When 'pens succeed, playoff wins follow

I have five bold postseason predictions:

1. The Philadelphia Phillies will not win the World Series.
2. The Los Angeles Angels of Anaheim will not win the World Series.
4. The Colorado Rockies will not win the World Series.
3. The Detroit Tigers (if they make the playoffs) will not win the World Series.
3. The Minnesota twins (if they make the playoffs) have a shot to win the World Series.

Here's why: Since 1969 (the start of the division-play era) only three teams have won a World Series with their closer owning an ERA over 3.00. That's three out of 39 World Series champions. Your odds aren't good if your ninth-inning guy isn't great.

The Phillies will use Ryan Madson (likely) or Brad Lidge (let's hope not) as their closer. Madson has a 3.26 ERA, which is four runs per game better than Lidge's 7.21, but still above our threshold.

With 48 saves, Brian Fuentes has proved you can rack up big numbers in that category with lackluster numbers everywhere else: 3.93 ERA, 1.40 WHIP, 46-24 strikeout-to-walk ratio, all while averaging less than an inning per outing.

The Tigers turn to Fernando Rodney and his 4.33 ERA, 1.43 WHIP and eight home runs allowed. Yes, Rodney has saved 37 games in 38 opportunities. You know many saves he has where he entered with a one-run lead? Nine. You know what his ERA is while pitching with zero or one day of rest? 5.54. You want him pitching to Mark Teixeira, Alex Rodriguez and Hideki Matsui with a one-run lead at Yankee Stadium?

Huston Street has a 3.06 ERA, although his other indicators are pretty solid, especially a 1.71 ERA on the road and only two blown saves. Still, he's battled triceps tendinitis and his ERA is over 3.00, in part because he's allowed seven home runs in only 60 innings.

The Twins, on the other hand, have Joe Nathan and his 2.15 ERA, .167 average allowed and 87 strikeouts in 67 innings.

Maybe you think this analysis is too simple; after all, playoff baseball involves clutch hitting and starting pitchers with guts and flashy leather and sometimes a little fan interference. All true, of course; but the chart doesn't lie:

Still not convinced? After all, doesn't logic dictate that most playoff teams have a good closer? Isn't this kind of like saying nobody wins with a first baseman who doesn't hit for power or a shortstop who leads the league in errors?

Maybe, but here's another statistic to activate your brain waves: Of the 56 teams to make the playoffs since the Division Series began in 1995, 30 had a closer with an ERA over 3.00 (and two more had no set closer entering the playoffs).

Only five of those 32 teams reached the World Series: the 1996 Braves (Mark Wohlers, 3.03); the 1997 Marlins (Robb Nen, 3.89); the 2006 Cardinals (Adam Wainwright, 3.12); the 2006 Tigers (Todd Jones, 3.94); and the 2008 Rays (no set closer).

So, while more than 50 percent of all playoff teams since 1995 had a closer with an ERA over 3.00, only 15 percent of them made the World Series.

Still not convinced that this 3.00 threshold is important?

During Division Series play, the team with the better regular-season record actually has a losing record, at 26-28. The team that scored more runs per game in the regular season has gone just 22-32. Meanwhile, the team with the closer who had a better ERA has gone 29-24 and the team with the better cumulative bullpen ERA has gone 33-23. Those series records are even slightly better than teams that had the better ERA from their starters.

Here are the complete breakdowns, listing series wins by Division Series, League Championship Series and the combined total. (World Series results aren't included due to the statistical differences between the NL and AL.)

Team with better record
Division: 26 W, 28 L (two series had teams with the same record)
LCS: 16 W, 11 L (one with same record)
Total: 42 W, 39 L

Team with the closer with the better ERA
Division: 29 W, 24 L (two with same ERA, one team had no set closer)
LCS: 17 W, 11 L
Total: 46 W, 35L

Team with the better bullpen ERA
Division: 33 W, 23 L
LCS: 18 W, 10 L
Total: 51 W, 33 L

Team with better starters' ERA
Division: 30 W, 25 L (one with same ERA)
LCS: 14 W, 13 L (one with same ERA)
Total: 44 W, 38L

Team with more runs scored per game
Division: 22 W, 32 L (two teams with same)
LCS: 14 W, 14 L
Total: 36 W, 46 L

A few caveats: The more sabermetrically inclined will protest at the simplicity of the study, that adjustments should be made for home park or that cause and effect hasn't been proved. Also, it's possible that over time these comparisons will even out (maybe the higher-scoring teams will win their next 10 series, for example).

