Boone, Kay, O’Neill: 3 Blind Mice

Aaron Boone certainly isn’t alone among Major League Baseball managers who get out managed by the likes of Alex Cora, A.J. Hinch, and Bob Melvin.  But last night the Yankees fearless leader showed it doesn’t really matter who the opposing manager is by again making what I would call sub-optimal strategic decisions.  And also once again, left to their own intellectual laziness, Michael Kay and Paul O’Neill not only didn’t point out Boone’s poor decision making, they actually condoned it with their version of statistical analysis.  That is to say, the type of analysis with logic in which a 6th-grade debate student could punch holes.

I don’t know where David Cone is, but he’s sorely missed.

Bottom of the 8^th, Yankees batting, leading by a score of 1-0 with the top of the order coming to bat:

D.J. LeMahieu and Gary Sanchez both singled to set up 1^st and 2^nd with nobody out and Brett Gardner due up with Clint Frazier (or “Fraser” as O’Neill says) and Gleyber Torres to follow.  Michael Kay, in his infinite wisdom, immediately suggested that a sacrifice bunt was a good play in this situation.  And we’re off…

He had someone check for him, then came back to tell us that the Yankees chance of scoring with 1^st and 2^nd and nobody out was 63%, and with 2^nd and 3^rd and one out, it was 66%.  Therefore, this might be one of those times a bunt is a good play.

(Breathes deeply, exhales, grabs Jameson’s…)

First:  Again, those numbers assume that the bunt will be successful.  We know – well, those of us that pay attention anyway – that a bunt attempt fails more than 40% of the time.  So in this instance, there’s close to a 50/50 chance that the Yankees would end up with 1^st and 2^nd and one out, which has 41% chance of a run scoring.  Only a 20 something percentage drop, no biggie…(rolls eyes…)

Secondly, those are the percentages for one, and only one, run scoring.  Why in the name of Earl Weaver would anybody play for one run before the bottom of the 9^th?  Pro tip:  If you’re the home team in the 8^th inning, the other team gets to bat again and they sure as hell aren’t going to play for one run.  They’re going to try to get 2, 4, 6, etc.  If you’re the Yankees, and you do get one run across, the game is…not…over.  If you get two, three, five runs – the game is over.  You do not play for one run unless it’s the bottom of the 9^th or later – EVER.

“So, what is the run expectancy when going for multiple runs in that spot?”

That’s a good question.  If you asked it, you ask better questions than Boone, Kay, and O’Neill.

Run expectancy for 1^st and 2^nd and zero outs is 1.52.  With 2^nd and 3^rd and one out, it’s 1.32.

Stop playing for one run and STOP BUNTING.

In case you missed it, here’s what happened:  Gardner missed two bunts attempts, putting him in a 1-2 count in which he then tapped a grounder to 2^nd base with an 88 mph exit velocity for a force out.  With that, the Yankees’ run expectancy went from 1.52 to 1.16.  Had Gardner, who is nowhere near a double play threat, been allowed to swing from the get-go, my guess is he would’ve done better than a 36 point drop in run expectancy.

Then Gardner, with the 4^th and 5^th batters in the lineup coming up, stole 2^nd base.  An attempt in which, had he been thrown out, would’ve dropped the run expectancy to 38%.  But he made it, which was then followed by a Frazier strikeout and an intentional walk to Torres.  Gio Urshela then followed with a 2 run single to right field which moved the win expectancy from 87% to 97% and essentially ended the game.

Again, just because it worked out doesn’t mean the right calls were made.  It worked out because 3 hits and an intentional walk led to 2 runs – not because of managerial decisions.

There is a chance, come October, that the Yankees will have Aaron Judge, Giancarlo Stanton, Aaron Hicks, Luis Severino, Dellin Betances and Didi Gregorius on the field, and they will be better over 7 games than the Red Sox, Astros, and Dodgers.  But we had better hope so because if it comes down to Xs and Os, we have a problem.