Thursday, April 08, 2010

On Imaginary Stats and Real Games

It didn’t take long for a flawed imaginary stat to impact a real game - negatively.

It happened in the fifth inning of the season opener, when Scott Baker was trying to preserve a 3-3 tie. He was at his pitch limit, he had two runners on base and a dangerous left-handed hitter was coming up. What’s more, he looked gassed, and relief was ready. The next move was obvious: pull him.

But he stayed in the game for three more pitches: a strike looking, a foul ball, and then a mistake. A fastball drifted towards the middle of the zone, Angles batter Hideki Matsui singled, and the lead was lost. Baker was pulled for reliever Jesse Crain. But the Twins would never make up that run.

Why did Baker pitch to Matsui? Because of the “Win” statistic, a sloppy short-hand stat which unfortunately is usually the first one mentioned when evaluating a starting pitcher. He was 11-7 last year with a 3.48 ERA. A starting pitcher earns a win when they pitch at least five innings and when they were the last guy on the mound when their team took the lead for good.

Baker wasn’t going to be pitching in the sixth inning in any case. But if Baker would have been able to get Matsui out, and then the Twins would have scored in the top of the sixth (and held that lead for the remainder of the game), Baker would have got the “W.” Manager Ron Gardenhire was giving his Opening Day starting pitcher a chance to get the win. Once Baker blew that chance, he was pulled.

If you pay attention, you will see similar decisions made by Gardenhire the rest of this year. There is no question that he would have rather had Crain facing Matsui in that situation. He’ll even manage differently late in the season when there is no room for error. We’ve seen as much the last two Septembers.

However, most managers feel that leaving Baker in to face that last batter has long-term positive effects that offset the short-term risk of an unfavorable matchup. For instance, it gives Baker an opportunity to stretch his skills a bit. But mostly, it shows Baker that Gardenhire is on his side.

Because imaginary statistics like wins and saves motivate ballplayers. Just like hitters want to claim they hit 20 home runs or hit .300 last year, pitchers want to be known as a 15-game winner or a guy who saved 30 games. There are plenty of reasons why: personal, communal, and financial. And for a manager of any team, let alone a team of kids pulling down a minimum of $400,000/year, having a motivational tool is really valuable.

It’s about this time that someone out there is saying that those ballplayers should have all the motivation they need because of the money and fame that goes with the position. After all, don’t we all work as hard as we can each and every day in the office?

The difference is that for most of us, our job doesn’t stretch out abilities to their absolute limit. Baseball players’ jobs do. You’re talking about a game played by millions and millions of people all over the world. And from that set, the top 350 who can throw really well compete daily against the top 350 that can hit really well. It’s different that way than most other jobs.

Let’s compare it to a job I’m personally familiar with and which is often used as a straw dog in these sort of arguments: a teacher. I was a good math teacher, but let’s say I could REALLY teach and even be one of the top 350 teachers in the world. In a normal teaching job, I look brilliant, the same way that these guys did in high school or college ball.

But I’ve gradually been given tougher and tougher assignment, and now on a daily basis, I face one of the top 350 toughest teaching assignments in the world. So I’m in a under funded school in a ghetto, teaching Algebra II to underprivileged or at-risk kids with enormously overfilled classrooms. What does my day look like?

I need to be prepared long before I get into the classroom. I need to handle distractions outside my work. I need to show up early, stay late, make calls to parents and take care of myself. And when I’m in those classrooms, I need to be fully engaged and on my game – taking risks, teaching in different ways, reaching kids who don’t trust anyone’s reach. Sometimes, when I have a good day or a bad day, I don’t even know why, so I adopt silly superstitions. That’s how stretched I am – beyond any sort of reason.

Even if I was paid really well, even if I loved my work, the chances of me slipping are high. Any motivation I find, no matter how silly it might be, that drives me to stretch a little further is welcome.

And it’s especially welcomed by my boss, who is judged by my results.

This is how wins and saves have become a management tool in today’s game. It was silly and counterproductive to leave Baker in to face Matsui on Monday if a team must win that game. But the team didn’t need to win that game. It does need a dedicated and motivated starting pitching staff, all of whom want those imaginary stats, and all of whom are watching to see if the manager is willing to stick his neck out to get them some.

