Wednesday, January 13, 2010

Play Makers

In the age of the baseball statistical explosion, one thing I continue to appreciate about basketball and football is the emphasis that still exists on the athletes themselves, the human element of the game. There are plenty of numbers that are helpful in both sports, but the nuances of athletic ability remain paramount. Players can have similar numbers, but strong distinctions will be made between those that make plays to help their team win games and those that don't. These distinctions are derived from what we see while watching games, not what we read in the box score or on the stat sheet.

To this end, DV is right when he talks about sabermetrics having gone to far. What was once merely an immensely helpful set of statistics have now become rules in many instances. This is obvious in many different places, and one example is Pythagorean winning percentage. Many will point to this winning percentage rather than actual winning percentage as the true barometer of how good a team is. Logically this just doesn't seem right. The Yankees, for example, outperformed their Pythag. winning percentage every single year from 96-03, and did so by an average of something like 6 wins per season. Are you going to tell me that it was mere coincidence that a team on what is now considered one of the best runs in baseball history just happened to be better than their Pythag., or "lucky", every single year? I don't think so. There we were again in 2009, with the Yankees again outperformed their Pythag., this time by 7 wins. They also were the best team in baseball and won the World Series. Again, coincidence, or do good teams just find ways to win?

Sure enough, it is now being proven by Bill James himself, one of the founders of sabermetrics, that there are team qualities that defy Pythag. James theorizes that the Angels, for example, who consistently outperform their Pythag., are doing so because they consistently run the bases so tremendously. Here we are, one of the great minds who brought about the statistical revolution pointing to traditional athletic ability to explain how statistic doesn't exactly work as well as we thought.

This general theme of the game not being played on paper is a potential problem for the 2010 Yankees in my mind. Key word here is potential, as I obviously have no idea.

The Yankees have replaced Hideki Matsui, Johnny Damon, and Melky Cabrera with Curtis Granderson, Nick Johnson, and Brett Gardner. On paper, from a production over 162 standpoint, that looks great. But in my opinion 162 production is not the only thing.

In the biggest spots, late in games, I knew I was getting Matsui's and Damon's best at bats. Just look at the World Series. Matsui's two homers off Pedro in critical games. Damon's at bat against Lidge. Matsui's 6 RBI Game 6. Damon's go-ahead double in Game 3, where the Yankees never looked back in from there in the Series. Damon's double in the first inning of Game 4. These were guys that stepped up and made plays to help their team win games. Melky Cabrera, who isn't the player that Matsui or Damon are, was second in the majors in walk-off hits last year, and has had a penchant for the clutch since he first saw extended playing time in 2006. His teammates and coaches described him as someone who made things happen. These are things that are not going to show up over 162 games. But they matter in terms of winning baseball games, especially in the playoffs.

Can Curtis Granderson, Nick Johnson, and Brett Gardner replace this? That remains a very legitimate question. I have no idea, and none of us do. But if I'm asking all of you Boston fans right now, 9th inning, runners on first and second, two out, tie game, who scares you more, Hideki Matsui or Nick Johnson? Is it even close? I like Nick Johnson as a player, a lot. The guy works counts and gets on base. Very useful piece. But that stuff doesn't determine who gets it done when it matters. Damon vs. Granderson? Probably a little closer than Matsui vs. Johnson, but I would still think Damon's peskiness scares you more, especially considering Granderson's strikeouts. Melky vs. Gardner is not as substantial, especially since Granderson has also shown himself to be a playmaker.

Stat heads will tell me this stuff doesn't matter. I really believe that it does. Even if it doesn't always pan out, we all know how draining it is as fans to face guys that scare you. Imagine what it's like for opposing pitchers? It also goes back something all of us who played or have been around basketball know to be true. 16 points and 8 boards on average is not the same for all players outside of the numbers themselves. Some guys are going to get 20 and 14 against weaker teams, and then disappear with 8 and 4 when it matters against the best teams. Others are going to be someone you can rely on for 16 and 8 no matter who you are playing. The same numbers on average, but which one would you rather have? In baseball, we only talk about this stuff when we want to discredit someone as a "stat padder". But there is a story to be told within every single player's numbers. Some guys are play makers, that you can count on when it matters most. Others are just guys who put together some good numbers. I'm not saying they can't do it - and I certainly hope they do - but the Yankees as a team have a long way to go to replace the late inning at bats they are going to be losing from Matsui, and possibly Damon as well.


Anonymous said...


