Better comparison would be Rivera/Maddux, the Big Unit or Pedro.

Would you rather have Rivera as a closer or one of the three above as a starter for their entire year? I think the answer has to be Maddux, Johnson, or Pedro (maybe not Pedro because of injury concerns.

Also, I would like to point out that pitchers that have started and closed during their career believe they made a big impact as a starter. For example, Smoltz who was adamant about returning to the starter role.

Another key point is that a starter can close if a team is in a pinch, but a closer can’t start in a pinch. For example, Johnson was more valuable than Rivera during the 2001 WS because he won Game Six and served as the closer in Game Seven.

You would rather have Smoltz as a closer than a starter? Between 1990 (at age 23) and 1997, Smoltz started 30 games every year but two (and one of them was the strike year). He started 35 games four times. He led the league in innings pitched twice. And he still managed to keep his ERA under 4 every year (except ’94), keep it at or under 3 three times, and lead the league in strikeouts twice (once he was best in the majors. As a closer…he was excellent for three years, in which he pitched 225 innings. On what basis would you rather have the second guy than the first? Who made a bigger impact on his teams?

As for Gagne: Yes, he probably would have gotten hurt anyway. That’s part of the point: if someone has the stuff to start, like Feliz seems to, wouldn’t you want to get as many innings out of him as you can while he’s pitching? And how can you say, “Eh, Gagne would have gotten hurt anyway,” while simultaneously being certain that Joba would have developed into Mariano Rivera if the Yankees hadn’t put him in the rotation? Do you see why I say that you would have been saying the same stuff about Gagne, or Isringhuasen, or any number of other can’t-miss closers, that you’re now saying about Soria et al?

It’s true that there are some people who are better for one inning than seven, but they’re not as rare as you think. The A’s have shown that they can develop closer after closer–Isringhausen, Street, Bailey–without getting much drop-off between them. And as someone pointed out above, if great closers are so important, why aren’t they paid like it?

Part of the problem here is that you won’t say *why* you’d rather have a great closer than a very good starter. Do they win more games for their teams than good starters do? Do they last longer? What? All you’ll say (twice) is that they’ve shown they have “balls,” which while honest–I think that’s what most anti-stat arguments basically amount to, “he has balls”–isn’t a great reason to sign one player over another. A good reason to root for a player, but not to sign him.

And for the guy continuing to harp on Gagne…relax man. Gagne would have gotten hurt whether he was a starter or a closer, right? So what. He was a dominant closer and I would take his years as that over trying to throw him out there as a starter any day.

Remember, this all started with Feliz being the closer and you guys whining about it. Feliz has shown he has the balls to close games. Why change him? Same thing happened to Joba…dude was unhittable when he came up. Until those bugs flew into his face, nobody was going to score on him. Shoot, even WITH the bugs, he still almost got out of that inning. Then they go and try to make a starter out of him and screwed him up. He could have been Mariano part deux…instead, he’s just a fat toad.

I probably would take Rivera’s *career* over that of a above-average starter, sure–but what, are you absolutely certain that Nathan, Wilson, Soria, and Feliz are all going to be Mariano Rivera? There’s less than one Rivera in an average generation.

Chris I ask you again, if statistical analysis is used in life changing and life saving fields like science and medicine to look for meaningful patterns that represent reality well enough to choose drugs that save lives, why would that not be the proper method to analyze data stemming from baseball outcomes? Seriously if you trust the medicine your doctor gives you, you are a believer in sabermetrics because it uses the same methods and rigor as medical research. You cannot deny the validity of statistical analysis in one field, baseball research, while accepting it in another, medicine without explaining why the best method for analyzing data in life saving research is not the best way to analyze baseball data.

Whenever they do move Feliz to the rotation, and I’m almost certain they eventually will, I’m not too worried about the increased innings. CJ Wilson pitched over 228 innings last year counting the playoffs, which was more than his last 4 years combined, and he was just as strong at the end as he was in the beginning. Feliz has a remarkably easy and stress-free delivery, and he should be able to make the adjustment.

I even cringe when I hear that freaking song on the radio.

Seems like you might want to stretch him out a bit from last year’s ~70 IP before penciling him in for 200 IP as a full-year starter – maybe start him in the pen, and move him in to starting later in the year?

http://www.fangraphs.com/statss.aspx?playerid=18&position=P

But this is why advanced metrics are so useful, because some of them apply to this situation in that they attempt to measure total wins contribution by players. For instance Rivera’s 2010 WAR was 1.7, and Cain’s was 4.0. That’s Wins Above Replacement.

Brian Wilson is a great closer? I’m a Giants fan and I know that’s nonsense. Brian Wilson was very good *this year,* but I sure hope his future years aren’t like the years he had pre-2010. He had a 4.6 ERA two years ago–*and still had 41 saves.* Does that give you a sense of why people say saves are worthless? And how many closers stay good year in and year out, like Matt Cain does?

You’d want to make sure you’re not breaking your star reliever before running this experiment, right?

Who would you rather have…

Neftali Feliz for 180-200 innings, or Neftali Feliz for 60-70 innings?

I am talking about the real baseball world here, not fantasy. Who would you take…Cain or Rivera/Soria/Wilson/Nathan???

It’s one thing to drool over pitchers with high K/BB and K/9 ratios and imagine the championships they one day might bring, but another to ignore the contributions players have already banked. Unless you’re prepared to argue that WPA has no predictive value . . . I’m just wondering if that’s true for high-level closers.

It seems to me intuitively that WPA is pretty well correlated with actually winning. If you want to take a step back, if you’ve got a pitcher that could put in six good innings every fifth day . . . wouldn’t you rather put them in starting at the fourth inning of a 0-0 game, than happen to start them a day the offense puts up ten runs?

Who would you rather have…

Matt Cain or (Mariano Rivera, Joe Nathan, Joakim Soria, or Brian Wilson)

That’s the question…I think Matt Cain is a very good pitcher. I think those four are great closers. Who ya got?

Who would you rather have…

Matt Cain or any one of Mariano Rivera, Joe Nathan, Joakim Soria, or Brian Wilson

That’s the question…I think Matt Cain is a very good pitcher. I think those four are great closers. Who ya got?

Rule #1 Stats are nothing more than a record of actual occurrences

Rule #2 These occurrences may be used to determine the effects of each type of occurrence on the winning of baseball games.

Rule #3 Objective analysis of data is always more reliable than memory or personal perceptions, each of which are horrible incomplete.

Rule #4 Decisions based on objective analysis will be more effective than those based on “eye-ball tests” or “guts”.

Rule #5 Stats do not replace the love of baseball, but are simply used to increase understanding of it, which typically increases enjoyment of baseball.

4/14 12-, two out, 8th inning, ahead 4 (s)
5/3 1–, no out, ahead 2 (s)
5/11 –3, one out, ahead 1 (bs, team L)
5/20, 123, two outs, ahead 6 (no record)
6/9, 12-, one out, ahead 10 (no record)
7/3 –3, one out, ahead 2 (s)
7/22 nobody on, one out, ahead 1 (s)
8/20 –3, two out, ahead 2 (s)
8/27 1–, nobody out, ahead 4 (no record)
9/12 1–, nobody out, ahead 3 (s)
9/25 1–, two out, 8th inning, ahead 1 (s)
10/2 12-, nobody out, ahead 4 (s)

Nope. “Bases Loaded, zero out, ahead by one” never came up. The only thing close was on May 11, and he blew that one. He did come in with a 4+ run lead 14 times, for which he earned two saves. That was 13 of his 69 IP.

So yeah, most starters don’t do all that.