Thursday, April 5, 2007

Oh what a relief it.... isn't

Once again, time to revisit the middle reliever question, and once again, I have to respectfully disagree with a fellow poster regarding the value of relief pitching. I won't pay more than five million dollars for any relief pitcher. I might consider paying those dollars to two or three guys, but after that, it's minimum salaries only. Let's take a look at why.

I wrote last year that of all the positions in GDR, the most volatile and unpredictable is relief pitching. This, from last years blog:


"From GDR, the top 15 relief pitchers from last year were:

Rivera, Mariano
Nathan, Joe
Shields, Scot
Turnbow, Derrick
Lidge, Brad
Wagner, Billy
Cordero, Chad
Rodriguez, Francisco
Timlin, Mike
Jones, Todd
Gordon, Tom
Dempster, Ryan
Farnsworth, Kyle
Hoffman, Trevor
Ryan, B.J.

How many of them were paid over 3 million dollars? Seven of the fifteen. In particular, Turnbow, Jones, Cordero, Timlin and Farnsworth can relatively cheaply, all under two million. Now, how many of the top 15 salaried relief pitchers made this list. Again, less than half (7) of them were able to get into the top 15."

I went on to write that 13 of the top paid 15 fifteen first basemen were also top performers, a much more reliable group.

Now that we've taken a look back at 2005, let's look at the top performers from 2006, and make some comparisons:

Top 15 GDR relief pitchers for 2006:

FRod
J.J. Putz
Joe Nathan
B.J. Ryan
Takashi Saito
Billy Wagner
Trevor Hoffman
Jonathan Papelbon
Francisco Cordero
Bobby Jenks
Scott Linebrink
Joel Zumaya
Scott Shields
Mariano Rivera
Akinori Otsuka

Right away you'll notice that most of these guys are closers. Since you should only have one closer in your lineup each week (if you have to ask why, find another game!), there is clearly lots of talent to go around at the closer position. If you paid a ton for FRod or Nathan last year, you missed out on the chance to purchase Putz, Saito, Papelbon or Otsuka for the minimum. While you can expect Nathan and Rodriguez to score well over 600 points, there isn't another RP in GDR that you can reasonably expect to score that many. So while a higher salary for Nathan can be somewhat justified, a higher salary for Zumaya cannot. Even old reliable Scott Shields dropped nearly 100 points from 2005 to 2006. Zumaya, even as dominant as he was last year, scored the same as Shields (who had a 'down' year) and Linebrink, and only marginally better than Otsuka, Proctor and Dan Wheeler, all of who could be had for far less money.

So, how many guys on last years list could have been had for under 3 million. Again, the answer is seven (Putz, Saito, Papelbon, Jenks, Linebrink, Zumaya, Otsuka). It simply makes no sense to overpay at this position because of the large variations in scoring.

I'll pay big bucks only for Pujols, Santana, Utley, and Reyes, because they are very likely to score 75 to 100 more points than ANY player at their respective positions. After that, especially in 12 team leagues, there just isn't enough difference in talent to justify a large investment in any one player, and that's especially true for relievers.

6 comments:

Blackjack Attack said...

The MR position value is a question of it’s variability. And the value applied to top MR’s is one of the most difficult to assess. However your stats are for relievers in general not MR directed. Now I do not agree with the assessment of the MR value. First good MR’s tend to move to SU and CL and more rarely SP, in doing so does their “value” change, it’s the same pitcher, same team(in this case). No the value hasn’t changed, but it may be a team already has two good CL’s and to that team the value has been destroyed. In the difference to overall value vs. team value the matter of team structure is important, but league structure is just and important(6vs16 team league). The variable nature of relievers delivers more value(damn the FR position, it does screw things up) that hitters. A hitter that’s 4 FP a game vs. 3 FP a game, may one week get you an extra 20 FP’s but more often will get an extra 5-10. A MR that averages 8 FP a game may get 0 one week and the next may get 40. With three relievers to go with, give me the shot at the blow out by reliever, rather than a slightly surer but very thin 7 point lead.

Re-run the data with MR’s stats and lets see where things are.

My data after two days scoring MR Heilman 23FP, SU Shields 23FP(Not a typo), CL Nathan 25FP…. And the opponents down about 50 FP and used more starters. With no chance of catching up without his relievers appearing in every game for the rest of the week.

