Attempting to regress Ryan Goins’ defense to reality

Feb 22, 2014; Dunedin, FL, USA; Toronto Blue Jays second baseman Ryan Goins (17) throws the ball as the Blue Jays work out at the Bobby Mattick Training Center. Mandatory Credit: David Manning-USA TODAY Sports

On Friday I was reading an interesting article from Jeff Sullivan of FanGraphs, which, because of the way my brain works, suddenly forced me to start thinking about the Toronto Blue Jays.

Sullivan’s piece, titled How much to make of Juan Lagares defense, takes a look at what the New York Mets should reasonably expect from their young outfielder going forward. The write-up goes on to explain why he’s not a lock for the starting lineup despite putting up 2.9 wins above replacement in 2013.

The problem with advanced fielding metrics, as pointed out by an executive quoted in Sullivan’s article, is that they generally less reliable year-to-year in comparison to their more established offensive counterparts.

Sullivan concludes that Lagares defensive numbers should be heavily regressed, which should give a more realistic projection of what to expect from soon-to-be 25-year-old.

Lagares’ situation with the Mets struck a chord with me. It seems very similar to the situation the Toronto Blue Jays currently find themselves in with rookie second baseman Ryan Goins.

The parallels are there – both were second half additions, both significantly upgraded their teams’ defenses and both provided negative value at the plate. You could argue which position has more value (CF or 2B) but both play premium up-the-middle spots defensively.

The way UZR evaluates infielders is slightly different from the way it does for outfielders. Both use range as a factor, which basically tells us whether or not they get to more balls than the average fielder. Both also look at the number of errors made compared to a league-average fielder. Where the metric mostly differs is that UZR measures outfield arm runs for outfielders and double play runs for infielders.

In theory, range shouldn’t change that much on a year-to-year basis. As long as a player is healthy (and not losing foot speed) he should be able to get to roughly the same number of balls as the year before.

Sullivan is a machine of analysis and in his article provided much better analysis than I’m about to. Jeff calculated the standard deviation between Range/150 numbers and Arm/150 numbers with the goal of better determining the general range of year-to-year numbers.

I nearly failed my university stats class, which parlayed with my general laziness means I won’t be looking at things nearly as scientifically. However to get a better feel for how transferable range is on a year-to-year basis I decided to take a look at Cincinnati Reds’ second baseman Brandon Phillips.

Even at 33, Phillips has arguably the best range at second base in the game. I initially planned to look at Dustin Pedroia from the Boston Red Sox, who is another second baseman with elite range but injuries in 2010 and 2012 threw a wrench in my comparison.

So back to Phillips. The past three seasons, he has produced a range runs above average of 9.1, 7.0 and 8.1 in 2011, 2012 and 2013 respectively. Each year he played at least 1,250 innings at second base.

Last season, Ryan Goins had a 4.0 range runs above average in 262.1 innings. Extrapolated over 1,250 innings that works out to a range factor of 20 runs above average for the entire season.

Goins looked great at second base but is it reasonable to believe that his range would save twice as many runs as Phillips in an entire season? It’s highly unlikely, if not impossible. The leader among qualified second baseman for range runs above average was Ben Zobrist of the Tampa Bay Rays at 9.4. So even if you do believe Goins is an elite defensive second baseman we should probably cut his expected range factor in half, and that’s probably even being a bit generous.

For some reason, fielders with elite or even decent range at second base don’t necessarily turn any more double plays than your average player. In fact, according to FanGraphs since 2002 only nine qualified second baseman have a produced a positive double play runs above average.

DPR could very well be a flawed metric. It takes two to tango and whether a play turns into a double play or a force out at second may very well depend on a player’s infield teammates rather than the individual player’s skill. Regardless it’s still a factor in determining UZR so we can’t just ignore it all together.

Goins did look smooth as silk turning the double play last year and was credited with 1.7 DPR. As a converted shortstop, his arm strength and foot work should allow him to make an above average number of double plays. However the problem becomes whether or not he will receive enough opportunities to duplicate his gaudy DPR from 2013.

For comparison sake, Jose Altuve led all qualified second baseman with a 1.8 DPR last season and had the most double plays from the position with 106. Goins had 25 in just over 20% of the playing time that Altuve received. If you crudely extract those numbers Goins would have made about 120 double plays over a full season of work.

As is the danger with small sample sizes, we can’t just simply extrapolate a small amount of data over a full season and assume that it will hold up. Goins could very well be one of the game’s best at turning two but based on the historical precedent set by those before him the chance he adds significant value in this area is slim.

Error runs above average is an area that I haven’t discussed yet. Goins only made one fielding error at second base in 148 chances last season. There’s not enough data to make a reasonable assumption about his sure-handedness but based on the eye test alone I’d be surprised to see a negative number from Goins in 2014. However he might not be able to duplicate his .993 fielding percentage from 2013, which if he qualified would have tied for the league lead at second base.

So where does that leave us with Goins? If we assume he does have elite range at second we could probably assume a range runs above average of somewhere between five and eight. Double play runs are a bit more of a wild card and to be safe we probably shouldn’t assume anything more than a nil value. He also might add another run or two in value if he can maintain a fielding percentage of around 99%.

Based on these numbers, I think it’s reasonable to expect a UZR of between 6 and 11 for Ryan Goins in 2014. Including a positional adjustment of +2.5 because he plays second base, his total defensive value would like be in the range of 8.5 to 13.5. For comparisons sake, Oliver projects a defensive value of 15.8 for Goins in 2014.

At the end of the day, Ryan Goins success at the major league level will very much come down to what he is able to do with the bat. His defensive proficiency does decrease the level of expectation but in order to become anything more than a replacement level player he’ll need to improve on the .267 wOBA he put up in 2013.