# What to do on 3 and 2

Last Monday the Cards entered the 9th inning down 2-0 to the Marlins and were looking for a comeback. They began the 9th going single, single, walk to load the bases for pinch hitter __Kolten Wong__. Wong worked the count to 3-2 and then this happened.

The Cardinals managed to get 1 run in but ended up losing 2-1. After the game – and during the 9th as well – much of Cards’ fans’ anger was at the umpire who called that pitch to Wong a strike. A ball puts Wong on 1st, narrows the gap to 1, and the Cards still have the bases loaded and no one out. It was a huge call in the course of the game. The leverage index for Wong’s PA was 6.08 (1.00 represents an average PA during the game) so this was about as high-stakes as it gets and the leverage index for that 3-2 pitch must have been much higher than that if LI’s for individual pitches were made publicly available. The enormity of that call cannot be overstated. The Cards __had a run expectancy__ of 2.33 going into Wong’s PA – which would have put the Cards in the lead – and a run expectancy of 1.56 after the PA. The Cards only got 1 and lost by 1.

Make no mistake, the pitch was a ball. Wong was right. However, as you can see from the plot below, that pitch also fell exactly on the line of what is typically called a strike against left-handed batters.

It’s a typical “lefty strike” as it’s come to be known. Now, I don’t want to focus on whether or not the umpire was right but rather I want to focus on whether or not Wong was right to take the pitch and, more broadly, in what circumstances a batter should take (or swing) at a close 3-2 pitch.

It was surprising to me that there was so little vitriol for Wong in that situation. After all, it was a 3-2 pitch and the bases were loaded. He’s got to get that run in. The team can’t afford a strikeout, a popup, or a double play. From the time kids pick up a bat, they hear their coaches and parents say to them, “with 2 strikes, you’ve got to be swinging if it’s close” and “don’t let the umpire determine your AB.” With 2 strikes, “swing if it’s close.” For probably 20 years, Wong has heard some variation of that mantra and he didn’t. But maybe it’s not Wong that was wrong; maybe it’s the mantra that’s wrong.

Baseball Savant and statcast data breaks the area in which pitches are thrown into 25 identifiable zones. They call this their “detailed zones.” The first 9 are obvious strikes and 8 more are obvious balls. Then there are 8 more zones on the edges of the strike zone – top, bottom, right, and left edges – that could, presumably, be called either balls or strikes. The pitch Wong took clearly falls into one of these zones as it was one of those pitches that could be called either a strike or a ball. You would have to think that, when Wong took that pitch, he wasn’t sure that it would be called a ball. He probably thought it was outside but we would all have to agree that it was pretty close – within an inch or 2 of the outside corner.

So far this season, there have been 201,986 pitches that have crossed the plate in 1 of those 8 zones and 10,310 of those have occurred on 3-2 counts like the pitch that Wong took. Just to be clear, we’re not talking about a small sample. The data below tells you what the results of those 3-2 pitches has been so far in 2018 (thanks to __www.baseballsavant.com__ for the help).

Batters are batting .170 on those 3-2 pitches on the edges of the zone and their batting average is .294 if they put those balls in play. Their slugging percentage on those pitches is a paltry .265. It shouldn’t come as any surprise, therefore, that when batters swing at these corner pitches, they’re not hitting them with much authority. This isn’t a fluke, either, as their expected slugging percentage is just .267.

So what does this tell us? For one, if Wong had swung at the pitch, he probably was not going to get a hit. He has about a 17% chance of getting a hit in that situation. (In case you think 2018 is an aberration, the batting average in all of the 2017 season on similar pitches was just .183 so, if you like, we can give Wong an 18.3% chance of getting a hit. Regardless, it’s still pretty small.) For another, if Wong had gotten a hit, it was probably going to be a single. Doubles and other extra base hits do occur on those “edge” pitches but not very often – thus, the .265 SLG.

There have been a total of 7845 swings on 3-2 pitches on the edge of the zone and batters have swung-and-missed at 1655 of them (21%) so there was a fairly decent chance that Wong would have struck out even if he had swung. And if he does put the ball in play, there is still the possibility that he hits into a double play – a result that is even worse than the strikeout. Ninety-one of these AB’s have ended in a double play. Still, he could have hit a sac fly or moved the runners over with a groundout. Even a force at 2nd not only gets the run in but also moves the runner on 2nd to 3rd and allows the next batter up to tie the game. (As it turns out, Matt Carpenter walked in the next PA so, had Wong gotten the runner in without reaching base himself, Carp’s walk would have just loaded the bases. Wong, of course, had no way of knowing that in his PA and, in fact, Carp’s PA may have ended differently if Wong’s hadn’t ended the way it did.)

