The Saber's Edge: Can Desmond Survive Moving to OF?

Written by

Updated on March 3, 2016 6:51PM EST

Ian Desmond has finally signed, but I am pretty sure most didn't expect him to be playing outfield for the Rangers. Probably just as many don't expect this move to work after seeing Hanley Ramirez struggle when Boston moved him from left side of the infield to the outfield. Ramirez's struggles are evident as he now moves to first base for 2016. Just using a sample of one is far from ideal. Players have made the move without such struggles. Jose Bautista moved from third base to outfield and has not missed a step offensively. Let's look at the consistent production by players moving full time from the infield to the outfield (and vice versa).

Comparisons between Ramirez and Desmond littered the internet immediately after the signing. Each writer may have thought they were the only ones to make the link. It makes for an easy narrative after the signing. Well, the data exists to see if the narrative is correct. It is time to check the historic data instead of basing our conclusions on a single sample.

Note: With just a few coding changes, I could look at outfielders making the move to the infield. While Desmond isn't making this move, I figured it would be nice to have the data available anyway.

First, the one item I didn't want to examine was the utility player who constantly moves around the diamond over the course of the season (see Willie Bloomquist). Instead, I want to look at players

Note: With just a few coding changes, I could look at outfielders making the move to the infield. While Desmond isn't making this move, I figured it would be nice to have the data available anyway.

First, the one item I didn't want to examine was the utility player who constantly moves around the diamond over the course of the season (see Willie Bloomquist). Instead, I want to look at players moving from playing one position dominantly to another. I decided to initially use at least 81 games played at each position in the paired seasons. Additionally, I only looked at players only moving from second base, third base and shortstop. I wanted the player to have some defensive value and not be a first basemen forced to the outfield or vice versa.

The reason people don't remember many of the moves from full time infield to outfield is that they are not common. In the last 10 years, only nine players have moved from infield to outfield. Since 1950, only 43 players made the transition. For outfielders moving to the infield, the numbers are similar with 11 players making the transition over the last 10 season and 37 since 1950. While I will adjust the games played later, here is how the players produced comparing their first season to the second.

	GAME	AVG	OBP	SLG	ISO	wOBA	AGE Y1	AGE Y2	COUNT
OF to IF since 1950	1	0.002	0.001	-0.005	-0.002	0.001	28	29	37
OF to IF since 2005	-6	0.006	0.005	-0.006	-0.009	0.003	28	29	11
IF to OF since 1950	5	-0.008	-0.007	-0.044	-0.016	-0.016	27	28	43
IF to OF since 2005	5	-0.016	-0.031	-0.074	-0.027	-0.045	26	27	9

The infield-to-outfield transition really seems to negatively drag offensive production with the most recent seasons being worse. Historically, players have seen their wOBA drop 16 points with the decline coming from both on-base rate and power. An issue to keep in mind, we are only looking at 11 players, so the drop could be from random variations in output from small samples.

The change from outfield to infield doesn't seem to negatively affect a player's overall offensive production. If anything, the move seems to improve the player as seen by their wOBA slightly increasing.

One note from the information is that teams seem to want to move younger players (median age 28 or 29) from infield to outfield. When they move from outfield to infield, players are usually one or two years older. Also, verall games played was generally constant.

Another issue I am worried about with this data, besides sample size, is that I am getting some survivor bias. With survivor bias, the players who performed well stay with the transition. If the player is struggling, the position change could be blamed. If this is true, the above values may be positively inflated. To see if this bias may be the case, I dropped the season two games-played limit to 61 games. Here are the results.

	GAME	AVG	OBP	SLG	ISO	wOBA	AGE Y1	AGE Y2	COUNT
OF to IF since 1950	-16	0.002	0.001	-0.001	-0.001	0.001	28	29	53
OF to IF since 2005	-7	0.006	0.005	-0.006	-0.009	0.003	27	28	13
IF to OF since 1950	-15	-0.007	-0.004	-0.016	-0.005	-0.008	27	28	69
IF to OF since 2005	-15	-0.005	0.000	0.002	-0.002	-0.001	27	28	15

It doesn't look like survivor bias is an issue, but sample size. As the sample increases, the variance in values for players moving to outfield was halved for the overall group and dropped to zero for the recent group.

Offensively, I see no reason to downgrade Desmond much for the move to the outfield. His production might take a small hit, but not much. The bigger worry is if his 2015 first half (and whole season) production is his new talent level or will he get back his 2012-2014 production when he had more than 20 stolen bases and 20 home runs each season.

Another possible issue brought up is the injury chances of a player making the transition to an unfamiliar position. Hanley Ramirez was off to great start last season until he ran into a wall and injured himself. Well, since Hanley did it, Desmond is surely to do it also. MLB.com's Cory Schwartz and Sports Illustrated's Joe Sheehan debated this very point on twitter a few days back.

@joe_sheehan Ah, of course, you'd never want to price in the injury risk for a guy with his history, who's changing positions. ¯\_(ツ)_/¯
— Cory Schwartz (@schwartzstops) February 29, 2016

Well, this isn't too hard to check using my injury database. I will use a little different time frame (2002 to current) because that is how far back my database goes. I looked at the chances for every player who played 81 games in two seasons to those who made the transition. For each of the comparisons, I have added the P-value, which is the chance that the difference is due to random chance.

	ALL	IF to OF	OF to IF	COMBINED MOVERS
Total in Sample	3,044	12	15	27
On DL in Y2	946	4	4	8
Percentage	31.1	33.3	26.7	29.6
P-value		0.90	0.79	0.91

The results show no significant change from the overall DL chances because the change is so small considering the small player sample.

With the injury information, I am even more likely to value Desmond higher than the crowd. I can see some saying they will just stay away with all the questions surrounding him. The biggest question for me will be his playing time. With several players in the mix, will he be replaced if he struggles?

Conclusion

Hanley Ramirez's disaster in left field last season will stick in people's mind as Ian Desmond makes the same transition. Don't make a big deal about the move as players have maintained performance making such moves. Also don't worry about him getting hurt more in the unfamiliar confines of the outfield. I would just go with a combination of projections and value him accordingly. I expect a final performance somewhere between his 2014 and 2015 seasons.