Ok here's an excerpt from Chapter 8...
So often in baseball, a squander by one team is followed by the other team scoring the very next time up. Boston had a great chance to cut into New York’s lead in the seventh, and even had a real possibility of tying the game or taking the lead with one big swing of Manny Ramirez’ bat. But Tom Gordon escaped, and the Yankees exhaled, and came up to bat looking to add insurance runs and put Boston away.
Timlin stayed on to face Cairo, Jeter, and Rodriguez. Cairo led off and ripped the first pitch he saw into deep left center. It one-hopped the Green Monster. Damon tracked it down and fired – well, with Damon’s subpar arm, “fired” probably isn’t the right word– it to Cabrera, but Cairo had a stand-up double to start the inning.
At this point, the Yankees had two main options. Old-school baseball dictated a sacrifice bunt by Jeter, which could advance Cairo to third, which would have meant that a productive out could have scored a crucial insurance run. But the modern game is played differently.
Baseball statistician Tom Tango has done great work helping to explain the mathematics of baseball. One question that comes up all the time is: given the current situation, how many runs can we expect to score? And if we bunt here and move a runner along, how many runs can we expect to score in that situation? If the second situation has a lower run expectancy than the first, bunting makes little sense. Why, after all, put yourself in a situation that reduces the number of runs you can expect to score?
Statisticians have analyzed literally hundreds of thousands of baseball games to determine the expected runs in any given situation, based on real-world data. There are three primary situations where managers typically call for a sacrifice (this is not counting when a pitcher is up, in which case sacrificing is almost always the normal – and correct – play). The first is when there’s nobody out and a runner at first. Managers will sacrifice in order to move the runner to second – “scoring position” – in the hopes that one of the following hitters will get a single. That single likely will drive the runner in. But hits of any kind are hard to come by in baseball – even the best batters make outs far more than they get hits.
In the case of a runner at first and nobody out, the expected runs are 0.86. That is, if you get a runner at first to lead off the inning, you can expect, over time, to score 0.86 runs. This is because even though he’s another base further from home, there are still three outs to play with, which means more opportunities for an extra-base hit, or a play that can extend the inning. But if you sacrifice the runner to second, your expected runs drops to 0.66. In other words, you’re not only sacrificing an out, you’re sacrificing 20% of a chance to score a run in the process. Outs are the most precious currency in the game. Normally, a full 9-inning game provides 27 outs to spend. Giving away outs to marginally increase your chances of scoring a run is generally bad business, but it’s especially bad business if it decreases your chances of scoring.
A second case where we often see sacrifice bunts is when there’s a runner at second with no outs, and he is sacrificed to third. This is the situation the Yankees were in after Cairo’s double. Traditional baseball thinking held that if you have a runner at second, you need a hit to the outfield to score him, but if you have a runner at third with less than two outs, he could score on a wild pitch, a passed ball, a balk, an infield ground out, or a fly ball to the outfield, in addition to a hit of any kind. Scoring from third – even with one out – seemed more likely than scoring from second with no outs.
Run expectancy tells a different story. If you have a runner at second with nobody out, you can expect to score 1.10 runs. But if you have a runner at third with one out, you can expect to score 0.95 runs. Yes, in fact, a sacrifice in this situation yields fewer expected runs than a runner at second and nobody out. By Tango’s methodology, Boston’s win expectancy would actually rise marginally if Torre elected to have Jeter sacrifice Cairo to third.
By the strict mathematics, Joe Torre’s best play, especially with Jeter, Rodriguez, and Sheffield due up, was to leave Cairo at second and take three shots at getting him in.
The third most common scenario is when you have runners at first and second and nobody out. In that situation, your expected runs are 1.44, but if you bunt them over, now you’re looking at runners at second and third with one out, and your expected runs drop to 1.38. A small drop, but a drop nonetheless.
The bunt has a long and complicated history in the game of baseball. It showed up first in the early years of the game, in the 1870s, and fans criticized it as being unmanly. The Boston Globe in 1873 called bunting “the black game” (not in reference to skin tone but rather that bunting was a sign that a hitter knew he wasn’t very good and had to resort to trickery).
