2017 Tigers Game 100: in tertia, campanam anulerunt

This blog had its origins in a season I spent following the 2009 Washington Nationals for a time. I was doing an MA in Ontario, and the sacking of Jim Bowden just as spring training began drew my attention to a train wreck about to happen. Despite the horror story that unfolded that spring and summer, I had a lot of fun playing around with the data.

So after 100 games, it feels as if the Tigers are worse than my memory of the 2009 Washington Nationals because of two recent blow-outs against Kansas City.

I decided to check.

On record alone, there's no contest.

2017 Tigers 45-55
2009 Nationals 32-68

It was at about this point that my engagement with the 2009 Nationals would end, as I was about to travel back to England to spend August with my wife and daughters. The 2009 Nationals would 'improve' in the sense that they would finish 27-35, or a .435 winning percentage against a .320 winning percentage after 100 games. The Tigers' current .450 winning percentage is better than either.

But, of course, the Tigers have been getting beaten in blow-outs you know, that's why this feels worse.

Let's see how that stacks up. The definition of a blow-out is a defeat by five runs or more.

Through 100 games in 2009, the Nationals had lost sixteen such games.

The 2017 Tigers have lost... sixteen such games. But, in fairness, that's a higher percentage of losses (29 per cent to the 2009 Nationals' 23.5 per cent). Some of that has to be down to the DH rule increasing run-scoring. How much? I dunno and I don't think I care. Someone could either deduct one from all the Tigers' losses or add one to the 2009 Nationals' losses to get a vague idea, I would think.

So, possibly it's the proximity of two such heavy defeats by the Royals in the past two weeks. And here I think I am going to be critical of the Tigers' management.

On 20th July, the Tigers played very badly in the first inning of a Michael Fulmer start. They then recovered some ground as Chad Bell pitched 2 1/3rd innings of solid long relief. However, in the third inning of his stint the Royals handled Bell's offerings quite roughly.

On 26th July, the Tigers kept the game within reach through six, in no small part due to 2 1/3rd innings of solid long relief from Chad Bell. Bell comes out for the third inning and BOOM, he gets cracked like his cousin Liberty Bell giving up four hits and a hit-by-pitch putting the game almost out of reach. (Warwick Saupould would come on after Bell had hit Alex Gordon and make sure of that.)

Letting the same thing happen within a week strikes me as some inattentive management, but I'm just some humble blogger. I am very supportive of Brad Ausmus as a manager, but in this case the fault is plainly his.

Of course, there are other issues, such as Mikie Mahtook getting doubled off first by Jorge Bonifacio or the aforementioned fielding disaster on 20th July in which I'm not sure I have enough fingers to point at the culprits.

But Wednesday night's bell-ringing could have been anticipated simply by remembering what happened last time.

Nationals' 2009 Postmortem #4: Leadoff Spot

Hmm, I had the chart done about ten days again, then I wound up with too much to do. (Some of that was watching the World Series.)

The Nationals did suffer from a problem here this past season. For some reason, even the supposedly sabermetrically minded Acta seemed to look for a 'classic' leadoff hitter to fill the spot. They did have a 'sabermetric' leadoff hitter, until they traded him to Florida. The chart tells us more than any amount of words by me could possibly do:

Remember, this is not a measure of quality, but of how the different elements in a player's stat line convert to Runs Created. So neither Nyjer Morgan nor Cristian Guzmán walk anything like as much as an archetypal NL leadoff hitter. Most of their offensive 'oomph' comes from batting average, way more than the archetypal NL anything. Willie Harris does walk a lot more, and hits for power, but he doesn't hit for average. Also, Harris is a good player in the sense that he's versatile, but is he really a major-league regular?

The Nationals may feel that they solved their leadoff issues by trading for Morgan, but I'd say they should still be looking.

