If I recall it correctly, Bill James' argument for a 4-man rotation is essentially the same. Why take 32 starts from your four best guys and give it to someone who can't break into the rotation?
He also adds to the argument the issue of use of a player's talent and skills. Back in the day when you mostly expected a player to stay with one team for most of his career, a team was wise to nurture that player along and not burn him out, to give him more good years with your team. But in this day when most players play for half a dozen teams or more in their career, a team is better advised to use every bit of a player they can use while they have that player.
Sam,
That's a good point and one I've never really understood. Sure it's heartless to a degree but if I'm a team, why don't I use a player to the betterment of my team. Do I really care as an organization if he blows his arm out for the next guy?
Have your read Art DeVany's paper on Steroids in baseball?
I think that make sense. but one of the question is that would those roster spots be better served by additional bench players? sure. having 1 or 2 more bench guys would be nice (particularly in the NL) but 3 or 4?
The comparason (espically with the relievers) isn't neccesary the extra relievers value vs the 3-4 best once. but its also those extra relievers value vs bench guys that will never play.
I think what your saying looks at the picture too much from the marco persepective and not the micro. yes. those last 2-3 relievers looks useless in the big picture and is just taking innings from the better onces. but if you run into a bad stretch where your starters keep stringing together short outtings wether from poor performane or injuries. Would those extra relievers REALLY be useless? or are they saving your butt and not blowing out the better RP's arm.
I think we can safely say that RPs being used 06 Scott Proctor or Scott Shield style is probably not a good idea in the long run.
To add on, I think the point your overlooking a bit is..
1. those extra roster spot would probably be even more useless in the hands of 2-3 bench player that will never play, or take playing time away from your starting fielders.
2. The bad cases where you keep running into short outtings and/or extra innings all strung together in a short stretch. It is a lot more likely for this to happen in a season than for your starting fielders to run into so many injuries that you run out of bench players (And if it REALLY gets that bad, you COULD always have a pitcher play the field. which is actually not quiet as bad as having a fielder pitch.)
3. Having guys that can go long typically mean they are pretty effective starters in their own right. It is pretty hard to convince any (effective) starting pitcher to take a contract to be your long man these days. and it is kinda counter productive to use your best young pitchers in this role due to development reasons.
A couple of comments:
Josh,
No I haven't read the article. Can you send me a copy via e-mail or a link to it?
RW,
I think at the heart of the matter is that the game is significantly broken. For example, I'm just pulling a team out of my posterior but let's look at Pittsburgh and see where it takes us. If you look at their rotation you have:
1)Duke 2)Maholm 3)Ohlendorf 4)Snell
Outside of Duke everyone has a below average ERA+ but let's just for the sake of argument say those are our 4 and leave Kartsens out of it. Right now the 5th starter would theoretically get 30-31 starts per year. That's 19% of starts. That's too many for a backend starter that is probably nothing more than a AAAA starter.
So the thought is can you go to a 4-man rotation? I don't know. What I do know is that studies have shown that 5-man rotations don't descrease frequency or severity of arm injury.
So why do we need a 5-man rotation? My guess is that it is something they tried to avert arm injuries but now you don’t have the prevention so why the need especially if it actually goes towards you losing more ballgames?
My guess also is that the union has something to do with it in that it’s perceived that more rest is better although empirically it hasn’t.
But I think the other notion too is that relievers can’t go more than 1IP per inning. You talk about not using pitchers like Scot Shields for 100IP but back in the 60s-70s relievers were routiniely used like this and to great success. That is the whole argument surrounding Tim Lincecum and Joba Chamberlain when they came up. Would they be more useful as 120IP long relievers used in high leverage situations? It’s a helluva point as an inning is an inning is an inning.
So you lop a starter off but why not have a couple of long relievers/spot starters on the teams? For Pittsburgh you obviously would have Matt Capps, John Grabow, Jesse Chavez and Sean Burentt (used to). That’s 4 relievers. Now we are doing math problems.
