As you know, Kam Loe started the season poorly, so much so that he was sent to the minors for a couple of days, whereupon he came back and put together a hot streak. Right after the All-Star break he slumped in two bad starts, but then improved in his last start on Tuesday. His next start should be tomorrow in Kansas City. At this point the Rangers consider him the number two starter behind Millwood, although arguably it should be McCarthy, Loe, then Millwood, but the argument can be made for any of the three in any order. In fact, their season ERA+ is almost identical, Millwood and McCarthy at 83 and Loe at 82. Depending on how far back you go, and how you slice and dice the numbers, you can get any of them to be the “best” this season. What bodes well for the Rangers is that there are three pitchers who are competitive, and who over anything from one to three months have performed well, with ERAs under 4 depending on the time period you choose. If they can put that together for a longer timeframe (e.g. the 2008 season), the Rangers might be able to step up and contend.
Today I’m focusing on Loe’s last four starts, because that contains his two bad starts surrounded by two good ones. Having crunched a bunch of numbers, I surprisingly did not come up with any great conclusions based on things like his release point, what he was throwing, break, etc. For a good start with a lot of sink on his pitch, I also found a bad start with a lot of sink. Movement in and out, the same. Break, the same. Speed, the same.
Looking just at speed as an example, we can see there were slight differences there, but nothing that jumps out and shows us something. In his good start on July 6, his fastball was running 85-91. The bad start on July 14 it was 83-89. Bad start on July 19 was 87-90. Good start on July 24 was 87-92. Of these starts, three were at home and the speed was measured at 50 feet from home plate, while the fourth, the July 14 start, was in Anaheim and measured at 40 feet, which would suppress the speed a little. Given that the speed as it crossed the plate was almost identical between the four starts, I’d suspect the difference between a 40 foot and 50 foot measurement is about two miles an hour, which would push the July 14 speed up to 85-91, the same as the prior start. So he showed pretty much the same speed in all starts, with a low from 85-87 and a high from 90-92.
The one place I have not looked at in great detail was where the ball crosses the plate. This could be considered the ultimate position to look at, because it makes all the difference between a ball and a strike, and whether a hitter is going to swing. If it’s too far off, advantage batter. Too close, and you better have good stuff or they’re going to tee off on you. So looking at this might tell us something.
First of all, a little explanation of what I did. I broke down each start by the description of the pitch in the data. I then took the balls and called strikes and charted them, to get an idea of where the strike zone was being called on that day. This can differ quite significantly. By using just balls and called strikes, we’re not seeing how the batter influenced the call (by swinging his bat or hitting the ball), so this should be what the umpire’s influence is. John Walsh has done some great work on how the strike zone is being interpreted compared to Gameday, much more detailed that I am doing here. In my analysis of these four starts, I found the strike zone varied a little, but the left side (from the catcher’s viewpoint) was around 1.2 feet off the center of the plate, and the right side about 0.5 feet off center, except the last start where it was 1.1 feet off. This seems to show a huge bias toward the left side, or the inside for a right handed batter. All I’m doing is looking at a chart of the pitches and see where the umpire stopped calling strikes and started calling balls, which of course can get a little mixed up, as sometimes they call balls inside of where they call strikes. So not necessarily mathematically accurate, but I think good enough for my purposes. As for vertical, this gets a lot more confusing, because players are all different heights, but in general the strike zone was about 1.8 to 3.2 feet off the ground in these starts. I’ll show the strike zone for each start, and you can see how it varies from game to game.
All these charts are sized the same, from -2 to 2 feet horizontally, and 0 to 4 feet vertically. Although I continually reference a right hand batter, for orientation purposes, the charts do show all pitches regardless of who was batting. These charts show fouls, swinging strikes, balls in play that were outs, and balls in play that were hits (which includes the variation they term “in play, run(s)”, although technically something like a sac fly is an out as well, but I chose to include it here because it was a bad thing for Loe). So effectively this is a chart of “what the batter swung at”.
The 7-6 start was Loe’s best of the four starts, with a game score of 62. He pitched six innings, gave up five hits and three walks, struck out three, and didn’t give up a run. You can see that they mostly only swung in the strike zone, or a little down. The chart of balls and called strikes (not shown) shows a lot of pitches down, a lot of pitches inside, a few outside, and a few up and in to a right hand batter. The pitches swung at down out of the strike zone suggest that he was getting good vertical movement. There is no great pattern to see here for individual result types. The hits (red dots) are a little more central, as you’d expect, and the swinging strikes are all down and in (to a right hand batter), but remember the small sample size effect, there are only five and three of these respectively.
