Lastings Milledge: A Roll of the Dice and a Baseball Philosophy

Tom AuSenior Analyst IIAugust 3, 2009

ATLANTA - APRIL 11: Outfielder Lastings Milledge #85 of the Washington Nationals bats against the Atlanta Braves on April 11, 2009 at Turner Field in Atlanta, Georgia.  (Photo by Al Messerschmidt/Getty Images)

Which would you rather have, five dollars, or the chance to roll a single die and earn $1 times the number on the die?

The "fiver" is Nyjer Morgan with his five-win (over replacement) trajectory, and the die roll is Lastings Milledge.

That's because Milledge could be worth anywhere from one win (Brandon Moss) to six wins (Alex Rodriguez) over replacement. For now, we'll assume that the die is fair, and that these results, plus two, three, four and five, are equally likely.

If you are the Pittsburgh Pirates, your answer is, "Throw in an extra dollar, and I'll take the bet." Sabermetrically, at least, the trade of Sean Burnett for Joel Hanrahan garnered an extra (expected) win.

The thing is, if Milledge is worth one to six equally weighted extra wins, the average is three-and-a-half. Throw in the reliever trade, and we're getting four-and-a-half, less than the five we're giving up.

A more interesting wrinkle occurred when Andrew Kaufman asked in the comments section of my previous piece, "Is Nyjer Morgan actually worth five wins?"

The short answer was, "Yes, but."

Video Play Button
Videos you might like

His BABIP (at the time of trade) was unusually high, .346, and probably seen as unsustainable, even though it's higher today. And his defensive value is also far "off the chart" this year.

Assume more "normal" values for the two metrics, and Morgan is more like a three-and-a-half win player. That may be why Pirates' management wouldn't do Morgan-Milledge "straight up" but would do the deal with the added incentive of Burnett-Hanrahan.

But, one might have said, "Nyjer Morgan is ALREADY producing five extra wins, close to the top of Milledge's POTENTIAL. I'll take that to the bank any day."

That's what yours truly would say.

We know that Morgan's average means that he makes good contact. And his BABIP has been high his whole career, meaning it might actually be sustainable. If we want that sixth win, we could work on his swing, and maybe find him some power.

And we'd also do something about the fact that he is caught stealing too many times to make his base speed worthwhile. Certainly all of the above represents a better chance than gambling on Milledge. Why would Pirates' management think differently?

The answer may be a belief in trading algorithms driven by sabermetrics, and other "quant" measures. In this regard, Pirates' management is not so different from "algorithmic" hedge fund traders  (a community I used to be part of).

It was Andrew Kaufman again, who clued me that this might be the rationale for what management is doing. Do a bunch of these trades, and pick up one (sabermetric) win for every two players traded,

Trade 20 people, and you might pick up ten potential wins for the Pirates in the process. Combine that with recent lucky "draws"  of new talent (Andrew McCutchen, Garrett Jones), and development of new players, and we have a potentially winning team.

Analogous versions of this formula has made many Wall Street traders filthy rich. It has also accounted for some of the spectacular failures like Long Term Capital Management a decade ago, and the collapse of most Wall Street firms last year. The main problem is that in doing "quaint" trades mechanically, you take your eyes off key underlying assumptions.

Here's an illustrative, and not necessarily far fetched example that shows why it might not a good idea to trade veterans for rookies.

For instance, the Pirates have about the fifth best home game record in the National League. That is the mark of a fundamentally strong team, and could be good enough to contend for at least a wild card slot, if their road game record were comparable.

But it's not.

The team is the third worst in the National League in this regard,with only the Washington Nationals and the San Diego Padres being worse. In 2007, on the other hand, the home game record stank as much the road game record, meaning that the Pirates were weaker then.

That's why it might be have been a good idea to keep veterans like Freddy Sanchez and Jack Wilson around, even though Wilson is only "league average," while Sanchez is one to two games above that. Such people might have pulled us closer to "league average" on the road, while their mediocrity doesn't cost so much at home with our newer, better players.

So if you want to do trades, do so for specific reasons.

The Pirates' Achilles heel, recently, has been pitching. In 2008, for example, pitchers cost the team something like 12 games below that of the league average staff. That accounts for most of the 14 wins below .500 (67 total).

The two losses or so that could be attributed to the lineup, stemmed from trading Jason Bay and (dumping Jose Bautista) for Andy LaRoche and Brandon Moss.  (Andy LaRoche was a downgrade from Bautista last year.)

Trading Xavier Nady (and Damasco Marte) for starting pitching brought us two pitchers (Ross Ohlendorf and Jeff Karstens), that were about three games better than Tom Gorzelanny, and the "tag team" of Matt Morris, while providing potential long-term replacements (Jose Tabata and Dan McCutheon) for themselves.

And if you replace Nady with Nyjer Morgan, you pick up an additional one or two wins for a total gain of four or five wins. (Management apparently didn't realize this in trading for, then using, Brandon Moss.)

In hindsight, trading Nate McLouth wasn't such a bad idea, either. Andrew McCutchen is an adequate replacement for him, so the Pirates didn't lose anything on that switch. But McLouth brought Charlie Morton, a three win upgrade over the replacement starter called Jeff Karstens, (plus two other prospects).

Trading expensive Adam LaRoche for what we can get would not be a bad idea, if you can put Garrett Jones, and his two win upgrade (over LaRoche) at first. And given that we need Jones in the outfield, trading LaRoche back to Atlanta for Casey Kotchman would have been a better deal than the nondescript rookies we got for him.

But you should refrain from doing "spec trades" with the likes of Nyger Morgan, Freddy Sanchez, Jack Wilson, Tom Gorzelanny, and Ian Snell. For instance, Snell is two wins below what he was worth in 2007. In 2008, Gorzelanny was four wins worse than two years ago, and he appears to have gained them back, based on his recent minor league record. Get those six wins back (by taking them back to 2007), and your rotation is an asset, instead of a liability.

And if Lastings Milledge was available, why not at least offer the Washington Nationals your "head case," Ian Snell, in a like for like trade? You could have thrown in Brandon Moss or Steve Pearce to balance the trade.

Why not find out more about what Morgan, Gorzelanny and Snell can do before trading them. And exercise your club options on Wilson and Sanchez for 2010 to buy the necessary time.

One of my bosses had some very sage advice for us stock traders: "You guys are spending too much time chasing 'new' stocks. Why don't you spend more time studying what you own instead of starting over with a new bunch?"