Merits of Bob Melvin's bullpen philosophy?

[Thurston Fluff]

We've covered this a little in the past in both the Yost thread and somewhere in the year but I thought perhaps it was deserving of it's own thread so we could discuss the merit of the philosophy. If the mods feel it's already covered, sorry, feel free to lock. It's about Melvin's use of his bullpens worst performers in blowouts.

 

Baldkin was kind enough to link this article in theYost thread.

 

http://www.hardballtimes.com/main/article/no-mirage-in-arizona/

 

I see the merits of this but I don't know if it's something that is viable as a blanket sort of approach or if it's more exclusive to a weak offensive team with a good pitching with a huge dropoff past the top six relievers. There can be no doubt it worked quite well for Melvin this year and no doubt was the perfect approach for that team. But if he had a better offense would it have made the same amount of sense to concede games? If he didn't have as many good relievers to win so many one run games would concedeing games have been as smart as they were for him this year? I guess part of me really likes that approach but I just couldn't see giving up on many games with the type of offense we have. Or the lack of quality arms in the pen that we had last year.

So was it the perfect approach for that team or is it something we might see other teams with other circumstances adapt in the future?

I'm sure every manager does it to some extent in big blowouts when the pen is already depleted. Melvin seemed to do it more often, earlier and in tighter games than some would consider a blowout.

Maybe we could list circumstances when it seems viable for certain types of team.

For myself if the bullpen is tired and the team is down 8 runs in the fifth and the starter has already has a high pitch count maybe it's time to suck it up and play for tommorrow.

[sbrylski]

I read this article and don't get it. The secret to success is not having crappy long relievers, that just explains why their pythagorean sucks. If you want to defy Pythagorus, have crappy long relief. If you want to be good, have 5 great bullpen pitchers to go with a solid rotation and lineup.

 

I see nothing here revloutionary, nor any type of model to follow.

[dadofandrew]

I agree with Sam. Every manager would be a genius if he had 5-6 guys with a 3.5 ERA or lower to turn to.

[mothershipconnection]

That's always been my criticism of pythagorean W/L too. If you have a couple weak pitchers, unreliable starter, or a string of injuries that force you to put up with the 2006 Brewers fill-ins, you'll look worse than you really are. In the end, it doesn't matter if you have a 5.00 ERA starter or a 7.00 ERA starter if the opponent only allows 4 runs per game.

Comparing runs scored and runs against sallows blowouts to skew the picture. I think it'd be better to look at median runs scored and allowed per game. I'd bet the D-Backs look better in that department.

[rluzinski]

It's true that you want your offense to be as consistent as possible and your pitchers to be as inconsistent as possible. After all, if your pitchers are going to have a 4.5 ERA average, better that they sometimes give up 8 runs a game than 6. You are probably going to lose either way. That said, most teams have a fairly typical runs per game distribution over the course of the year, which the standard pythagorean equation assumes.

 

Sometimes, it assumes wrong, however and the D-Backs's 2007 season is a good example of that happening. "SABRMatt" came up with a way to account for that, with his "pythagenmatt" metric:

 

http://detectovision.com/?p=1054

 

As of 8/24, the DBacks had a much better pythagenmatt record than pythagenpat record.

[dadofandrew]

Note how Valverde was completely worthless after 4 outs or so last night. The funny thing is, many think closers are just machines and can go 6 outs as easy as 3...when in reality, it only takes common sense to figure out it's twice as difficult, and your success rate will only be half as high.

 

I certainly do not feel it was a terrible decision to leave your best reliever on the mound, but you even have to wonder if Jose will be 100% after a day off, going 42 pitches.

[TooLiveBrew]

There was a thread similar to this after we'd played AZ earlier this season. The same allegation (that Melvin only uses his lower relievers when trailing by 'a bit' was made then, and when looking through the game logs, I could really only find a few games that'd fit this theory at that time. I'm admittedly too lazy to do it right now, but I think this whole thing is way more a product of having a deep group of very solid/good relievers than any genius managerial strategy.