President William Taft joined the chorus in 1904 saying that fans should see players “hit it out for all that is in them”. In the movie Moneyball, Oakland A’s General Manager Billy Beane, played by Brad Pitt, shares with his players his philosophy on bunting. He says, “No bunting whatsoever. And if someone bunts on us, just pick it up and throw it to first. Don’t try to be a hero and go to second. Let them make the mistakes. And when your enemy’s making mistakes, don’t interrupt them. They’re giving you an out, man, they’re just giving it to you. Take it and say thank you.”
But even teams that understand the statistical reality behind bunting will bunt more in the playoffs than the regular season. Jeter, for example, bunted twice as often in the postseason (as a rate stat) than in the regular season. Teams play more “small ball” in the postseason, falling back on traditional thinking and run manufacturing than during the regular season. Part of this may have to do with the notion that they’re facing better pitching in the postseason, making it more difficult to score runs. Therefore, bunting and sacrificing and making productive outs become more important.
Whatever Torre’s rationale was, he decided to have Jeter bunt. Just as McCarver said, “He could be bunting here”, and pointed out that Jeter had executed a sacrifice bunt the game before, Timlin spun around to bluff a throw to second, trying to keep Cairo close to the bag. Any extra step Cairo could get towards third would make it more likely he could advance safely on a sacrifice.
“I think you give up too much by bunting here, however,” McCarver said. “Jeter’s natural stroke is the other way. Millar’s in (at first).” The idea, of course, is that with Jeter especially, his preference for going to the right side of the field – like he did in his previous at-bat on the three-run double – meant that even if he hit the ball on the ground at an infielder, that would be enough to get Cairo to third. So why not try to get a hit to right field, knowing that even if it resulted in an out, it would accomplish the same thing as giving yourself up in a sacrifice bunt.
Timlin delivered and Jeter squared to bunt. The pitch was low and Jeter let it go by for ball one. Sutcliffe pointed out how, with the infield in to deal with a sacrifice, huge holes opened up if Jeter were to swing away. But he squared around again and bunted a chopper in front of the mound. Timlin fielded it, took a quick glance at third, realized he had no play there, and took the safe out at first.
Torre played his card and now had Cairo at third with one out and Rodriguez coming up. The previous game, in the 11th inning, the Yankees had roughly the same scenario. In a tie game, Cairo had led off with a single, and Jeter advanced him to second with a sacrifice. Rodriguez had the chance to drive him in, but lined out to short. After Sheffield and Matsui walked, Bernie Williams flied out to end the threat. This time, Rodriguez came up only needing a fly ball to get Cairo in.
Boston brought the infield in, hoping to cut off a ground ball and prevent Cairo from scoring. Advanced metrics are helpful in this situation as well. On balls put in play, batters hit .296 with the infield at regular depth, but the runner from third scores 63% of the time. When the infield is in, batters hit .366, a huge increase in the likelihood of a hit, however, the run scores from third only 49% of the time.
Timlin’s first pitch was a strike. He missed with the next pitch to even the count. Boston could ill afford to let this run in. Overcoming a two-run deficit would be hard enough against Gordon and Rivera. Rodriguez was a more than proficient run-producer. During the 2004 season, he had come up with a runner on third with less than two out some 30 times. In those situations, he batted .333 with a .413 on-base percentage and a 1.046 OPS. Moreover, he drove in 31 runs in those 30 opportunities. He was the right man for New York to have up at the plate.
On the 1-1 pitch, Timlin missed again to make the count two balls and one strike. The difference in baseball between 2-1 and 1-2 is dramatic. For Rodriguez, his odds improved dramatically with that one change in the count. In 2004, with a 1-2 count, he hit .142 with a .148 on-base percentage and .412 OPS in 106 at-bats. But when he got the count to 2-1, he hit .476 with a .477 on-base percentage and 1.239 OPS in 42 at-bats. Timlin missing on the third pitch put him in a huge hole. Rodriguez’ OPS jumped to 1.528 on a 3-1 count, but back down to a .587 OPS on a 2-2 count. The next pitch was huge.
Buck said, “Timlin pitched a scoreless seventh. He’s going to have to do some work to pitch a scoreless eighth after the double by Cairo.”
Timlin threw his best pitch of the night, an 85-mph slider low and away that Rodriguez hacked at and missed. On the 2-2 pitch, Timlin fired a fastball high and tight that Rodriguez swung at and missed for strike three.
“Mike Timlin just found a hole,” Sutcliffe said. “He went up and in to get ahead in the count, and finished him off in the exact same spot.”
Obviously I could have picked a bunch of excerpts, but for the SOSH crowd, this one works.
