THE BOOK--Playing The Percentages In Baseball

126 assists on relay for MLB, or 4 per team.  It doesn’t seem that shocking that one shortstop would wind up with zero.

On relay throws from the OF, doesn’t the OF get credit for the assist and nothing for the infielder?

I don’t think you can have more than one assist per putout.

I am Misterdirt.  You can have multiple assists on a putout.  Hence the 126 assists credited to SS on relay throws in 2001.  That post (linked above on the word “facts") was the second one I made on that thread regarding Jeter.  The initial response is linked above on the word “Jeter” and gives his relay throw assists for 2003-2005.

Considering the presence of Bernie Williams’s arm in center for those years, and Hideki Matsui’s nearly-as-bad arm in left for 2003-5 (not 2001, obviously), I think it’s very likely relays from Jeter were rendered meaningless quite often. And even when he did have a play at the plate, he was probably making his throws from deeper in the outfield.

You do make a good point.

So, a good test for that would be to look at the Fans Scouting Report.  Look for other low-armed CF, and see how those SS do.  For example, Johnny Damon would be another.  Juan Pierre, etc.

Then, see how the strong-armed CF do, Beltran, Wells, Edmonds, etc.

Shortstops get most of their outfield relay assists from the left fielder, not the center fielder.  There were 340 direct from the outfield relay assists in 2003-2005; 155 from left, 126 from center, 59 from right.  Damon had 1 assist relayed by the shortstop, Pierre 1, williams 0.  Wells had 1, Edmonds 6, and Beltran 3.  These numbers are far too few for any meaningful observational or statistical analysis.  Furcal had the most direct relay assists of any shortstop with 18, 6 from each outfield position.

Is it hard to expand the sample beyond 2001,2003-5?

Anthony concentrated on arm strength, but perhaps there’s a park effect. Bigger parks => ball can go deeper => longer throws. Yankee Stadium would qualify as big in left. Don’t know how that would play out as a tradeoff in terms of creating more difficult relays vs creating more opportunities for runners to try to advance ...

It’s hard to conclude from this alone that Jeter is bad at relays; I see him play vs Red Sox fairly often and have never noticed a problem with his relay throws. But certainly it puts the burden of proof on the person arguing that Jeter is great at relays.

Agreed, no one is saying that Jeter is not good at relays.  But simply if someone is going to say he is good, at least present something compelling.

Joe, obviously we need LF/SS relay park factors!

Its not hard to expand the queries to other years, but really what’s the point?  This is a minor skill worth at most a run or two a year to a shortstop and there are too many factors involved to make any accurate assessment.  As Tango points out, its one thing to say that the statistical evidence is lacking to prove a shortstop is superior at making relay throws, and quite another to try and prove definitively that he is either bad or good at them.  The time is better spent trying to get a more accurate evaluation of a shortstop’s main job of fielding ground balls.

It reminds me of the Ichiro argument of having his arm in RF rather than his speed in CF.  An OF legs come into play far more than his arm, and a CF will get 30-35% more leg plays than a RF.  Even if a RF gets 30-35% more arm plays, the sheer number of leg plays overwhelms the arm plays.

For an IF, the leg plays far, far overwhelms arm plays.  Its nice if Jeter is great at relays, or its no big deal that he’s horrible at them.

This can go into the “baseclogging” argument, of having technical merit, but extremely limited in its practical utility.

***

This was my argument of having Ichiro in CF over RF:

Ok, now let’s talk about the value of an arm in RF compared to CF.  In all cases, I will give the benefit of the doubt to Ichiro in RF.

About 30% of all PAs has a runner on 1B.  An OF single is hit in about 14% of all PAs.  Let’s say that one-third are hit to CF and one-third to RF.  (If you have more accurate numbers, please chime in.) There are about 40 PAs per game.

So, this gives us: 162 * 40 * .3 * .14 * .333 = 91

That means 91 times, the RF or CF’s arms are in play.

I figure that 15% of the time, any runner can make it to 3B, regardless of the arm strength, and another 35% of the time that no runner can make it to 3B, regardless of their speed.  That gives us 50% of the time when the OF arm is *really* in play with a runner on 1B.  That’s 45 times.

Now, let’s compare Ichiro to an average RF arm.  Let’s say that the average RF allows the runner to get to 3B 15 times, throws him out at 3B 5 times, and holds him another 25 times.  What does Ichiro do?  Let’s give him 5,10,30.  Seems fair?

Now, we need to quantify that in runs.  A runner at 2B will score 43% of the time, a runner at 3B will score 60% of the time, and a runner thrown out will score 0% of the time (obviously), plus he costs the hitting team an extra out in the inning, which is worth say 0.18 runs.

The average RF will have the runner in a position to score .60*15 + .43*25 = 19.75 runs and the 5 extra outs are worth 0.90 runs, so that works out to 19 runs allowed.

Ichiro will be .60*5+.43*30 = 15.9 runs and the 10 extra outs are worth 1.80 runs, so that works out to 14 runs allowed.

Conclusion: Ichiro’s arm in RF is worth about 5 runs above average.

Now, how much could it be worth in CF?

Let’s say that instead of 45 times that the RF arm is in play, we’ll make it only 35 times for a CF.  Let’s say the average CF allows the runner to get to 3B 10 times, throws him out 3 times, and holds him another 22 times.  For Ichiro, we’ll be modest, and say he allows the runner to get to 3B 7 times, throws him out 5 times, and holds him 23 times.

Work it out, and, conservatively, Ichiro’s arm in CF is worth 2 runs.  I believe it would be worth more.

Anyway, so the value of his arm in RF, compared to CF, is 3 runs, at best.  And more likely, it’s 1 run.

How much is his range worth?  You make about 30 to 35% more plays in CF than RF.  If he’s worth 20-25 runs above average fielding in RF, he’s worth 27 to 33 runs above average in CF.  That’s a 7 to 8 run improvement.

Overall, shifting him to CF would add about 5 to 7 runs.

His legs come into play a few hundred times a year, but his arm comes into play a few dozen times a year.  And when his arm comes into play, it’s the small difference of the base value of 2B to 3b (which is worth .17 runs).  Turning a hit into an out is far more valuable.