Nationals' Postmortem #3 - 'Batting Third, Ryan Zimmerman'

So, armed with the Ashley Hexagon, let's take a look at the Nationals' lineup in more detail. In this series of posts, I propose to look at each lineup slot, and the players who batted there, to see if they are more or less like the average National League batter at that position. Then, we'll see what that might tell us about who the Nationals might like to target this offseason, in the 'drive to 75'.

As a benchmark, I'm going to focus in each position on the player who got at least 80 per cent of the Plate Appearances. If nobody reached 80 per cent, then I'll add in players, in order of total Plate Appearances (PA), starting with the largest, until I hit 80 per cent. To make my life easy, I'll start with the third spot, filled by Ryan Zimmerman for about 86 per cent of the total PA. First, let's see how his hitting profile matched the National League generally:


The right-hand axis is an indication of power. The closer one is to that side, the more likely one has power. The bottom axis shows batting average, the further to the left, the more runs are created through batting average. The left-hand axis is how much walks contribute to runs created. The chart shows the proportions of the average National-League batter at the various positions in the lineup. It doesn't actually tell us anything about the quantity of runs created (using Bill James' most basic formula), only how the different elements are weighted in the total.

The data is based not on Zimmerman's performance batting third, but on his totals for the year batting in all positions, and as a pinch hitter. The key thing to note here is that Ryan Zimmerman's 2009 profile is more like that of a cleanup hitter than that of a batter in the third spot. The question becomes whether the fact that he was hitting out the 'wrong' spot hurt the Nationals' ability to score runs. That's something I intend to cover in a later post, but there's food for thought here.

I see two alternatives. Either Zimmerman should be moved to the cleanup position, or one could consider the possibility that he should bat second, since he has the walk rate, but a lot more power. Sabermetric theory tells us the best hitter should bat second, to optimize his contribution in terms of getting to the plate with men on base.

It's an interesting question. Keep him where he is or move him? What would you do?

Nationals' Postmortem #'2 - Lineup Overview

This post is an introduction to an in-depth analysis of the 2009 Washington Nationals' batting lineup using some old-fashioned sabermetric methods. Throughout it all, I'm interested in how Nationals' players compared to the league average. Before we get to the actual measure of quality, I'm going to offer a profile of how the Nationals put together hits, power and walks in the different lineup slots in order to generate runs. As a first step, I'm going to excavate the Ashley Hexagon, which first appeared in the 1997 Big Bad Baseball Abstract.

The Ashley Hexagon was described by Gerry Myerson on pages 100-106 It doesn't tell you a whole lot about quality. What it analyses, using a basic form of Bill James' Runs Created, is where a given batter sources his offensive contribution. 'X' percent came from his hits, 'y' from walks and 'z' from power. Usefully, we can compare that to the league average.

Here's a graph (click on it to see it bigger) showing you the National League average. There are three components. The left-hand graph measures the walk component. The bottom is a contact component, but the right-had graph is power, so a negative number at the bottom means not much power.


Here's another graph, showing the Nationals' average by batting order position. As you can see, the Nationals' get more of their runs created by power than the league average in the heart of the lineup.


That 'z-axis' marks the right-hand boundary of the NL average, but the Nationals' push w-a-a-a-y over to make space for their cleanup hitter. Basically, the heart of the Nationals' order was more likely than the league to create runs through extra-base hits. They also tended to walk a bit more. However, at either end of the lineup there's a distinct shortage of both power and, in the leadoff spot especially, walks. It's a peculiar sort of bifurcation. Sock and eye is all in one place.

Finally, let's take a snapshot of actual quality by utilizing the Big Bad Baseball Annual's Offensive Index. Here, we have one line showing the league average of Runs Created per batting order position matched against another showing that of the Washington Nationals.


Well, there's a couple of surprises there. All that power didn't actually generate a lot of Runs Created by those 3, 4, 5 and 6 hitters. Who'd have thought the number three slot for the Nationals offered below-league-average production? And the leadoff slot was about average? What, you did? [Exeunt, red-faced.]