And yet some people expect the Oakland Athletics to score runs. I mean, they really do – and they have the numbers for it.
Most of you are probably familiar with the concept of run expectancy. It is a fancy expression telling us how many runs scored on average between a certain baserunners/outs situation and the end of that particular inning. That number tells us what we should expect, or better said, it tells us what other teams did when facing a similar situation in the past.
It is helpful, as there are only 24 different situations in baseball and roughly 200,000 plate appearances every year. Structuring the data and being able to analyze so many events helps us try to answer some of the very basic questions about strategy in baseball. Is it helpful to bunt that runner over from second base? Walk that slugger with the first base open? It also gives some indications why we lately saw the shift from evaluating on-base and slugging as being equally worthy (OPS) to accepting on-base to be more important (wOBA).
Let’s look at them graphs.
Basic Run Expectancy 2000-2009
The numbers in each diamond represent the average number of runs scored after the situation occurred. The first number is the given baserunners situation with no outs, then with one and then with two. They are based on little less than 2,000,000 game situations since 2000. 2010 is not included, as the retrosheet data for 2010 was not yet available at the time.
Some things are both intuitive and obvious. Having baserunners on is good. Having outs is bad. Yet, it is the precision of these number — thanks to the large data samples — that allows us to make some assumptions. And precise they are – the second decimal point is not listed for decorative purposes.
The first number in the first diamond is the simplest one. It almost corresponds to the average number of runs scored in an inning (the difference are the inning-leading home runs). So, with nobody out and nobody on, the teams will on average score 0.53 runs till the end of the inning. You can even imagine it literally, as if every team is given 0.53 runs at the start of every inning and asked to manage it. The teams that go about it carefully, don’t make many outs and advance their baserunners will grow that number to 1 or 2 or in some cases even 10. Most often though, they will lose that initial investment.
So, let’s take a closer look what happens with the run expectancy when the leadoff batter steps in the box.
Such calculations are the core of WPA stat. For every event the situation “before” and the situation “after” are considered and the difference is calculated. In addition to outs and baserunners, WPA also looks at the inning and the score, as well as adjusting the data for park effects.
If the batter makes an out, we have a new situation – nobody on, one out. On average, teams score 0.28 runs from there on. If the batter makes it to first however, say on a walk, the situation is runner on first, nobody out; the situation that yields 0.92 runs on average.
Reaching first base is the worst possible result a batter can do if he doesn’t make an out. Yet even such minimal gain from not making that out is huge – it is the difference between 0.92 and 0.28 runs that will be scored from that point on. Or in other words, drawing a leadoff walk puts 2/3 of a run in the team’s runs bank.
Marginal gains of hitting an extra base hit are all smaller than that. Hitting a double instead will add further 0.24 runs to the savings account, a triple 0.53 and a home run 0.61 (one run is in and the run expectancy remains at 0.53, making the new run expectancy to be 1.53).
You can think of it this way – if you lead the inning off twice, and draw a walk each time, you have done slightly more to help your team score than if you hit a home run and fly out.
Obviously, batting with bases empty is the most favorable scenario for base on balls or on-base-percentage in general to be compared to home runs and slugging. How about with bases loaded, you might say? I will not lead you through every example, and if you want exact numbers – there is wOBA, the stat that looked at such comparisons under every circumstance (home run and an out are indeed worth more than two base on balls, if one considers every imaginable situation).
But one of the main problems for baseball managers is that they can not just call their favorite player who is most suited for the situation when one arises. You can not place a high on-base-percentage player when bases are empty and then chose a slugger when bases are full. You can tweak the lineup a bit, but in the end it is all pretty much random. Sluggers will lead off, and Daric Bartons will bat with runners on the corners just as often as the other way around.
So, if on base skill is more important when nobody is on, and slugging when runners are on, why is it that OBP is accepted as more important? Exactly, because most pitches in the MLB are thrown from a windup.
More than a half of plate appearances occur with bases empty, 85% of them with no or one runner on. You can’t hit all that many grand slams, when bases loaded PA comes along only once in every 40 tries.
Should Daric Barton bunt? Should anybody bunt?
This is basically just a cheap attempt to attract passionate readers. There will be a special issue on bunting later in the week, but for now – let’s not delve into the analysis too deep. Let’s just look at the general averages. And, I admit, this has nothing to do with Daric Barton himself.
The left part of each box is the situation before the bunt, the right one is the situation after. The red and blue squares are placed around the run expectancies before and after the bunt.
Successful sacrifice bunt costs a team an average of 0.2 runs, depending on a situation – the most favorable being 1st and 2nd, nobody out. That’s not to say that bunt can not be a useful weapon, it’s all about when and why and with whom. But, on average, it is not really helpful. More on that in the bunting article.
Sometimes a prevailing feeling on AN is that the Athletics don’t win more games because they don’t play sound, fundamental baseball. Move the runners over more often. Score that runner from third when there are less than two outs. “It’s pathetic! The only thing you need to do is hit a fly ball, you are a Major League hitter, for chrissake!”, the outcries ensue every time a pathetic Athletic fails to do so.
Some other people believe that Major League pitchers are not completely deprived of skills either, and that it is not in their best interest to allow good contact. So, let’s take a general look what happens when that said runner is on third and there is one out.
First data column shows how many runs were scored on average from that point on. The second show the percentage of the innings where that runner from third scored. The data includes early naughts so there is some of Giambi magic in there, making the A’s numbers nicer than they are very recently. On a side note, did you know that Jason Giambi was signed by none other than Ed Crosby? Yes, that Ed Crosby.
A comparison with Anaheim is always good in such cases, because they are the father and the mother of all that is sound, fundamental, gritty and scrappy in baseball. And a quick look tells us that they actually do score that runner from third more often than A’s do. But in most cases it is not how often a team scores, but how much – the A’s actually capitalize more on this particular situation although they do it in fewer, but more productive innings.
But the real question is – if even the Angels fail to hit that sacrifice fly 30% of the time, is it really that automatic? Do you have any idea how often batters actually manage to hit a sacrifice fly when presented with an opportunity? The answer is below, in the last chart in this article. I hope you enjoyed it and that you will tune in for the next chapter – “9th Inning, Playing It By The Book”, coming soon, perhaps even tomorrow.
The most frequent results of plate appearances with runner on third and less than two outs. Sacrifice flies account for only about 13% of all outcomes. The A’s are in line with the rest of the league, with biggest discrepancy being the percentage of base on balls.