NLCS Preview: Mets vs. Cardinalsby Jeff Sackmann
October 11, 2006
Six years ago, the New York Mets and the St. Louis Cardinals met in the NLCS. That year, too, one team swept the short NLDS, the other won in four games. The Mets, of course, were on their way to the 2000 subway series and blew past the Cardinals in five games.
There isn't much that the 2000 matchup can tell us about this year's series: the storyline that has endured is the self-destruction of Rick Ankiel. Albert Pujols didn't arrive until 2001, and the only player from either team who will return from the previous NLCS is Jim Edmonds.
Like the 2000 Cardinals, who rode into the NLCS on Ankiel's heroics then couldn't rely on him in the series, these two playoff teams are notably different from the squads that got them here. Stalwarts like Pujols, Chris Carpenter, Carlos Delgado, and Jose Reyes are consistent features of each team, but with Cliff Floyd and Pedro Martinez missing from the Mets roster, Mark Mulder and Jason Isringhausen absent from the Cards, and a suddenly competent Jeff Weaver making the start for St. Louis in Game 1, the NLCS isn't just a crapshoot, it's a crapshoot between teams we don't know a whole lot about.
If we assume that the NLCS Mets are a 97-65 team and the NLCS Cardinals are an 83-78 squad, it's tough to make a case for St. Louis. Using Bill James's log5 method, a .599 club (the Mets) should beat a .516 club (the Cards) about 58.35% of the time. With a little more math, we can figure that, in a seven-game series, the Mets will advance 67.8% of the time. (I'm ignoring home-field advantage for two reasons: I don't like the adjustment very much, and it's more math than I can figure out how to do before the NLCS actually starts.)
I wouldn't have begun the previous paragraph with "if we assume" unless that assumption was questionable; I'm interested in seeing how we can reasonably alter those assumptions and, most importantly, what those changes do to each team's probability of advancing to the World Series.
The first way we can increase the accuracy of the series win probability is to use each team's Pythagorean record instead of their actual one. Because they are based on each team's actual runs scored and runs allowed, they may eliminate some of the luck inherent in their winning percentages. The Mets' Pythagorean record was 91-71, while the Cardinals' was 82-79, good for winning percentages of .562 and .509, respectively. Those records suggest that the Mets have a 61.5% chance of winning the series. This is still a commanding advantage for the New Yorkers, but a nice jump for St. Louis.
There are four major differences between the teams that played 162 games this year and the ones who will play this series. The first two are in the Mets rotation: John Maine and Oliver Perez will start instead of Pedro Martinez and Orlando Hernandez. Third, Endy Chavez is likely to be the Mets left fielder instead of Cliff Floyd. Last, Jeff Weaver is pitching like a major leaguer again.
There are myriad other adjustments we could make: Ronnie Belliard might improve the Cards' second base situation from the rotation of stopgaps that started the season; Scott Rolen might be weaker than usual as his shoulder recovers; and both bullpens have changed a great deal, especially adjusting to replace Jason Isringhausen and Duaner Sanchez. But because it would be difficult to quantify or foolish to predict the results of those changes over a seven-game series, we'll ignore them.
Further, we can safely disregard the difference between Floyd and Chavez. Chavez was actually the better player over the regular season: his 14 Win Shares exceeded Floyd's 9, and he accumulated them with less playing time. A comparison of their careers suggests the opposite, so we'll say the two considerations cancel each other out and call it a wash. Similarly, Pedro's weak campaign, measured by Win Shares, was roughly equivalent to John Maine's breakout season, allowing us to scratch that comparison. That leaves only two adjustments to be made.
The first is easy. In partial seasons, El Duque was worth 6 Win Shares to the Mets while Perez didn't contribute any. Accordingly, let's call the NLCS Mets an 89-73 team to reflect their 91 Pythagorean wins minus Hernandez's contribution.
The second is trickier. I'd imagine many Mets fans, as well as a handful of statheads less fanciful than I, object to the assumption that the Cardinals are a better team than their record indicates simply because Weaver has pitched well in his last four starts. However, it's an assumption I'm comfortable making. Weaver's career numbers are so much closer to his last month's performance than they are to his 2006 season line that it seems more likely he'll be a Jeff Suppan-esque innings eater than a Perez-esque disaster waiting to happen.
Essentially, we can say that the good Jeff Weaver is replacing the two-headed (and one-beer-gutted) monster that was Sidney Ponson and the mediocre Weaver. Ponson and Weaver combined for 6 Win Shares for the Cards this year, while Weaver contributed 12 and 13 Win Shares to his '04 and '05 teams, respectively. Let's say Weaver enters the NLCS as a 12 WS pitcher, adding 6 WS, or two wins, to the Cardinals' Pythagorean record, adjusting it to 84-77.
These adjustments, combined with the shift from using actual records to Pythagorean ones, makes this series out to be much closer than it originally appeared. Using adjusted winning percentages of .549 for the Mets and .522 for the Cardinals, the log5 method gives the Mets only a 52.8% chance of winning each game, and a 56.1% probability of taking the series.
On the basis of these calculations, should you be rushing out to put your money on the Cardinals? I hope not, if only because then you'd be put in the awkward position of rooting for Tony LaRussa and counting on Jeff Weaver to succeed. But if you didn't already view this short series as a crapshoot, there's all the more reason to do so now.
References and Resources
For more on the log5 method, read Tom Tippett's great piece on the topic. Tippett uses log5 and his own calculations to predict the results of postseason series; my method is similar to his, except that, unlike Tippett, I ignored home-field advantage.
If you're looking for a more traditional series preview, MLB.com offers a pair of position-by-position breakdowns. Here's the rotation analysis, and here's everything else. You might also be interested in the excellent NLDS previews offered here at THT, as much of what was true about these teams a week ago is still accurate. For the Cardinals, here's Bryan Tsao's and Vinay Kumar's preview; for the Mets, here's what Dave Studeman and Rich Lederer had to say.
Jeff Sackmann is the creator of MinorLeagueSplits.com. With Kent Bonham, he founded CollegeSplits.com. Jeff and Kent blog about college baseball and the draft, and you can follow them on Twitter for bite-sized snacks of minor league and college stats. Jeff also has an email address.
<< Return to Article