But 14 years of results indicate -- arguably -- that a good bullpen is a better predictor of playoff success than a team's win-loss record or its offense.

What's perhaps even more surprising is that the bullpen numbers have been a slightly better predictor than a team's starting rotation.

If you think about it, however, this isn't so surprising: Starting pitchers throw fewer innings than they used to (since 1995, 33.1 percent of all playoff starts have been seven-plus innings, compared with 44.7 percent from 1980 to 1993) and managers often manage to avoid criticism -- call it the Grady Little Effect. They'd rather pull a starter and have the bullpen lose a game than keep the starter in and have him blow a lead (and get murdered in the media for doing so). After all, it wasn't Pedro Martinez who drew the wrath after Boston lost Game 7 of the 2003 ALCS.

So managers turn tight games over to their bullpens -- from 1995 to 2003, starters went seven-plus innings in 38 percent of all postseason games (225 out of 592 games); since 2004 (post-Grady/Pedro), that percentage has fallen to 23.7. Relievers now factor into the decision in about a third of all postseason games. You need a good middle guy (or three) as much as you need a power-hitting cleanup man these days.

Since most teams usually employ their best reliever at closer, if your closer has an ERA over 3.00, it may suggest the quality of your middle relievers is lacking. Or named "Farnsworth." Bottom line: A 3.34 ERA for a starter is excellent; it's not a good ERA for a closer.

Which is why the Phillies shouldn't be nervous about only the ninth inning. Because even if Ryan Madson can do the job in the ninth, somebody has to pitch the seventh and eighth.

It should be noted -- for the sake of Phillies, Angels, Tigers and Rockies fans -- that three teams in the above chart did go all the way.

A closer look at those squads:

2006 Cardinals: Jason Isringhausen was the team's closer until a hip injury ended his season in early September. This was actually a positive development for the Cardinals, as Isringhausen had gone 4-8 with a 3.55 ERA, 10 home runs allowed in 58 1/3 innings and a terrible 52-38 strikeout-to-walk ratio. In other words, he wasn't exactly Dennis Eckersley, circa 1989. Adam Wainwright was a rookie reliever with only three saves, but he was better than Isringhausen and was lights-out in the postseason, pitching 9 2/3 scoreless innings.

1997 Marlins: Robb Nen had 35 saves and nine wins, but the '97 season was one of his "odd" years -- if you check his record, Nen alternated dominating seasons (1.95, 1.52, 1.50 and 2.20 ERAs in 1996, 1998, 2000 and 2002) with less-stellar seasons. He gave up four runs in a 14-11 victory in Game 3 of the World Series, but was solid otherwise, picking up two saves in both the NLCS and World Series.

1987 Twins: Veteran Jeff Reardon recorded 31 saves but was prone to the long ball (14 home runs), which resulted in a 4.48 ERA. He struggled in the ALCS victory over Detroit (but got a win and two saves) but tossed 4 2/3 scoreless innings in the World Series, including the ninth inning of Game 7 while protecting a 4-2 lead.

This may or may not be a coincidence, but those three teams also happen to be three of the worst World Series champions ever. The Cardinals won only 83 games and the Twins won 85 in the regular season, the two lowest winning percentages for a World Series winner. The Marlins were a wild-card team with 92 wins.

(The 2001 Diamondbacks managed to win the World Series despite the implosion of closer Byung-Hyun Kim. Of course, they won only when the Yankees blew their own ninth-inning lead in Game 7.)

One final note: This is not to suggest that a team should go out and spend $12 million on a closer. You can survive the regular season with a mediocre closer (or, in the case of the Phillies, about 20 stages worse than mediocre) and keep in mind that closers can come from nowhere: Wainwright was a rookie; the White Sox had their own rookie closer the year before in Bobby Jenks, a guy they had picked up off the scrap heap from the Angels; the Yankees lost John Wetteland after 1996 and replaced him with Mariano Rivera.

Of course, the other option is just to ride your starting pitcher. But considering there's been only one complete game over the past three postseasons combined, we don't expect Charlie Manuel to employ that strategy.

David Schoenfield is a senior editor for ESPN.com.