October 13, 1999

Prospectus Roundtable

Mets vs. Braves

by Baseball Prospectus

Before The Game

Keith Law: Braves in FOUR? Oh, we're going to get hate mail...

Joe Sheehan: I can see the Mets winning a game...I just can't fathom which one, or how.

Rany Jazayerli: The chances for a Braves' sweep are about as high as that of a career revival for anyone who has appeared in a 1-800-COLLECT commercial.

Greg Spira: Why on Earth wouldn't they win a game? Play the series out in Strat-O-Matic 1,000 times, and the average outcome will probably be Braves in seven games. Predicting sweeps of any kind with two relatively close matched teams goes against logic.

JS: But they're not that closely matched. They were separated by six games in the regular season, and that's the regular season. In the postseason, it's all about the front-line talent, and the Braves have that over the Mets. The Mets are a deeper team.

GS: I'm not sure where that difference is, except in the starting pitching. The Mets clearly have the better everyday players (as a result of the Braves' injuries), offensively and defensively, and probably a very slightly better bullpen overall.

Yes, the Braves' front four starters are significantly better than the Mets' starters, but the Mets' starters are all solid and reasonable; it's not like they're the Rangers. And, despite their performance during the year against the Braves, the Mets' main offensive weakness has been against left-handed pitchers.

KL: True, but they're hardly the Braves. Al Leiter is inconsistent, Kenny Rogers is gimpy, Rick Reed is taterrific...oh, wait, that three-hitter against a Giles/Kendall/Sprague-less Pirates changes everything, right Mr. McCarver?...Masato Yoshii is flammable...

Meanwhile, the Braves had two of the top five--arguably top three--starters in the National League this year, neither named Greg Maddux or Tom Glavine.

My objection is that it takes very little for a four-game sweep to go awry. If one Met starter turns this into a battle of bullpens, it's basically a toss-up, depending on which relievers appear in the game and how tired they are. Braves in five seems a lot more likely to me. If the point is just that this is a bigass mismatch, though, Braves in four makes that case better.

The one other thing with which I disagreed in that writeup was the mild dig at Kevin McGlinchy. He was decidedly mediocre in the first half, but in the second half, he was quite good: 26 innings, 28 baserunners, 23/5 strikeout-to-walk ratio. Maddux, in fact, was much worse in the second half, with 129 baserunners in 92 1/3 innings and a 62/16 ratio. I know it's barely comparable, and second-half stats don't mean a ton, but I just thought that the comment was undeserved.

RJ: 1) The Braves are the better team.

2) The Braves' talent distribution is better suited for postseason play: the bench is the worst part of the team.

3) That still doesn't mean predicting a sweep is wise.

If we assume the Braves have a 60% chance of winning each game, the chances for a sweep are 13%. There is a way to predict the winning percentage of two teams, given their winning percentages against the league; Bill James wrote about this in the early Abstracts. I think, just off the top of my head, that a six-game difference in the standings wouldn't translate into more than about a .550 winning percentage, but we'll call it 60% because of the Braves' frontline talent.

So they have a 13% chance of a sweep. Does that mean you should predict one? Let's put it this way: if I were a betting man, I would jump at the opportunity to bet against the sweep.

And I think everyone is overlooking just how significant the Braves' OBP problems are. You've got Gerald Williams, Bret Boone, Jose Hernandez, Eddie Perez and the pitcher - that's five spots in the lineup without even a .340 OBP. The Mets, 1 through 7, have no one worse than a .361 OBP. That's an enormous difference, and by itself reason enough to think the Mets won't go away in four games, or even in five.

Michael Wolverton: This got me curious to check out the probabilities. In case anyone else is interested:

Braves in 4: 13%
Braves in 5: 21%
Braves in 6: 21%
Braves in 7: 17%
Mets in 7:   11%
Mets in 6:    9%
Mets in 5:    6%
Mets in 4:    3%

That's assuming Braves beat Mets 60% of the time in a game, all games are independent, etc. All unrealistic assumptions, of course.

It's interesting that, if you accept the 60% number, "Braves in 5" is as good a prediction as any other. I would have thought that the safe money is always on a six- or seven-game series, except for a drastic overmatch situation (e.g., a Texas team vs. anybody in the postseason).

In order for a sweep to be the best prediction probability-wise, the Braves would have to beat the Mets at least 75% of the time in a game.

Chris Kahrl: It seems people have conceded the Mets' defensive superiority far too easily. Whatever the Mets' perceived advantages in the infield--and I think they're overstated--aren't the Braves' advantages in the outfield just as significant? Is the difference between John Olerud and Ryan Klesko really that big a deal? I'm reluctant to concede Piazza's a better glove than the Myers/Perez combo. The Robin Ventura versus Chipper Jones argument seems to be the only one where the Mets have a clear advantage defensively.

After The Game

JS: Oh...here's another reason: because Bobby Valentine is too unbelievably stupid to hit for Rey Ordonez when he's the tying run with two outs in the ninth inning.

RJ: Funny... I actually thought that Valentine's willingness to pinch-hit for Ordonez when losing late could be the key to the Mets' chances in this series, since their lineup 1-through-7 is pretty solid. Don't the Mets have any backup infielders on the roster?

One small point: if Benny Agbayani pinch-hits, he gets the intentional walk to bring up...Bobby Bonilla? At that point, do the Mets have anyone who's a big upgrade on Ordonez?

(Mind you, with the tying run on base, it's still the right move to get Ordonez's sorry butt out of there. Put Bonilla at third base and Ventura at shortstop if you have to, but get to the bottom of the ninth first.

JS: Uh...no. Bobby Cox would never intentionally walk the tying run with his dominant closer on the mound so he could give the winning run a chance to bat.

Jeff Hildebrand: Nope. Costas mentioned that Luis Lopez was left off the roster for this round of the playoffs. If that's really correct, I'm stunned by Valentine's stupidity. Amused, but stunned.

GS: Melvin Mora is the backup shortstop. Yes, it seems like he's an outfielder, but he was playing shortstop at Norfolk everyday before he was called up at mid-season.

JS: That's actually not such a bad idea, given that Valentine has apparently decided that Ordonez hits in any and all situations. Lopez is really a wasted roster spot, in much the same way Luis Sojo is on the Yankees.

I actually forgot about Shawon Dunston, who was still in the game at this point. He's actually a great person to have on this team, and would be the natural choice to play shortstop.