So there’s some expected value to swinging the bat but it’s not very high considering the likelihood that Wong will make an out. What about if he doesn’t swing the bat? As indicated earlier, there have been 1524 called balls and only 940 called strikes when hitters chose to take. So, 61.8% of the time a batter has taken a 3-2 pitch on the edges, as Wong did here, it’s called a ball and the batter, of course, walks. Without looking at the numbers, I would have guessed it would have been more like 50-50 but the numbers appear to favor hitters. That’s no aberration either. In 2017, 62.5% of the 3-2 pitches on the edge of the zone that were taken by the batter were called balls.

So let’s try and figure out what Wong should have reasonably expected from that situation. If he takes, then he has about a 62% chance of reaching base (but 1 base only) so his expected number of bases in this situation is .62. Since the slugging percentage of a hitter in that situation is .265 that tells us that the approximate number of expected bases if the batter swings in this situation is .265. This obviously doesn’t include reaching base on an error or a dropped 3rd strike, both of which could occur but won’t occur all that often.

Knowing this, the numbers appear to indicate that when a batter has a 3-2 count on him and the pitch is on the edge of the strike zone – since batters have no way of knowing whether the umpire plans to call the pitch a ball or a strike – in a generic situation the batter should almost ALWAYS take. Since the expected number of bases is about .35 more bases when the batter takes than when he swings, it’s really a no-brainer. Now, this may work on a micro level. Obviously, if batters as a whole start to act like this then umpires may change the way they call those edge pitches but 1 Kolten Wong PA is going to have so substantial effect on the macro numbers here. Bottom line – a batter should take the pitch when the count is 3-2 and the pitch is somewhere on the edge of the zone. (There may be slightly different results depending on whether we’re talking about the top or bottom, inside or outside corner, and I can delve into that a little deeper sometime in the future.)

Ok, so in Wong’s situation, he should probably take that pitch. His decision was probably correct. However, we cannot forget the fact that the leverage index was above 6, bases were loaded with no one out, and the team was down 2 runs in its final at bat. This is no normal situation. Batters in 2018 have only faced 218 pitches with bases loaded and a 3-2 count so far in 2018 so when you add in 2017, you get 535 similar situations to the one Wong faced last Monday.

In this particular situation batters are batting a predictable .171 but slugging a little better than in the generic situation, .316. These 535 pitches have led to 292 balls in play and 14 times the batter has grounded into a double play. This is, of course, the worst case scenario. So, theoretically, if Wong had swung the bat, he had about a 4.7% chance of hitting into a double play. Wong is faster than the average runner so maybe we drop that to 3 or 4% or so – not very likely.

Twenty four (4.4%) of these 535 swings have ended up as a sacrifice fly so while that’s a possible outcome, it’s also not an incredibly likely outcome. The Cards had __Paul DeJong__, an average base runner, on 3rd so there’s no reason to change that expectation. __Harrison Bader __was on 2nd so there’s a decent chance that a sacrifice fly would have also moved Bader to 3rd as long as it wasn’t hit to left field. Still, the sac fly is only slightly more common than the double play so that’s not very convincing that Wong should have swung the bat.

A ground ball fielder’s choice might have gotten the run home and some would have also advanced Bader to 3rd but some of the ground balls would have ended up forcing the runner at home and the team would have been in the same situation as it was with the strikeout. And a sac fly or ground out that scores a run but leaves Bader at 2nd doesn’t really help the Cards all that much since this is the last opportunity for the Cards to score. Getting 1 run is basically the same as getting 0 since it’s the 9th inning. The team’s win expectancy changes a lot if the run scores and Bader moves to third but doesn’t change much at all if the run scores and Bader stays at 2nd.

The point of all of this was to determine the following:

Should batters who have a 3-2 count swing at a pitch on the edge of the zone?

Should Wong have swung in this situation, considering the team was down 2 with the bases loaded in the 9th inning?

The answer to #1 appears to be no. In most situations, batters are better off taking the pitch on the edge of the zone instead of swinging at it. Yes, they run the risk of being called out on strikes but batters are far more likely to reach base if they take that pitch. The possibility of hitting an extra base hit isn’t so great that it overcomes the likelihood of making an out in that situation.

The answer to #2 is less clear but still seems to suggest that Wong was correct in taking that pitch. If he swings, it’s possible he gets a run in but a run in without advancing the runner on 2nd helps the team very little. It is possible that he moves the runner to 3rd in addition to getting the run in but it is also possible that a swing results in a double (or possibly even triple) play which is a worse outcome than the one determined by the home plate umpire.

The bottom line is that a batter has no control on whether or not the umpire calls a ball or strike on any particular pitch. Once Wong recognizes that the pitch is close to the edge of the zone, he knows that the umpire’s call is uncertain. Players are taught when they are kids to “swing at anything close with 2 strikes” but that mantra should not apply to 3-2 counts. The evidence seems to suggest that, regardless of the situation’s leverage, a batter should not swing at a pitch on the edge of the strike zone when the count is 3-2.

Thanks to fangraphs, baseball savant, and to @cardinalsgifs for the great pics.

Thanks to you all for reading.