So often in baseball, a squander by one team is followed by the other team scoring the very next time up. Boston had a great chance to cut into New York’s lead in the seventh, and even had a real possibility of tying the game or taking the lead with one big swing of Manny Ramirez’ bat. But Tom Gordon escaped, and the Yankees exhaled, and came up to bat looking to add insurance runs and put Boston away.
Timlin stayed on to face Cairo, Jeter, and Rodriguez. Cairo led off and ripped the first pitch he saw into deep left center. It one-hopped the Green Monster. Damon tracked it down and fired – well, with Damon’s subpar arm, “fired” probably isn’t the right word– it to Cabrera, but Cairo had a stand-up double to start the inning.
At this point, the Yankees had two main options. Old-school baseball dictated a sacrifice bunt by Jeter, which could advance Cairo to third, which would have meant that a productive out could have scored a crucial insurance run. But the modern game is played differently.
Baseball statistician Tom Tango has done great work helping to explain the mathematics of baseball. One question that comes up all the time is: given the current situation, how many runs can we expect to score? And if we bunt here and move a runner along, how many runs can we expect to score in that situation? If the second situation has a lower run expectancy than the first, bunting makes little sense. Why, after all, put yourself in a situation that reduces the number of runs you can expect to score?
Statisticians have analyzed literally hundreds of thousands of baseball games to determine the expected runs in any given situation, based on real-world data. There are three primary situations where managers typically call for a sacrifice (this is not counting when a pitcher is up, in which case sacrificing is almost always the normal – and correct – play). The first is when there’s nobody out and a runner at first. Managers will sacrifice in order to move the runner to second – “scoring position” – in the hopes that one of the following hitters will get a single. That single likely will drive the runner in. But hits of any kind are hard to come by in baseball – even the best batters make outs far more than they get hits.
In the case of a runner at first and nobody out, the expected runs are 0.86. That is, if you get a runner at first to lead off the inning, you can expect, over time, to score 0.86 runs. This is because even though he’s another base further from home, there are still three outs to play with, which means more opportunities for an extra-base hit, or a play that can extend the inning. But if you sacrifice the runner to second, your expected runs drops to 0.66. In other words, you’re not only sacrificing an out, you’re sacrificing 20% of a chance to score a run in the process. Outs are the most precious currency in the game. Normally, a full 9-inning game provides 27 outs to spend. Giving away outs to marginally increase your chances of scoring a run is generally bad business, but it’s especially bad business if it decreases your chances of scoring.
A second case where we often see sacrifice bunts is when there’s a runner at second with no outs, and he is sacrificed to third. This is the situation the Yankees were in after Cairo’s double. Traditional baseball thinking held that if you have a runner at second, you need a hit to the outfield to score him, but if you have a runner at third with less than two outs, he could score on a wild pitch, a passed ball, a balk, an infield ground out, or a fly ball to the outfield, in addition to a hit of any kind. Scoring from third – even with one out – seemed more likely than scoring from second with no outs.
Run expectancy tells a different story. If you have a runner at second with nobody out, you can expect to score 1.10 runs. But if you have a runner at third with one out, you can expect to score 0.95 runs. Yes, in fact, a sacrifice in this situation yields fewer expected runs than a runner at second and nobody out. By Tango’s methodology, Boston’s win expectancy would actually rise marginally if Torre elected to have Jeter sacrifice Cairo to third.
By the strict mathematics, Joe Torre’s best play, especially with Jeter, Rodriguez, and Sheffield due up, was to leave Cairo at second and take three shots at getting him in.
The third most common scenario is when you have runners at first and second and nobody out. In that situation, your expected runs are 1.44, but if you bunt them over, now you’re looking at runners at second and third with one out, and your expected runs drop to 1.38. A small drop, but a drop nonetheless.
The bunt has a long and complicated history in the game of baseball. It showed up first in the early years of the game, in the 1870s, and fans criticized it as being unmanly. The Boston Globe in 1873 called bunting “the black game” (not in reference to skin tone but rather that bunting was a sign that a hitter knew he wasn’t very good and had to resort to trickery).
President William Taft joined the chorus in 1904 saying that fans should see players “hit it out for all that is in them”. In the movie Moneyball, Oakland A’s General Manager Billy Beane, played by Brad Pitt, shares with his players his philosophy on bunting. He says, “No bunting whatsoever. And if someone bunts on us, just pick it up and throw it to first. Don’t try to be a hero and go to second. Let them make the mistakes. And when your enemy’s making mistakes, don’t interrupt them. They’re giving you an out, man, they’re just giving it to you. Take it and say thank you.”
