Pitching is something that is pretty clearly a key component of this business plan, of building a championship team. From the studies on success in the playoffs, we find that offense has not influence on the results, on winning World Championships, that it is pitching and fielding that wins World Championships. From the roster structure, we find that building with pitchers reduces the risk of developing good players, only to find that you need to trade to complete your team, because for pitching, the cream rise to the top. If you have 3 good pitchers, you have a great rotation, but if you have 3 good 1B (and only 1B), the team has to trade at least one of them (two for an NL team), which introduces a trade risk.
From the Pythagorean calculations, we find that when you have a great defense, you can actually win 90 games easily even when your offense is very sub-par. Thus a rebuild is faster when a team focuses on pitching first - as cream rises to the top - and supplementing the offense - due to Pythagorean - to complete the team. Now a study shows another positive aspect of developing your own pitching: that's the only sure way of obtaining good to great pitchers.
Trading Prospects is a Sucker's Bet Most Times, so is Signing Other Team's Free Agents
The Hardball Times publishes a Baseball Annual and the 2012 edition had a very interesting study in it, looking at how much of an advantage it is to know your own personnel. The chapter is "Down with Other People's Players" by Matt Swartz, page 207. He noted:
But does this extra information matter? Does another team's extra knowledge about a player affect the market? Should GMs think twice before signing a free agent from another team? Should they wonder why the other team is so willing to trade that highly ranked prospect?
Yes, a thousand times, yes.He calculated the $ per WAR for all types of free agent deals and he found that while the premium for hitters is only 17% when signing other team's players, it is 93% for starting pitchers and 139% for relief pitchers. He found that there were some hometown discounts given, but not enough to explain the premium.
Meanwhile, he examined players' performances after signing or resigning, versus Oliver projections. Re-signed hitters do not beat projections any better than newly signed hitters. More importantly, re-signed pitchers beat their projections, while newly signed free agent pitchers fell very short of their projections. That's a double-whammy right there, not only do you pay a lot more for these free agent pitchers, you also end up with less than projected performances as well, as if the former team knew that they were not going to continue to pitch as well as before.
Trades are no better. He found that "everyday players and starting pitchers both fall short of projections more often after being traded." Given the evidence above, this is not surprising, "we should expect that players who are traded will underperform projections more than players who stay on the same team." "Teams probably were able to detect injury risks that would not be observable to an outside projection system and send those players away in advance."
Strong Evidence That Teams Really Do Know Their Own Players Better Than Other Teams Do
He says that it is very strong evidence that teams know their own players - hitters and pitchers - better than even reasonable forecast systems do. "Caveat emptor." Trading is fraught with risk because other teams do not willingly give up good prospects to you. Signing free agents is particularly fraught with risk, and particularly with pitching, because you need to overpay (roughly 100% premium) for expected performance but meanwhile these free agent pitchers tend to underperform, giving you a double whammy. Again, teams generally know what they got and don't let go of players who they think are worth keeping.
The same goes for prospects. He examined prospects from Baseball America's Top 20 from each offseason from 1990 to 2007, covering 360 prospects, dividing them into those kept and those traded away. He found that "highly ranked prospects who are traded are more likely to be busts than the highly ranked prospects that teams retain." That was true by many angles, for while the average ranking was roughly the same, the average WAR for traded prospects was less than half that of the prospects kept, who on average ended up being good players (average WAR of 17.1). Furthermore, "the biggest difference was the number of elite prospects who managed 15 WAR, which 42 percent of untraded prospects were able to do, but only 18 percent of traded prospects did." "Teams are better able to tell which of their own minor leaguers are going to succeed than other teams' scouts, who see them less frequently."
In conclusion, he noted:
The safest bet is to develop your own players and keep them when they seem likely to maintain or improve their performance. When teams do need external help, they should fill holes with the reservation and suspicion derived from the knowledge that if a player is available, there is probably a reason.Duh: Teams Need to Develop Their Own Good Pitching
Therefore, it behooves teams to develop their own pitching, that is, those who will lead the way to championships, per the latest studies on winning championships. I know that was really obvious before, but this new study makes the imperative that much more greater to follow this rule. It is better to have the pitcher in hand than to rely on the trade or free agent market to acquire them, as the pitching you need in order to win championships most probably is not available in any other way, whether in trade or via free agency, and whatever is out there on the free agent market is way overpriced (100% or more). The exception there, of course, are Boras players, so they might be good, but they most likely will still be overpriced.
