First Dataset of the Year: Two Week Team Trends

I’ve waited two weeks and I can’t wait any longer. There isn’t much data to go on after half of April but I will find out what I can. I will take statistics from Fangraphs and look at them for a long time until something clicks, just like I always do. Players have had only 50 or so plate appearances, and we can’t find out very much from that, so I settled on whole-team statistics. Even better, I launched by delving into plate-discipline statistics.

Plate-discipline statisitcs tell you about the process of a plate appearance rather than the result. In small samples, you can learn more about a hitter’s process than you can about his ability to get results. In two weeks a hitter has 50 plate appearances but he may face 200 pitches. It’s a long enough period to constitute a hot streak or a slump, and it’s hard to distinguish those from genuine improvement or decline. Drastic fluctuations in plate-discipline statistics are less common because each professional hitter goes up there with his own unique agenda and mindset, which has been validated by his presence in the majors.

So we’ll look at plate-discipline statistics in aggregate plus some others. Perhaps by the end of this we can come to a better understanding of how offenses have operated this season.

We’ll start with O-Swing%, which stands for percentage of swings outside the strike zone. Imagine [Number of swings at pitches outside the zone] divided by [total number of pitches outside the zone]. The lower this number, the better, obviously, because pitches out of the strike zone don’t tend to be hit very often or very hard. Better to let them go for balls and force the pitcher to throw another. Below I have a table of our 10 teams, each next to their 2014 O-Swing%, which is itself next to that team’s 2013 O-Swing%. The difference from last year to this year is in the last column, and the first column of numbers signifies that team’s MLB rank for its 2014 O-Swing%. This is how all of these will be formatted. All data courtesy of Fangraphs and through Sunday’s games. Now, what can we learn?

oswing

I think the most interesting thing here is the overwhelming decrease from last year to the start of this year. Players all around are chasing pitches less, and I have to believe the figures will pull back toward normal. No one could stay at 26 percent for the whole season, but the A’s clearly have the best pedigree. And we can see that the Giants, Padres and Rockies are continuing their free-swinging ways. They did not have much roster turnover this offseason, so we should expect them to finish near the top of the league yet again. The Angels traded Mark Trumbo but that isn’t enough to explain their steep decline. The Astros have basically the same roster. My bet is that their big drop is a mirage.

contact

Another large-scale trend! This time we see that only two teams are making more contact than they made last season. Perhaps hitters are still shaking off some rust in April. Pitchers do start spring training earlier than hitters, which could conceivably put hitters at a disadvantage in the early part of the season. It is even possible that hitters anticipate this disadvantage and that’s why they are more selective with their swings.

The differences here are small compared to the differences we saw in O-Swing%. Those figures ranged from 26 to 32, these from 76 to 82, but each extra percentage point carries less weight here, just as an extra dollar matters more to you than it does to some rich jackass.

The Rangers are good at contact yet again, even though they’ve reworked their lineup considerably. The Giants have been scoring well, but behind that there’s a disconcerting combination of less contact and more swings at bad pitches.

swstr

Swinging Strike percentage is in many ways counterpart to Contact percentage. It measures how many of a player’s swings were whiffs. Here we see that most teams have been whiffing more, which must be the case if they are making less contact. There isn’t much more to learn from this one, but remember that if you are dealing with small sample sizes, SwStr% is a good proxy for Strikeout percentage, just as O-Swing% is a good proxy for Walk percentage.

ppa

Pitches per plate appearance will pull towards the center as the season goes on, the center being around 3.80. But the Twins have a P/PA over 4.00 for the second straight year, and the A’s are right up there with them, so there is room for some differentiation. In the west, the A’s and Rockies hold comfortably the edges of the spectrum. The Rockies are impatient but you would be too if you hit in the best hitters park in the history of the major leagues.

babip

Now let’s look at BABIP, Batting Average on Balls In Play, the most famously fickle stat out there. Except for the Astros, these numbers were more consistent than I expected them to be. The Rockies stay near the top thanks to Coors Field, but the Dodgers and Diamondbacks are also taking strong figures and building on them. Certainly both teams have plenty of good hitters.

We have not seen in these tables anything as dramatic as the drop in BABIP for the Astros. It is staggering but not unexpected; in fact it is just this kind of statistical noise that makes BABIP so renowned in analytic circles.

hrpa

For an entire season it is rare for a team to hit home runs in more than three percent of its plate appearances. So we see disappointment ahead for the Angels. On the other end the Rangers have had a curious power outage. Probably not what they expected when they signed Shin-Soo Choo and traded for Prince Fielder. Adrian Beltre and Jurickson Profar sit on the disabled list, meaning this year Texas will probably have to make do with less power than before, unless Fielder can summon his absolute best.

The Giants have more than doubled their rate of home runs: their output is 203 percent of what it was last year, making for a percentage difference of 103. Percentage difference frames the improvement or decline in terms of proportion to the original figure. Remember what I said about the rich jackass. Well, the Giants’ HR/PA spike is like giving a beggar $1.25 when they already have $1.20 in the cup. The rich man would also have to double his money to get the same percentage difference as that pauper.

With that in mind, I gathered the percentage differences for the stats listed above and put them into one table. At the end of the table I included the percentage difference in runs scored per game. I thought that by clever use of color gradients I could show which factors have been most influential in each team’s early season fortunes.

runsscoredmatric

This is the key table. It brings us back to the land of results, and we can assign credit for each team’s improvement or decline. If the runs have resulted from a large spike in one category, we should consider the pattern unsustainable. So it is with the Giants. Run scoring is up a robust 31 percent compared to last season and yet only the home runs have improved, and those in drastic fashion that cannot be found elsewhere on the table. Same goes for the Angels, although they also have a nice improvement in O-Swing% to lean on.

Only the Rockies have improved the offense without extra home runs. The BABIP is up, the whiffs are down. Perhaps they are hitting more doubles than before, or stringing their hits closer together. In any case, they are scoring more and there’s no obvious flaw in the foundations of that improvement so far as I can see.

In a helpful development, three of the four teams with reduced run scoring have also sharp declines in home runs. The Padres did not have a good offense last year and to take away what little power they have seems cruel. Joffrey is dead, so that kind of cruelty should abate quickly. The A’s and Rangers should near their figures for last season given more time. Watch out for that Rangers offense when it finally clicks.

The Astros are hitting home runs at a respectable rate, much more often than last year, but it is that sharp decrease in BABIP–an aberration–which is keeping their offense down right now. Besides that their offense seems rather improved. The rebuilding process in Houston marches onward and upward.

These percentage differences show you the differing levels of volatility associated with each stat. Pitches per plate appearance and contact percentage are steady as can be, which is why no team is more than five percent off last year’s figures. On the other end of the spectrum are BABIP and HR/PA. Homers have come in bunches for some and eluded others almost entirely (Kansas City has just one on the year). The Giants’ real work will start once the balls stop flying out of the ballpark.

Leave a Reply