |
| Ben Revere | 1.5 | 0.4 | -1.1 |
| Ender Inciarte | 1.4 | 0.3 | -1.1 |
| Omar Infante | -1.7 | -2.7 | -1.1 |
| Eddie Rosario | 0.1 | -1.1 | -1.2 |
| Salvador Perez | -0.2 | -1.4 | -1.2 |
| Xander Bogaerts | 2.0 | 0.8 | -1.2 |
| Dee Gordon | 2.5 | 0.9 | -1.6 |
I have computed the values of all players from 2015 who accumulated 350 or more at bats with regard to batting average and on base percentage as a z-score – standard deviations from the mean. Here are the 25 biggest gainers and losers from the change in scoring.
It’s not a shock to see that players with higher walk rates experienced a relative rise in value while players with lower walk rates dropped in value. The difference between batting average and on-base percentage is literally walks – with sacrifices and HBP making up a small difference. What I found interesting, is the scatterplot that I compiled below that attempt to determine the actual value of walks to fantasy value in on-base leagues.
This can be interpreted as follows. For each percentage point of walk rate above the league average – about eight percent – there is a corresponding increase of about .2 standard deviations. It’s easier to think about if you scale the data up to increments of five percentage points. For each five percentage points of increase in walk rate, a batter is worth an extra standard deviation above the mean than he would have been in a batting average league.
Considering that a decent player in fantasy is worth between five to ten standard deviations, this can help inform the value of a batter in a specialized league when you’re used to batting average leagues.
Consider the following example, Dee Gordon finished second in baseball with a .333 batting average while Joey Votto led baseball with a 20% walk rate and finished second in OBP with .459. In batting average, they scored 2.5 and 1.8 respectively. Based on my generalization above, and on Gordon’s walk rate being 5.2 percentage points below average and Votto’s being 12 percentage points above average, I would predict Votto’s OBP score to increase by 2.4 standard deviations and Gordon’s to decline by a little over 1 standard deviation. In fact, Votto’s increased by 2.1 and Gordon’s decreased by 1.5. It wasn’t perfect but it seems pretty close as long as you’re just looking for an approximation.
I realize that if you are using projections, you probably have OBP projections to go with the standard categories. It’s actually a little easier to project OBP and AVG due to the stability of walk rate versus BABIP. However, this exercise was intended to build a little intuition for valuing players with varying walk rates and determining the value of a walk. In short, assuming the average walk rate is eight percent, each five percentage point increments adds – or subtracts – about one standard deviation to a player’s value.
The Author