On top of that, given that it is so hard to find any prospect in the draft, the most efficient way for a team to rebuild a world championship-caliber contender is to focus first on pitching, for all the reasons outlined in this business plan, the flexibility, the control, the risk mitigation with regards to finding multiples, the ability to win with subpar offenses, all that adds up to a quicker turnaround when a team focuses first on finding and developing good to great pitchers.
Once you get that core group together, it is relatively easier over time to trade for or sign bullpen arms who can round out your group, and once you start winning and start getting back of the first round picks, where it is much harder to find talent (as per my draft research), you can locate useful hitters and pitchers (like the Giants did with Brown, Panik, Stratton), guys who can complement the good players who the team had found and developed previously. Meanwhile, once you got your great rotation, you still need to prime the pump, in case of injury (Lowry), reduced performance (Lincecum?, Sanchez), or free agency (Lincecum?), and still draft and develop pitchers. The Giants appear to be doing that, with Crick, Stratton, and Blackburn in hand.
1. This research destroys the pipe dream that sabermetricians can suss out superior players from stats, without substantial, informed scouting. The stats are equally available to everyone, so that a player's current team would have no edge over a team considering acquiring him.
ReplyDelete2. Because starters throw a great many more pitches, different in variety and location, than batters have plate appearances or opportunities to hit a great variety of good pitches, a scout can get much more evidence about a pitcher's capabilities than about a batter's. The advice to stock up on homegrown pitching would lead a team to hire where it had the best information.
Maybe hitting is just like the old local news adage: if it bleeds, it leads. They sure love them dingerz on highlights. But good pitching, even amazing pitching, gets highlighted seldom, maybe when Pedro was in his prime, or the first couple years of The Franchise. Sabean and Tidrow have completely re-worked the Giants.
ReplyDeleteOGC - I've seen you reference this piece, are there any links, or is just the annual?
The core, much like in 97-02, that is the key here. When you only have to add a couple pieces, you have the added advantage of being able to be pickier. Its one of the reasons I think the Giants draft up the middle players often. Those are the premium spots. The slugging corners is somewhat a thing of the past, I think fans expectations are still catching up, especially with left fielders.
Interesting thoughts OGC, nice post.
You know what, it might be from a post on-line, that makes a lot of sense, I'll see if I can figure that out for you, and if so, post it here.
DeleteYeah, that's actually been one of my thoughts which I've never touched upon, but my thought was more specific to SS: SS is the closest to a pitcher in terms of the flexibility that I've touted about pitching, where the creams rises to the top, and the pitcher can fill one of many spots in a pitching staff, whether starter or reliever.
One thing I've noted is how a pitcher can go to many different spots. Shortstops typically can move to 2B or 3B, as their next option, if they can hit well enough, particularly 2B. Some can make the move to CF too, others can handle the corner fielder positions at minimum.
Mark DeRosa is an example of that type of flexibility, he hit OK, but really is a platoon guy, but he could play at least 4 different positions with great defense - 2B, 3B, LF, RF - and probably plays CF and 1B passably before, and perhaps even SS too, in a pinch.
The Rays did that with Ben Zobrist.
And CF is the flexible player in the OF, capable of playing all three OF positions, we have seen a lot of that in recent years with Winn, Torres, Ross, Melky, Pagan, Blanco.
The Giants have shown that type of focus in the draft, especially since Barr came on. Crawford, Carter Jurica, Joe Panik are all SS they the Giants have drafted. To your point, Brown is an up the middle guy (and he used to play 2B I believe in high school), Payne too, even Parker was a CF before. And I'm sure there are others.
And the Giants are always picking up a catcher or two.
Swartz once wrote for BP, so while I could not find that exact chapter on-line, I found bits and pieces of the research that he's done over the years that make up the whole chapter:
Deletehttp://www.baseballprospectus.com/article.php?articleid=11003
http://www.baseballprospectus.com/article.php?articleid=10883
http://www.baseballprospectus.com/article.php?articleid=10096
Seems like he redid all that above BP research using THT/Fangraphs resources in order to write that chapter. He also wrote the following THT articles that had bits of pieces of the research in the chapter:
http://www.hardballtimes.com/main/article/big-long-deals/
There are other links in that article, but I could not get them to work, might have to search Fangraphs to get to the actual link.
Also, I ran into these two interesting articles regarding pitcher's velocity that is OT, but interesting still:
Deletehttp://www.hardballtimes.com/main/article/you-shall-know-our-velocity/
http://www.drivelinebaseball.com/2012/05/12/making-the-sabermetric-argument-for-increasing-fastball-velocity/
The latter one drives home one of my pet peeves about the Giants, and baseball in general, in that there are many ways a team can help their prospects reach their potential and yet they don't do anything much for them, really. If anything, they make it a really hard road.