Pittsburgh threw 1455 innings last season. If the starters can each go around 250IP then you have 1,000 innings right there. If you have 5 relievers to hack up the 455 other innings then you have 91IP per reliever which isn’t horrible to be honest. That’s a 9-man staff. You can go upwards of 11 for all I care. If the four starters can then go 200IP a piece then it’s 800IP with 655 to go between 7 guys which is 93IP per reliever and I don’t think that is asking a ton. If you have a couple of long relievers in the pen who can to 110-120IP then essentially you can garner 800IP from your 4 starters at 200IP/Starter. You go 240IP from your 2 long releivers at 120IP/Long man and that gives you 1,040 with only 415 to go. If you have an 11-man staff with 4 starters, 2 long men and 5 relievers you are still stuck with 83IP/Reliever and I don’t see why any of this is a problem. You have 11 pitchers (which may be 1-2 more than you need) on a staff which effectively cuts the staffs
Cont.
by 2 pitchers sometimes 3 in a given circumstance.
And keep in mind that this is relative. If you have a 4-man rotation then each starter gets 40-41 starts. If he’s only expected to go 200IP for the year he’s on average pitching between 4.9IP-5IP/GS. That’s ridiculous. If a pitcher averages just 5.3IP/GS then over 41 starts he’ll get 217.3IP. Multiply that by 4 and you are looking at 870IP. That’s 585 innings left over. A couple of long men at 110IP each knocks that down to 365IP. If you want 85IP/reliever then you need only 4 relievers leaving you with a 10-man staff.
If the fundamental problem is pitch counts then the real problem behind this is pitching to contact. Baseball today is all about the HR, K & BB. It’s a very slow plodding game that probably doesn’t rely on contact as much as it maybe should. Whether you like that or not it’s the way it is and a lot of teams have to be this way because that is what baseball is today. With every advantage going against the pitcher and ballparks being smaller and smaller, then the only way teams can be decent is to hit a lot of HR and do not pitch to contact because if you do then the other team is going to hit lots of HR. Guys like Greg Maddux can get away with it in this era, but Maddux is a rare animal that doesn’t come along all that often.
If you are pitching to contact then of course you are lowering pitch counts and this means more IP and less strain all the while pitching more innings. Efficiency is letting your defense do the work. Sure it’s impressive to strike out the side but you may need at a minimum of 9 pitches and sometimes you get up 13-15 pitches. For three groundball outs all you need is at a minimum 3 pitches. Stiking out the side is 4x-5x as much work as getting groundball outs.
Of course to a certain degree you have to expand the strikezone to that effect. That scenario I just describe makes it all but impossible to be more efficient with pitching staffs. The long reliever/spot starter is a relic of the past. Specialization is king now and thus you get 12-14 man pitching staffs staffed with players who really don’t belong in the major leagues.
So if you can fix that problem with Pittsburgh and come up with 10-men then that leaves 14-men on a 24-man roster. There is no need for the 25th man but again that has to be a collective bargaining issue because eliminating the 25th man costs 30 jobs at the major league level and there is no way the union is backing down on that.
But take Pittsburgh. You have Ryan Doumit at catcher. He’s solid both ways as a swith hitter but you have to have a backup catcher there. So there are 2 spots right there. Now we are up to 12-men. But look at LaRoche at 1B. You can argue you could platoon him at 1B against LHP given that his career OBP% against lefties is .315. That’s brutal especially from your 1B. Pittsburgh could obviously use a platoon guy with LaRoche against lefties. That’s 2 players right there giving you 14. S
Cont.
