Friday, December 05, 2003

Either/Or Meets Both:
Polyculture & the Playoffs  

In the last entry I was discussing the Scattershot magazine piece called Billy-Ball: Oakland's Uncanny Ability to Lose the Big Game. The essence of what tweaked my managerial interest was in the article's following paragraphs:

While Beane clearly has achieved consistently outsized returns in the regular season – due largely to factors unrelated to Beane’s different way of approaching his business – going 0-for-9 in games for playoff advancement is a remarkable feat of underachievement. Indeed, there is reason to wonder whether Beane’s approach to the sport is responsible for his team’s lack of success in the playoffs. [snip]

This approach to baseball also manifests itself in a general disdain for playing “small ball” – stealing bases, bunting runners over, and so forth. Lewis argues (and Rob Neyer before him, and Bill James well before them both) “the math works” in support of this position – squandering outs through small ball limits a team’s ability to score runs.

The author had many good ideas, but no numbers to support them. I don't view this weblog as my vehicle for original baseball research (an avocation of mine), but his contention triggered a need to run some numbers to confirm or deny his core thesis: that the approach one takes to achieve success in the regular season is the antithesis of what makes for success in the playoffs after. It's temptingly existential, teasingly Manichean. Since I've run a small sample (not ultimately significant, but indicative), I think it's probably wrong.

Thanks to the collaboration of Steve Nelson, the in-depth thinker of the Steve's Mariners and Other Stuff weblog, I can bring you some numbers that cast light on the thesis quoted previously. The conclusions are mine alone, so don't hold Nelson responsible.

I used the 2002 playoffs and World Series (02PWS) as my data source. I chose that year because it's the most recent set of play-by-play game sheets for playoffs on Retrosheet, and because that year, the Angels won the Series and that triggered a lot of gush about small ball tactics, the contention being the 2002 Angels were exemplars of why small ball tactics won Series.

There were 34 games in the 02PWS.

9 of those games were clearly won by Big Ball
3 of those games were clearly won by Little Ball approaches

I'm defining clear Big Ball as a game where the winning team scored more runs driven in by homers than the opposition scored in total. That's an unarguable Big Ball strategy. I'm defining Small Ball approaches solely as bunts (for hits or as sacrifices) and steals. I credited a team with small ball runs for any run scored in an inning subsequent to either a successful steal or bunt (but not counting runs that were driven in by homers or triples). This is pretty generous to Small Ball, because one could argue that the subsequent runs might have scored anyway. You can't know, because the infield defense changes position with a runner on a different base and a different number of outs, and the pitcher throws a different sequence.

In those 34 games, there were a total of 30 innings in which there was a Small Ball event and the offense scored a run or more.

If you remove the special cases, there were 21 innings in which there was a Small Ball event and the offense scored a run or more.

So in the 02PWS, there was fewer than a half-inning per game where a team garnered a small-ball run inning. In the special cases, the runs came in as the result of a hit that would likely have scored the runner without the small ball event. An example of a special case: Single, steal of second, triple.; if the runner hadn't stolen second, he would likely have scored from first on a triple, not assuredly, because the triple might not have happened. This is why I deliver both the total innings and the ones with the special cases removed.

It's pretty rare in the playoffs that a team executes a small ball tactic in an inning that results in a run (at the most, fewer than one time per game on mean average). But we remember these, because they're dramatic, like the Game 5 AL 9th inning victory for the Minnesota Twins over the Oakland As, where Christian Guzman stole second and scored the series winning run. Most people don't remember the Big Ball As stole a base in the first inning (that didn't ultimately result in a score). Or the dramatic Game 2 victory of the Angels over the Yankees where the winning run came in during the 8th, when a baserunner who stole a base and moved to third came in on a sacrifice fly.

The conclusion one has to draw from this single year is that small ball tactics don't outweigh big ball tactics in the post-season games. The argument isn't supported by this single year of data, not significant, but indicative.

However, it's not Either/Or Big Ball/Small Ball. As I just noted, the As used a small ball tactic. Here's an interesting counter-indication.

Total 02PWS Innings where the game-winning team used a small ball tactic (successful or not): 30
Total 02PWS Innings where the game-losing team used a small ball tactic (successful or not): 24
Total 02PWS games where the winning team tried more small ball tactic attempts than the loser: 12
Total 02PWS games where the losing team tried more small ball tactic attempts than the winner: 9

That is, teams that went on to win games are somewhat more likely to use a small ball tactic. This isn't the same as winning the games because of small ball (already indicated as less likely previously in this entry). What it means is that teams are somewhat more likely to win in the 02PWS if they execute small ball tactics.

The Point

The point is, there are things you do in management and in baseball that don't, in themselves, make success, but they are things you need to know how to do and execute properly if you want success. The Dot-Bomb era was filled with primitive little MBA functional-morons who believed that you could succeed as long as a stock price went up and you could deliver goods at a lower price because you could evade taxes. They forgot that factors like customer service, and product quality (nothing that would get you on the cover of Red Herring or eWeek) affect long-term customer choice. It's not Either/Or, it's some of both.

Big Ball doesn't guarantee success (though the odds of making it to the playoffs without is are not good). Small ball doesn't guarantee success (though the odds of winning playoff baseball go down if you don't know when to roll it out and have a little luck with it). Teams that have polycultural offenses have more different ways to win games than those with monocultural offenses.

Managing organizations requires a full toolbox, a "portfolio" of tools to face shifting conditions, antagonists, customers. Just taking things at the first level of apparent correlation is destructive. As blog-buddy Steve Nelson said to me:

The reviews I have seen show that when you go beyond simple association and investigate whether the assumed cause and effect relationship is actually in play in the underlying competition, you get a different picture.

And there is also a management lesson in that - good managers probe the simple assertions; bad managers take loose correlations at face value.


This page is powered by Blogger. Isn't yours?

free website counter