Here is an article that latter article's author wrote about a prospect: http://www.hardballtimes.com/main/blog_article/how-baseball-failed-steve-delabar/
Why give tools to the players then have nobody in the organization who can explain it to the player and help them understand how to benefit from the analysis (which probably was expensive to boot!).
I can't imagine that it would cost all that much either. Off the top of my head, $1M per minor league farm team, to add instructors and tools there that all the prospects can utilize and learn/grow from, that's, what, $5M, the price of 1 WAR in free agent signing. So if that converts one prospect into a player who can provide one average player season, or 2 WAR, for a couple of seasons, it pays for itself if you can find one of these players every 2-3 years.
But I know why this don't fly in a business environment. There is no proof that the investment of $5M (or whatever millions) is going to yield anything. It is all belief, theory, and supposition.
Boddy's article is good for showing how ROI could be measured and generated, so that is a start, but that's presuming management believes that you can easily add 4 MPH to any pitcher. Why a team don't take up his offer, I don't know.
It could be make good too. Maybe the team gives him living expenses, which is not that much, while he trains their prospects at one minor league team, and he gets some set fee for every pitcher who increases their velocity 4 MPH. That's the kind of gain share pricing that is done by some companies in technology deals, paying based on a shared risk/reward business benefit type of deal. Of course, this forces him to put some skin in the game too, by investing his time doing this and perhaps coming out with nothing but living expenses (or whatever he thinks he can negotiate from the team).
I've thought similarly with the guy who Andres Torres used to learn how to proper hit a baseball. He was taught to slap at the ball, but this guy taught him how Pujols hits, which is basically what is taught in Ted Williams classic book, The Science of Hitting.
DeleteThe Giant should just hire him or a guy like him to stay with teach minor league team, and to work with any player wanting the extra help. I'm a firm believer that you can't lead the horse to water, and why waste time with guys who think they know it all and won't listen anyway?
As a pilot, they could start with the Volcanoes or maybe the Rookie team, and do that, and then expand if it proves successful. I personally would do it at Augusta first, that's where most of the borderline prospects end up first anyway, as the better ones are jumped higher to begin with, and if that proves to help, the move it to all the farm system teams. That seems the better test of this new way to run player development.
If a team invests $1M per year, converting one failed prospect out every 30 years into an average ballplayer who gives you 3 average years out of 6 (or 6 WAR) as the rate of production (getting an average player seems a lot to ask, I realized), assuming 1 WAR = $5M. Of course, we need to subtract the value of the contracts he earns with that, probably reduces it to one every 20 years. Seems like a pretty low rate of production to justify the expense.
If you put in $5M per year, you need to convert one player every 6 years.
And there is the rub: who is going to invest up to $25-30M over a 5-6 year period to test this idea out? That's a huge hole if this theory is wrong. We need a team willing to chip in $1M per year as the pilot and see how that goes.
And I think baseball has a good example of this in its history. The Royals used to run some sort of baseball academy that I recall they used to either find or develop George Brett and a number of other players. That would seem to validate the idea of having some sort of academy to teach your prospects the same concepts and stuff, though honestly I know that academy more by reputation and other sources, than reading the exact history of the endeavor.
Ultimately, it comes to the bang for the effort. It is like what I complained about in the MCC discussion of the draft last season. Why do all that extra action and not get much return out of it?
DeleteSame for this: you want the big bang, but the success rate to justify the investment is so low that you've spent a lot of money before you realize that it is not working. Of course, a team could just do it a year or two, but what if that is not a proper enough test, what if you realistically needed 3-5 years to find that average player?
It would be easier if you knew within a year or even two if you get a big payoff. But the rate of finding a good or even just an average player is so low that you cannot easily measure the success of the theory without investing a lot of years into it.
And, if you prove the theory is right, then every other team would copy you and there goes your advantage, after you invested all that money into it, proving it correct. Why stir things up just to have a competitive advantage for a year or three? And also risk losing a lot of money before you adequately test the theory?
Those are all good reasons why a team won't take this route.
Or rather, take that first step.
DeleteSome of the articles Matt linked to were on Fangraphs, and you can see them in the author's list:
Deletehttp://www.fangraphs.com/blogs/index.php/author/matts/
Good stuff, and he's the developer of SIERA.
Thanks for links. Read em tonight when I'm off me phone.
ReplyDeleteRomo signed, 2 years! Bobby Evans is the man. No more arbs.