On 7-14 in Anaheim, Loe had his worst start of the four, only lasting 67 pitches in 2.2 innings, ending up with five runs on five hits and five walks, with two strikeouts. The balls and called strikes were all either down or inside, mostly inside. This is the start where they said afterwards that he’d had a lot of trouble keeping it in the strike zone. Again not much of a pattern, a lot of pitches down and away, or down in the strikezone, not much up. This almost seems the opposite of what you’d expect, in that usually the advice is to keep the ball down and you’ll stay out of trouble. In this case, they’re swinging at a lot fewer pitches up high, and got better results from it. Could it be possible that he was getting too much movement, and it was moving out of the strike zone? Comparing the called strike zone, very similar to the last start vertically, but horizontally it was called much more inside to the right hand batter. This is also counter-intuitive, you’d expect a wider strike zone to benefit the pitcher.
This start on 7-19 looks closer to 7-6 than to 7-14, although really what I see is that hitters are swinging at a lot more pitches out of the strike zone. Again, that should mean advantage pitcher, but note how much narrower the strike zone is, the right hand side stays the same but the left side is down to almost one foot, and the smaller strike zone means advantage hitter. This start ended up being six innings, six runs, on six hits and four walks, with two strikeouts. Balls and called strikes were much wider inside and outside than the previous two starts, and not as many were low. This time the vertical hold was okay, but the horizontal hold was messed up. Although it is much more variable, there is still no pattern to the result types, with them coming everywhere across the strike zone. If you look closely, you’ll see the red dot at bottom right in the strike zone (just down and right of a green dot), that red dot is the only home run he gave up in these four games.
His last start was back to being good (although with a game score of 51, good is a relative term, maybe average would be more like it). 6.2 innings, three runs on seven hits and two walks, five strikeouts. Any pattern here? Results are throughout the strike zone. Hits are all on the left side of the plate. A lot of swinging strikes down. Balls and swinging strikes show everything down, or inside. But still, overall, nothing that leaps out and grabs you. The strike zone is much wider, with the right side finally getting away from the six inch mark, and jumping over a foot to be even with the left. I’ll have to go back and take a look at other pitchers, and see if they showed similar variation, and whether it can be traced to different umpires, or a different way the Gameday system is measuring the strike zone. I would have thought it would measure evenly across the plate, since that is a fixed position, but it’s possible it could be off for a while, since many of the other stats have shown a bias in one direction or the other.
So what’s the grand result of all this? Not much at all. From these starts I cannot prove anything about location in the strike zone causing specific results. There may simply be too little data to show the patterns we desire. He’s giving up hits everywhere, foul balls everywhere, and making outs everywhere. Swings and misses are down more often than not, which you would have predicted knowing that he’s a sinkerballer who will drop more than the usual pitcher, causing batters to swing over the ball as it drops below the bat.
I’m a little disappointed in this conclusion though. I would have loved to see something stand out and be able to say “oh, that’s why he was being beaten”, but there’s nothing there. At best, I can look at the strike percentages and see a pattern: in his two good starts, he threw 61% and 66% strikes, but in both bad starts it was 55%. Doesn’t seem like much, does it, but if you’re throwing a hundred pitches it’s six or eleven extra strikes, which means either fewer walks or more balls swung at.
There was a famous study a few years ago which created a stat called BABIP, which stands for Batting Average on Balls in Play. Basically it took away strikeouts and home runs, things that fielders couldn’t affect, and looked at what happened to balls actually hit by batters. It showed that for pitchers, there was no consistency, in that anyone from a Roger Clemens to a Vicente Padilla could have a similar BABIP. The deciding factor on how good pitchers turned out to be was the strikeouts, home runs, and walks, in that a good pitcher gave up less of these, and a bad pitcher gave up more. My very tentative conclusion, looking at just these four starts from Kam Loe, is kind of similar: if he throws balls, he will end up with bad results, but if he throws strikes, it is not what he threw but what the batter does that matters. Once it is in the strike zone, the result across the zone is so variable it can’t be seen (in these starts). It’s when the batter makes contact that counts, and what the pitcher is throwing doesn’t seem to affect the outcome. Is this true? Hardly believable, but then neither was the BABIP study. Maybe this is simply a manifestation of what BABIP shows, that if the pitcher isn’t throwing a ball, the result after that is random. Before I make this a firm conclusion though, I’d want to look at a lot more pitchers and a lot more games.