Sanchez has a massive platoon split against righties. His OPS against RHP is .716. Against LHP is .896! Plus Sanchez is an average defender at 2B. I don’t think it would be a horrible idea for Pittsburgh to have a defensive replacement for Sanchez with a RHP on the mound. Also one against RHP is a possibility. Whether it’s another 2B or a MI it doesn’t matter. You need both which now ups your roster to 16. Jack Wilson is another who can’t hit RHP. He’s not a great hitter anyway, but he does hit LHP (OPS = .762) and he has an outstanding glove. Wouldn’t Wilson make a much better platoon player as a SS against LHP and a defensive replacement late in games with the lead? Because both Sanchez & Wilson are challenged against RHP you can theoretically use the same MI backup but that means you are now up to 17 players. Andy LaRoche is fairly even at 3B although he’s far from good. At this point you don’t really need the extra player so you are up to 18 players on the roster which leaves 6 spots on the 24-man roster. In the OF you look at having 5 players, the 3 outfielders and your 4th & 5th OF. That gives you 23 men and you could easily pick up a 3rd catcher or another special pinch hitter to make up th 24 man roster still keeping with only 10 pitchers.
That gives you: 4SP, 2LR, 4RP, 3C, 2-1B, 1-2B, 1-3B, 1SS, 1MI, 5OF = 24 Players
I just feel like you lose massive flexibility when constructing a lineup around 12-13 pitchers.
I write all that to essentially say that I think there is inefficiency in baseball, but now I think it's roster construction and I'm not sure how you get around that because it's obvious the powers that be want offense and lots of it.
It's also obvious that pitching to contact isn't coming back and that rosters aren't going to get cut.
Still, the inefficiency is using ballplayers more than what they normally would be playing or using them less than what they should be. I'd rather have Josh Beckett pitching 290IP if he can get away with it on the pitch counts than sign guys to make a start every 5th day. It's ludicrous.
I'd also rather have Mariano Rivera go 95IP a year than 65IP if the other 30IP are going to backend relievers who don't belong in the majors.
Teams that gamble on this and make it pay off I think will be the next ones to completely explode this market inefficiency and start hammering away at the competition.
A few thought.
A. Yes, I kinda agree with the 4 man rotation thing.
B. Relievers going past 1 inning. I think it is possible. but it really end up depenindg on what the guy has. a two pitch guy tend to be quiet effective the first time you face him but not the second time round . there's a good reason that a lot of these good / great reliever weren't much of a starter (Rivera , Gagne , Gossage) because their 3rd pitch (or in Mo's case. second pitch) is just not quiet there.
The evolution of the recent decades have been about that. there are pitchers who really have 2 GREAT pitch but can't field a 3rd decent one. Or they simply don't have the frame / stamina that they ended up gertting crushed the second / 3rd time around the order as a starter. However if they are used in shorter stints they endup being great.
RW,
No question about the pitch repetoires. For the most part you have to have at least a 3rd pitch although great pitchers who were starters could get by on 2. Think of Roger Clemens and that cutter he used for the last few years of his career. Or even Randy Johnson who has made a career out of throwing a fastball/slider combination.
But even with your relievers, if you can get through the lineup once on just a 2-pitch arsenal then you are looking at a minimum of 120IP in just 60 appearances.
If you have 5 relievers doing that then you are looking at 600IP. Using the Pirates example earlier that leaves 855 for your starters or roughly 227 per starter which isn't a horrible amount if you keep pitch counts at 120 pitches/GS and never go over 30 pitches per inning which is the commonly accepted levels to prevent injury.
Also with the 4 man rotation you are knocking a lot of starters down a notch. A lot of 4th & 5th starters now become either AAA starters who are still trying to refine their game or they are long relievers out of the bullpen.
A lot of these guys are like you said, pitchers who probably aren't bad at getting through a lineup the first time, but who simply can't go through a 2nd time for whatever reason.
Still on average you are looking at 2IP to get through a lineup once and if you can get a reliever to give you 60-70 appearances at 2IP a clip then how much of an advantage do you have simply by limiting the specialization of your pen and opening doors to better use your team for winning purposes such as platoon players or defensive replacements?
It's a massive inefficiency but to exploit it, what can you do? There are so many things that would need to be changed just to get it going I'm not sure a team would be that committed to it.
I would think a team like Boston, Toronto or maybe San Diego would be on the precipice of something like that but how far away, if you ever would see it, is baseball from this?
Commenting by HaloScan
