The statistical impact of switching leagues (for pitchers)

Javier Vazquez disappointed a lot of fantasy owners in the American League; how will the move to the National League affect him? (Icon/SMI)

Last week, I looked at how a hitter’s stat line would be affected simply from moving from one league to the other. Overall, we confirmed the general wisdom that moving to the National League is favorable for hitter. Today, let’s take a look at pitchers and see if the same holds true.

These results could end up having a big influence on pitchers like C.C. Sabathia, Jake Peavy, A.J. Burnett, Derek Lowe, Francisco Rodriguez, and numerous others in 2009. Javy Vazquez will certainly be affected given his recent trade to the Atlanta Braves (on this note, be sure to read Josh Kalk’s great Vazquez article from yesterday if you haven’t already).


My process is slightly changed from when I looked at hitters as I thought it would be appropriate to use multiple years to increase the sample size and to try and drown out single-year noise. I’ll be re-posting the hitter results using this altered process sometime soon.

I used seasonal data pairs from 2004 through 2008 (i.e. 2004-2005, 2005-2006, etc). I examined all players who played in one league in year one and the other league in year two. The results I’ll present are the aggregate of all data pairs included in the sample, which totals 415. Each player’s contribution to these results were weighted based on the lower of his at-bat, plate appearance, or other such denominator total for each data pair.

The year one numbers were age-adjusted to put them on par with the year two numbers, and the numbers for both years were park-neutralized. Finally, year one numbers were also adjusted for differences in league average.

In the tables presented, the first column gives the total weighted denominator, as explained above. The second column gives the aggregate change simply as a result of switching leagues. These numbers are to be read as if an AL pitcher moves to the NL, and you would simply take the inverse for an NL player moving to the AL.

Pitching results


+-------+----------+| wIP   | AL -> NL |+-------+----------+| 18685 |   + 0.57 |+-------+----------+


+-------+----------+| wIP  | AL -> NL |+-------+----------+| 18685 |   - 0.05 |+-------+----------+


+-------+----------+| wIP   | AL -> NL |+-------+----------+| 18685 |   - 0.02 |+-------+----------+


+--------+----------+| wOF FB | AL -> NL |+--------+----------+|  19340 |  + 0.21% |+--------+----------+


+-------+----------+| wBIP  | AL -> NL |+-------+----------+| 56796 |  - 0.008 |+-------+----------+


+--------+----------+| wDenom | AL -> NL |+--------+----------+|  23655 |  + 0.47% |+--------+----------+


+-------+----------+| wTPA  | AL -> NL |+-------+----------+| 55240 |  + 0.33% |+-------+----------+


+-------+----------+| wBIP  | AL -> NL |+-------+----------+| 59273 |  + 0.11% |+-------+----------+


+-------+----------+| wBIP  | AL -> NL |+-------+----------+| 59273 |  + 0.37% |+-------+----------+


+-------+----------+| wBIP  | AL -> NL |+-------+----------+| 59273 |   - 0.42 |+-------+----------+

Approximated based on RA results

+-------+----------+| wTPA  | AL -> NL |+-------+----------+| 18685 |   - 0.41 |+-------+----------+


Just like when we looked at hitters, we find that pitchers do better moving from the American League to the National League (as expected). In some cases, the changes are extremely significant. In moving to the NL, a pitcher’s strikeout rate (K/9) would be expected to rise more than a half-point. That’s huge. Walk rate is less significant at just an 0.05 difference.

[Non-fantasy comment] As a general comment, compare this to the relatively large gain the hitters saw in switching leagues (roughly 5-to-1 per 630 plate appearances/batters faced). We know that the NL issues/draws more walks than the AL each year (8,816 to 7,521 this past year), but these studies seem to indicate that this discrepancy is mostly due to the talent level of the pitchers, not the hitters. With guys like Ryan Howard and Adam Dunn and Barry Bonds playing in the NL in recent years, one might be under the impression that the batters are the cause of this. Not really so.

To put it simply, NL pitchers have worse control than their AL counterparts; NL batters don’t necessarily have better patience (or at least not much better). This isn’t of super-importance for fantasy owners, but I thought it was interesting enough to note. [/Non-fantasy comment]

A pitcher moving to the NL would drop .008 points in BABIP—very favorable. This is particularly strange because the pitcher would also give up a few more line drives, though not enough to make too large of a dent in the BABIP. A pitcher would also induce a few more ground balls and infield flies and strand more of the runners that he allows on base.

Aside from the aforementioned line drive rate, the only other stat that a pitcher would be better off moving to the AL for is HR/FB. We see a small increase here, but it would really only equate to three-tenths (0.30) of a home run over 150 innings or so.

Overall, a pitcher moving from the AL to the NL would see his ERA drop by about 0.41 points, a very large difference and one well worth noting on Draft Day. In a traditional mixed league, this would make the difference between drafting a player in the mid-teens (4.10 ERA) and not drafting him at all (4.51 ERA). A lot of this comes from the strikeout increase (which is the best outcome for fantasy owners since strikeouts are a category in-and-of themselves), though the little gains in walks, hit-by-pitches, ground ball rate, infield fly rate, BABIP, and LOB% all add up to contribute.

Concluding thoughts

Hopefully you guys found these articles interesting and will find some use for the results. I’ll certainly be referencing them in player profiles. If you have questions about any of this (or anything fantasy related at all) or suggestions for future studies, always feel free to drop me an e-mail. Also, as I mentioned earlier, I’ll be sure to post the results for hitters using multiple years’ worth of data shortly.

Comments are closed.