As promised in the March 29 newsletter, I wanted to study the impact of the most infamous tactic of the Orix Buffaloes’ 2020 season, putting runners on the corners in motion with two outs. It all looks cut and dried because, well, former Orix manager Norifumi Nishimura is an old-school guy, and the visual impact of having the third-base runner walk into an inning-ending out is hard to put out of your mind.
So how bad were the Buffaloes in those situations? They were bad, although not the worst in Japan, but Nishimura was perhaps the most engaged in this tactic last season.
I pointed out in the newsletter that under Nishimura, the Buffaloes scored three runs in those situations, but were caught stealing five times. That should be two runs since I’ve decided to eliminate wild pitches, passed balls, and runners picked off and stuck to those things teams initiate with their baserunning.
Under his successor, Satoshi Nakajima, the Buffaloes ran out of one inning in 15 such situations, but never scored on a play initiated by their base runners.
Over 720 NPB games last year, managers in those situations tried to steal 141 times with 14 of those being run-and-hit plays in which the batter swung and missed. They made 19 outs and scored four runs.
We knew Nishimura did it a lot, but before this I didn’t have a sense for how he really owned that tactic. In 2020, he stole that mantel as it were from Seibu Lions manager Hatsuhiko Tsuji, who led NPB in stolen base attempts with two-outs and runners on the corners each season from 2017 to 2019.
The manager who owns it, in the sense of being the best at it, is the one with the pinch-runner fetish, the Yomiuri Giants’ Tatsunori Hara, who scored three runs off those plays in 2019 while going 17-for-19. In 2020, Hara’s team scored once while going 18-for-19.
Nishimura only managed half a season in 2020 but during that time his team scored half of NPB’s runs and made a quarter of the outs. In 2019, Nishimura’s Buffaloes ran into three outs on these plays, but because they scored once — on a rundown that ended the inning — and collected seven other free bases, they came off better than the three teams that were 3-for-4 stealing without scoring.
To get a sense of the risks and benefits of sending the runners, here are the average run expectancies in NPB from 2017 and 2018.
2 outs, Runners 1st, 3rd = .497.
The most common results when running in that situation are:
- 3 outs, no runs = .000
- 3 outs + 1 run = 1.000
- 2 outs, Runners 2nd, 3rd = .524
- 2 outs, Runner 2nd + 1 run = 1.33
- 2 outs, Runner 3rd + 1 run = 1.34
Improving to runners on 2nd and 3rd is the most common outcome. It’s practically a free base with a half-way fast runner on first. The payoff, however, is small, less than a three percent increase in the chance of scoring, and the cost high, going from a 49.7 percent chance of scoring one run to zero. To cash in on that easy gamble, teams need to be successful 38 out of 39 times – if they never try for home plate. If they try to score from third, they need to be successful half the time to make it work.
Of course, context matters, and this data doesn’t control for that. The cost is a bit higher with the cleanup hitter than the No. 8 or nine guy, and less – in terms of wins and losses if it’s a one-sided game in the late innings.
If we’re ONLY talking about the runner on first going, the success rate needs to be 98 percent to break even, and from 2017 to 2019, NPB teams were successful 89 percent of the time. The upside is the possibility of scoring a run.
Under Nakajima, the Buffaloes took those free bases but didn’t push it.
That’s the Alex Ramirez way. My more detailed baserunning data starts from 2017, and from then until 2020, the BayStars went 20-0 stealing second, but never scored except on passed balls and wild pitches.
NPB results
The table below gives the total number of situations, the number of runs scored, outs made, the average expected run gain per play and the total cost for the season. The winner for 2020 was actually a manager, who lost his job, Rakuten’s Hajime Miki, and his total gain of scoring one run from stealing nine times without being caught was an increase of 1.05 runs over 120 games.
The Eagles were worst in 2019, in a tie with the Yakult Swallows, who were worst again in 2020, making three of the four CL’s outs. The other thing you’ll notice is that this tactic is more of a PL thing. Even so, even the worst team at this is not costing itself a lot of runs.
To put it in perspective, having the pitcher bat eighth in your lineup will probably give you four or five extra runs a year.
| Team | Total | Runs | Outs | Avg gain | 2020 Gain | 2019 Gain |
|---|---|---|---|---|---|---|
| Giants | 19 | 1 | 1 | 0.042 | 0.80 | 1.82 |
| Dragons | 5 | 0 | 0 | 0.027 | 0.14 | 1.22 |
| Carp | 6 | 0 | 0 | 0.027 | 0.16 | 1.02 |
| Tigers | 16 | 0 | 0 | 0.027 | 0.43 | 0.73 |
| Lions | 10 | 0 | 1 | -0.078 | -0.78 | 0.60 |
| 'Stars | 4 | 0 | 0 | 0.027 | 0.11 | 0.24 |
| F's | 12 | 0 | 1 | -0.060 | -0.72 | 0.11 |
| Hawks | 10 | 0 | 2 | -0.078 | -0.78 | 0.10 |
| Nakajima* | 15 | 0 | 1 | -0.008 | -0.12 | 0 |
| Nishimura* | 13 | 2 | 5 | -0.049 | -0.64 | -0.28 |
| Orix * | 28 | 2 | 6 | -0.027 | -0.76 | -0.28 |
| Eagles | 9 | 1 | 0 | 0.117 | 1.05 | -0.42 |
| M's | 9 | 0 | 1 | -0.089 | -0.8 | -0.42 |
| Swallows | 13 | 0 | 3 | -0.094 | -1.22 | -0.42 |