Wins and saves are sloppy statistics that are questionable for player evaluation. But for managers looking for motivational techniques, they serve a deeper purpose. And that’s how flawed imaginary statistics can impact a real season - positively.

Think the teaching analogy was a bit of a stretch? Let me hear about it in the comments below. Or, better yet...

Let me hear about it in person while we dissect Baker's next start. That will be Saturday, and we'll be watching at the Twins Viewing Party at Major's in Blaine starting at 12:00 along with Sooz, the TwinsCentric guys and host of other writers. We'll also be sipping $2 pints, wolfing down two-for-one appetizers, raffling off goodies (including two row 6 Twins tickets) and watching the Twins dominate the hated White Sox. Last months we had 60+ people show up, so get there by 11:30 if you want a good seat.

- Wondering who is in AAA that might help the Twins this year? Check out Seth's Rochester Red Wings preview.

- If someone wants to teach me how to waltz, I'm game.

- Want to know why Baker struggled Monday? Parker points out that Baker lost confidence in his off-speed pitches and relied purely on his fastball towards the end of the game.


Anonymous said...

This is good stuff, thanks. The "he makes $___ million so he should have been able to ____" argument is really tiresome, and I like your perspective here.

It would still be nice to scrap or change the "Win" statistic, and have motivational tools that are a little more meaningful. Not that I have any suggestions, of course.

TT said...

I think you are mostly imagining this as the reason. If a manager takes out his starter every time they get in a jam in the 5th inning, they don't have a bullpen left for very long.

First, its not clear that "Baker wasn’t going to be pitching in the sixth inning in any case." He had thrown less than 90 pitches when Matsui came to bat and he may well have come out to start the 6th.

Second, I don't know when Crain started throwing, but I would doubt Gardy had him up because Aybar got an infield single. When Hunter came to the plate Baker had only 82 pitches. Which means Baker was in a jam only after Hunter walked. No time at all to get anyone up and throwing before Matsui batted.

I think stathead's sometimes imagine that the people who play the game are as obsessed with statistics as they are. But I think managers and players mostly worry about one statistic - the score.

John said...

For the record TT, Crain was ready. He was standing in the bullpen waiting.

Jim H said...

John, you are right that managers do manage sometimes to get their starters wins. I am not so sure that wins are anymore imaginary than some of the stats bloggers like, like UZR for example. TT is also right, you need your starters to go more than 5 innings (especially your "ace") or you won't have a bullpen.

I kind of like wins as a stat. Generally if a pitcher wins 15 games, he has had a good year. If you approach 20 wins you pretty much have to have had a great year. Nearly all stats are flawed in some manner or another, the trick is be aware of what those flaws are, and put the particular stat in context.

By the way, I was a math teacher too. Also a social studies teacher and a coach. I think having an appreciation for numbers helps me enjoy baseball even more.

BeefMaster said...

I am not so sure that wins are anymore imaginary than some of the stats bloggers like, like UZR for example.

It's true that lots of the advanced stats are imaginary, in the sense that they don't simply record objective events. The problem with wins is that in addition to being imaginary, it's not useful for evaluating players, but lots of people think it is.

The fact that a guy won 15 games could mean that he had a good season... or it could mean that he played on a team that scored a ton of runs.

Anonymous said...

90 pitches is a lot for April 5!

I would hope he would not have come out for the 6th innng, even if he did retire Matsui.

Brad said...

"W" was definitely the issue & I agree it makes sense during the typical "regular season game" to give it some weight in your managing. Shoot, from what I've read, middle relief types who wind up with a bunch of wins get needled, etc. for "poaching" the starters' wins.

For me, there's another reason to leave Baker out there: at some point, we're going to need him to get out of his own mess to save the bullpen arms for later. Even if whoever comes in to start the next inning, it saves wear & tear on the bullpen - much easier to start with the bases empty, etc. than to come into the starter's mess. Give guys the opportunity to succeed, blah blah blah.

For me, the biggest thing is "Scott, you gotta find a way to get rid of these hitters more quickly" - the way guys extend ABs against him with 2-strike foul balls is maddening.