But even teams that understand the statistical reality behind bunting will bunt more in the playoffs than the regular season. Jeter, for example, bunted twice as often in the postseason (as a rate stat) than in the regular season. Teams play more “small ball” in the postseason, falling back on traditional thinking and run manufacturing than during the regular season. Part of this may have to do with the notion that they’re facing better pitching in the postseason, making it more difficult to score runs. Therefore, bunting and sacrificing and making productive outs become more important.
Whatever Torre’s rationale was, he decided to have Jeter bunt. Just as McCarver said, “He could be bunting here”, and pointed out that Jeter had executed a sacrifice bunt the game before, Timlin spun around to bluff a throw to second, trying to keep Cairo close to the bag. Any extra step Cairo could get towards third would make it more likely he could advance safely on a sacrifice.
“I think you give up too much by bunting here, however,” McCarver said. “Jeter’s natural stroke is the other way. Millar’s in (at first).” The idea, of course, is that with Jeter especially, his preference for going to the right side of the field – like he did in his previous at-bat on the three-run double – meant that even if he hit the ball on the ground at an infielder, that would be enough to get Cairo to third. So why not try to get a hit to right field, knowing that even if it resulted in an out, it would accomplish the same thing as giving yourself up in a sacrifice bunt.
Timlin delivered and Jeter squared to bunt. The pitch was low and Jeter let it go by for ball one. Sutcliffe pointed out how, with the infield in to deal with a sacrifice, huge holes opened up if Jeter were to swing away. But he squared around again and bunted a chopper in front of the mound. Timlin fielded it, took a quick glance at third, realized he had no play there, and took the safe out at first.
Torre played his card and now had Cairo at third with one out and Rodriguez coming up. The previous game, in the 11th inning, the Yankees had roughly the same scenario. In a tie game, Cairo had led off with a single, and Jeter advanced him to second with a sacrifice. Rodriguez had the chance to drive him in, but lined out to short. After Sheffield and Matsui walked, Bernie Williams flied out to end the threat. This time, Rodriguez came up only needing a fly ball to get Cairo in.
Boston brought the infield in, hoping to cut off a ground ball and prevent Cairo from scoring. Advanced metrics are helpful in this situation as well. On balls put in play, batters hit .296 with the infield at regular depth, but the runner from third scores 63% of the time. When the infield is in, batters hit .366, a huge increase in the likelihood of a hit, however, the run scores from third only 49% of the time.
Timlin’s first pitch was a strike. He missed with the next pitch to even the count. Boston could ill afford to let this run in. Overcoming a two-run deficit would be hard enough against Gordon and Rivera. Rodriguez was a more than proficient run-producer. During the 2004 season, he had come up with a runner on third with less than two out some 30 times. In those situations, he batted .333 with a .413 on-base percentage and a 1.046 OPS. Moreover, he drove in 31 runs in those 30 opportunities. He was the right man for New York to have up at the plate.
On the 1-1 pitch, Timlin missed again to make the count two balls and one strike. The difference in baseball between 2-1 and 1-2 is dramatic. For Rodriguez, his odds improved dramatically with that one change in the count. In 2004, with a 1-2 count, he hit .142 with a .148 on-base percentage and .412 OPS in 106 at-bats. But when he got the count to 2-1, he hit .476 with a .477 on-base percentage and 1.239 OPS in 42 at-bats. Timlin missing on the third pitch put him in a huge hole. Rodriguez’ OPS jumped to 1.528 on a 3-1 count, but back down to a .587 OPS on a 2-2 count. The next pitch was huge.
Buck said, “Timlin pitched a scoreless seventh. He’s going to have to do some work to pitch a scoreless eighth after the double by Cairo.”
Timlin threw his best pitch of the night, an 85-mph slider low and away that Rodriguez hacked at and missed. On the 2-2 pitch, Timlin fired a fastball high and tight that Rodriguez swung at and missed for strike three.
“Mike Timlin just found a hole,” Sutcliffe said. “He went up and in to get ahead in the count, and finished him off in the exact same spot.”
Obviously I could have picked a bunch of excerpts, but for the SOSH crowd, this one works.