I’ve explained sabermetric concepts on this blog before. Some, such as leverage and WAR, can take a while to explain for the uninitiated.

Run expectancy, despite its fancy-sounding name, is not one of those things.

Run expectancy, as a concept, is one of the easiest to understand: given situation x, you may expect y runs to score. For example, with a man on first base and two outs in the inning, you can expect y numbers of runs to score.

The mathematically-inclined stats geeks (ie, not me), have done the actual, hard-core math, and found out the exact numbers.

They’ve done funky database things, and now you can go here, and then set your criteria, so you can figure out the run expectancy for whatever parameters you want to set.*

So if we leave it at the default for 2010, what the columns read, left to right are thus:

1) Year
2) How many men on base. “1″ means runner on first, “2″, runner on second, etc. A “0″ means no runner on that base.
3) Run expectancy with 0 outs
4) Run expectancy with 1 out
5) Run expectancy with 2 out.

The higher number, the more likely you’ll score a run.

As you would expect, with the bases loaded and no one out, you have the highest chance of scoring a run–in this case, 2.47 runs. The lowest expectancy comes with two outs and no one on base.

Really, you don’t need sabermetrics to understand that the more runners and fewer outs you have on base, the more likely you are to score runs. It’s really a very basic concept to understand.

Now, in terms of tonight’s game, we have the following set of data for the ninth inning:

1) With a runner on second, and no one out, the run expectancy is 1.11.
2) With a runner on third, and one out, the run expectancy drops to 0.927.

This isn’t exactly dropping off a cliff here, but when we’re dealing with numbers in these ranges, it is significant.

Now, if you’ll excuse me, we’re going to take a brief detour, but I promise I’ll get back to the point, quickly.

In baseball, there is no clock in the sense that we think of one–nothing ticking down, stopwatch-ing or otherwise counting to 0:00.

What you have, instead, are outs. These are precious commodities, things you guard with your life, like porn in a jail cell.

The entire premise of Moneyball, sabermetrics, and modern baseball statistical geekery, as relates to a team’s offense, is predicated on the notion that you never, ever, ever, ever give away an out.

So, then, understanding this, let’s get back to the issue for discussion here: bunting.

When Francisco Cervelli bunted with Canó on second base and no one out, he changed the run expectancy situation from #1 above to #2–that is, lessened the number of runs the Yankees could be expected to score.

Now, there is an argument–and a legitimate one–that you sacrifice the out to have Canó on third with less than two outs, because Thames, Miranda and Winn were on deck, who are, well, not the best hitters in the Yankee lineup. With Canó at third, he can score on a wild pitch, on a long fly ball or fortuitously placed ground out, lessening the necessity of Thames, Miranda or Winn to actually get a hit.

The problems, however are these:

1) Cervelli, thus far, has shown himself to be a decent, if lucky, hitter. Miranda and Winn are much more likely to make outs against Papelbon, if their current season statistics count for anything.

2) If Cervelli reaches, and there is still no one out (and Canó hasn’t scored, so a walk or an error or some such), then only one of Thames/Miranda/Winn would need to actually reach base to ensure, barring a double play, that Derek Jeter gets an at bat. I know I’ve been hard on Jeter as of late, but the top of the line up is what you want batting in the bottom of the ninth, not the bottom of the lineup.

3) The current pitcher, Jonathan Papelbon, had been able to successfully retire two of the last eight batters he faced. Which, you know, says that he wasn’t performing well and raises the likelihood Cervelli may have actually done something with the bat.

4) Giving up the out decreases the run expectancy, as stated above.

In The Book, Tom Tango and his co-authors argue that one should bunt more with low-OBP hitters on deck. In that case, Thames/Miranda/Winn certainly fill that category, as the bottom of the line up usually does, but it’s relative.

Thames/Miranda/Winn, whatever their faults, are not Peña/anyoneontheMarinersPiratesorAstros/pitcher. All three of the Yankees, not including last night’s game, have OBPs of .300 or over. No, that might not be sparkly, but it’s far shinier than what you’d expect from an NL ninth hitter–which is probably what Mr. Tango, et. al, had in mind when they wrote that passage.

So basically, if we want to keep it simple, we can sum up in this: by bunting, Cervelli decreased the number of runs the Yankees could expect to score in that situation, given the data thus far assembled for the 2010 season.

Outs are shiny, precious nuggets. They are Guttenberg Bibles, Tolkien manuscrips, Honus Wagner rookie cards, Mariner runs.

They are near and dear and the most valuable asset a baseball team has to work with.

They should be parted with willingly under no circumstance, while a baseball team is batting.

Francisco Cervelli should not have bunted.

*there is another table to look at here, and the entire article is worth a read, but their data comes from years past as opposed to this year.