Those pythagorean numbers you mentioned for the Yankee teams from the Torre era are exactly why I think Joe Torre was such a great manager. At some point a team either outperforms or underperforms. Those Yankee teams outperformed. The numbers suggested the Yankees were very good, but those teams were actually great. And to me, the difference is the intangibles. Prevailing thought is that chemistry doesn't matter in baseball. But I'm sure that it does, just not to the extent that it does in basketball or football. But ultimately, I feel like a team that should win 91 (according to pythagorean stats) games but wins 99 games probably owes a debt of gratitude to its manager for playing the right hunches, mixing things up at the right time, and not murdering them with poor line-up/pitching moves.

--the Gunn

Patrick said...

i think part of it definitely has to do with the manager, no question, and this is definitely true in the case of torre. and i also think that good managment is part of having a good team, and that any extra wins a manager can get his team shouldn't be viewed as somehow less legitimate, which is the picture some try to paint when they point to wins above pythag. win pct. but i don't think this is the whole thing. i think, just like in any sport, good teams find ways to win games, maybe even more games than their on paper talent would indicate. we are seeing that good teams are consistently outperforming their pythag. i don't think we can chalk that up to luck at all. i think good management is a start. but i think there are other things, like finding ways to win one run games, having late inning relievers like mariano rivera (who some are now theorizing is the reason the yankees always almost outperform their pythag., which makes sense), doing the little things well all the time, etc. again, i'm not saying pythag. is not incredibly useful. it, like most stats, are really a great guide. it gives us a general idea of where a team should be. but it isn't a rule. outperforming your pythag. doesn't mean you overperformed, and underperforming it doesn't mean you underperformed. there are things that happen in athletic performance that dictate these occurrences. it doesn't just happen by chance. all teams who outperform their pythag. aren't overperformers, somoe are teams that do other things well to perform at the level they ultimately arrived at the end of the season. in some cases. in other cases, maybe teams just played over/under their heads. we have to look beyond these numbers to arrive at our own rules and conclusions, not use these numbers as the rule and the conclusions.

TimC said...

To me, the Pythag argument is a problem of nonlinearity. The formula, I believe, measures all contributions as having an equal impact on a projected W-L record. But we all know that in baseball, a home run hit in the ninth inning of a 10-0 game has ZERO impact on a W-L record whereas a home run in a 1-1 game has a massive effect. In this case, an equal amount of effort and production leads to two very different results (hence, nonlinear). The statistic needs to be adjusted to account for this.

In my mind, Theo does not get this. He has consistently made moves aimed at bolstering 162-production at the cost of players who can swing games late. (Which is interesting as in '04 his moves, famously, seemed aimed at doing the exact opposite). Going by Pat's comments, it looks like the Yankees, led by stat-minded Cashman, are going the same way. Tough break.

Patrick said...

absolutely timc. and that type of thinking applies to a lot of stats, if not all stats, in my opinion. that is a lot of what i'm saying about matsui/damon vs. granderson/johnson here. sure, they're overall stats/production may be similar. but we know that matsui and damon are game changers, and can "swing" a game late as you said. i love that term "swing", because that is really what it is. your changing a game, a game changer, a play maker. this is what i'm talking about here, it's all very similar. do we know that about granderson/johnson? not in a yankees uniform. they may be, they may not. maybe one and not the other. so while they may replace the production pretty adequately over 162, 162 production is not the whole story. there are other things that happen in games that matter on a more micro level, especially in the playoffs. if the yankees have in fact lost something in this regard, it's a big loss. we'll have to wait and see what granderson and johnson can do. either way, i don't like the thinking that it is just about replacing on-paper, macro production. there are other intangibles that matter, and they go beyond the numbers. you have to look at the athlete and their abilities.

Ross Kaplan said...

I was never a huge fan of the whole stats movement partly because I never really understood what UZR, RAZW, OBS and a thousand other acronyms really meant. You can quantify just about any stat you like, but you can't solely depend on formulas and statistics, sometimes you also have to go with your gut.

That was another lesson I learned from "The Torre Years." In 03-04 while Theo, Bill James and their team of math whizzes were crunching numbers to acquire key players, the Yankees were still depending on old school scouts who went with their gut and used the usual, ERA, batting avg, RBI's for player evaluation.

Of course these new age stats are useful, but to agree with Pat you can't depend on them alone. Some evaluation needs to come from observations and gut feelings. Things like Damon's 2 stolen bases in one play and Matsui's dominance against Pedro just are not quantifiable.