Top 5-10 relievers are worth cash because they can carry a team to a blow out win when they get 3-5 games in a week.

Blackjack Attack

schlesinj said...

I agree with the original poster, do not pay for MR and closers at all. In both the Ruth League and my regular league both of which started last year several owners either dumped their closers and MR they paid big bucks for last year or just had to keep a $6 million Kyle Farnsworth. Has anyone heard of Cla Meredith before last year besides Cla's parents? Ironically, this year Cla went for to much so it is time to find the next guy. How about Tony Pena or Henry Owens?

leehaak said...

blackjack, it appears you base your argument on the variability of the position. To wit, you conclude that variances are to your advantage, therefore paying more for variable positions mkes sense. While that's an interesting argument, it flys in the face of basic statistical analysis.

If you have a good team, your goal is to perform as close as possible to the statistical norm for your team. Want proof? If every team in a league performed exactly to the norm, the best team would win every game. Conclusion: Better teams are harmed by wide statistical variations. This is one of the reasons I like the virtual series, it will tend to even those variations out, and the best teams are less likely to experience wide variances in record.

Woody said...

You're wrong. In fact, you're so blatantly wrong that its digscusting.
Like you had said, the top 15 relievers were
"From GDR, the top 15 relief pitchers from last year were:

Rivera, Mariano
Nathan, Joe
Shields, Scot
Turnbow, Derrick
Lidge, Brad
Wagner, Billy
Cordero, Chad
Rodriguez, Francisco
Timlin, Mike
Jones, Todd
Gordon, Tom
Dempster, Ryan
Farnsworth, Kyle
Hoffman, Trevor
Ryan, B.J.

then you said
Turnbow, Jones, Cordero, Timlin and Farnsworth can relatively cheaply, all under two million.

So basically what your saying is that if you didn't pay a lot of money, your chances of getting a decent (not great closer) was 50%. Essentially your argument is that if you spend big on relievers, there are always relievers who have good seasons that can be had cheaply (i.e. Putz, Papelbon) but for every Putz or Papelbon there are countless relievers who people will WASTE .5, 1, or sometimes even 2 million dollars on only to realize that their success was just a flash in the pan. When you pay big bucks for those pitchers you KNOW that those relievers are going to get you a huge amount of points that can make up for almost any other problems on the team. So paying over 5 for a setup who has established himself as a guarantee every year (Shields), has a filthy arsenal of pitches (Zumaya), or had absolutely insane numbers (Linebrink) you KNOW that (barring injury) they will be successful. When people signed Papelbon or Putz, they were one of many 'flavor of the week' relievers who seemed like they'd be worth taking a shot on with .5. Just because 2 succeeded doesn't mean that that is the strategy that needs to be used. May I remind you that Brandon Lyon was almost as successful as Papelbon to start the season? What if someone chose him instead of Papelbon, their bullpen would have been horrible last year.

leehaak said...

Lol, I didn't realize we got points for 'a filthy arsenal' of pitches. It's obvious you've missed the point here. My point was, is, and will continue to be that it is difficult predict the success of relief pitchers, and therefore you are much better off allocating that money to some other position, and not paying big for relief pitching. I'm sorry you didn't understand that. I'm not arguing that a relief staff that is assembled using this strategy will be better than one you paid 40 million for. I'm just saying that the utilization of your budget is counterproductive if you spend big money on anticipated performances of relief pitchers as opposed to hitters. I'm not sure what you're trying to say here, but hopefully this will help you better understand my point.

Final point, my chances of getting a decent (or maybe even great) closer is probably much closer to 100%. I don't need a top 15. I've employed this strategy in pretty much every league I've been in, and never found myself without a solid closer. Same with middle relievers.

leehaak said...

Woody, final comment for you. You've paraphrased my statement to mean that 'if you didn't pay alot of money, your chances of getting a decent (not great) closer was 50%'.

I never said, or even implied that, so to put those words in my mouth takes a leap of logic that is so great that it's hardly worth addressing, but I don't think it should be left unchallenged.

Perhaps you can explain how you figured out that I was implying that, because there seem to be two ways to get to your conclusion. One is based on a factually incorrect assumption, the other appears to be based on my entirely random decision to show the top 15, rather than the top 20, or 30. Maybe you can explain how you came up with that conclusion?