If there’s one group of free agents that, over the last five years, has been the most consistently undervalued by the market, it is clearly position players from Japan. Take a look at the following players who have come over, how much their teams paid them per season (posting fee included if applicable), and their average WPA/LI during the contract.

Ichiro Suzuki, Seattle Mariners – 2001 to 2003, $9 million – 2.0 WPA/LI per season
Kenji Johjima, Seattle Mariners – 2006 to 2008, $5 million – -0.3 WPA/LI per season
Hideki Matsui, New York Yankees – 2003 to 2005, $7 million – 2.1 WPA/LI per season
Tadahito Iguchi, Chicago White Sox – 2005 to 2007, $2.75 million – 1.1 WPA/LI per season
Akinori Iwamura, Tampa Bay Rays – 2007 to 2009, $4 million – -0.1 WPA/LI per season
So Taguchi, St. Louis Cardinals – 2002 to 2004, $1 million – 0.0 WPA/LI per season
Kazuo Matsui, New York Mets – 2004 to 2006, $7 million – -0.2 WPA/LI per season
Kosuke Fukodome, Chicago Cubs – 2008 to 2011, $12 million – -0.3 WPA/LI per season

Obviously, WPA/LI doesn’t adjust for position or defense, but the conclusion is still obvious – as a group, these guys have been a massive success. The original contracts for Ichiro, Matsui, Iguchi, Iwamura, and Johjima especially were ridiculous bargains. Not one of them cost their team more than $10 million per season, and all of them were above average players, including Ichiro and Matsui proving to be all-star caliber players. Even the so called busts, such as Kaz Matsui, were reasonably productive players at a not ridiculous price.

Maybe Fukodome’s contract from last winter signaled a shift in how teams view Japanese position players, and they’ll be fairly valued going forward, but there was a huge undervaluation of Japanese players for six solid years, and it’s hard to imagine that it all disappeared in a single winter. While I’ll leave the specific names of potential bargains to those who know NPB ball better than myself, it seems wise that teams looking for a good bargain would target that segment of potential free agents as an opportunity to find a quality player at a lower than expected cost.

For our final free agent bargain, we take a look at another starting pitcher who simply doesn’t get the recognition he deserves – Derek Lowe. I’ve written about him several times this year, but it bears repeating: Lowe had the 7th best FIP of any starting pitcher in baseball this year. Better than Brandon Webb. Better than Johan Santana. Better than breakthrough stars Ervin Santana and Ryan Dempster or established aces like Jake Peavy and Roy Oswalt.

Lowe was tremendous for the Dodgers this year, continuing his run as a durable, front line starter. Because he’s one of the most extreme groundball starters in the majors and he’s learned how to command his sinker, his success is built off limiting walks and home runs. He did both of those better than almost everyone else in baseball, which makes up for the fact that he posts just average strikeout rates. Getting batters to swing and miss is great and all, but it’s not the only way to succeed – Lowe has gotten so good at the other aspects of baseball that he doesn’t need to blow hitters away. Weak groundballs turn into outs with enough frequency too.

Because of his age, Lowe isn’t going to be in the market for the five to seven year deal that CC Sabathia will be looking for. The length of the contract we should expect Lowe to get is three or four years, which makes him more attractive than others on the market just for that reason. How much should Lowe get?

If we assume that Lowe’s true talent level will see him give up 4.25 runs per nine innings next year, and we project him to throw 200 innings, that makes him about 38 runs above a replacement level starting pitcher. Given that, we can call him a +3.5 to +4 win pitcher. Given an expected going rate of about $5 million per win this winter, we’d expect Lowe to get something like $17.5 to $20 million per year if he was valued correctly.

I highly doubt that Lowe will get that much money, though – the perception of his abilities across the game don’t match his actual abilities, and I’d put the expected range of his salary at about $15 million per year. If I’m right, Lowe’s going to be worth about $2.5 to $5 million in asset value to whoever signs him for 2009.

Teams like Atlanta, who are considering giving up the farm for the right to pay Jake Peavy the same amount that we’re projecting Lowe will sign for, should look at him as a viable alternative – similar caliber of pitcher, similar money, and you get to keep all your prospects.

Moving back to the mound in our series on free agent bargains, we find an interesting name – the tallest player in baseball and a sure fire Hall of Famer, you wouldn’t think Randy Johnson would fly under too many radars. But that’s the state of the game – overemphasis on ERA as a tool to evaluate pitching, and for whatever reason, a premature willingness to put great talents out to pasture.

Sure, Randy Johnson may not be what he once was, but he’s still a pretty terrific starting pitcher. Check out his standing among his peers this year:

8.46 K/9, 6th best in NL
2.15 BB/9, 11th best in NL
3.9 K/BB, 3rd best in NL
3.76 FIP, 15th best in NL

Or, if you prefer, his closest comparable pitcher in the NL this year was Cole Hamels – their walk rates are nearly identical, Johnson’s got a slightly higher K/9, and Hamels has a slightly lower HR/9, but the final product is almost exactly the same. Can you imagine what kind of money Cole Hamels would get if he was a free agent this winter? Now, obviously, there’s a huge age difference, and Johnson’s not going to be pitching for another 10 years like Hamels will be, but their current value is almost identical.

Despite all this, and the fact that he’s unlikely to ask for a long term deal given his age, the D’Backs are still talking about only being interested in Johnson if he takes a significant pay cut. That’s just nutty.

Even if we expect Johnson to regress significantly, giving up about 4.5 runs per nine innings, and only throw 150 innings next year, he’d still be 25 runs above a replacement level starting pitcher. He’s easily worth $10-$15 million a year for a single year, and considering how well he’s fought off decline, a two year deal shouldn’t even be out of the question.

Randy Johnson’s still a pretty terrific pitcher, and teams who just see him as an old guy with back problems are going to miss out.

As we move behind the plate for today’s free agent bargain, we take a look at a guy with one of the most inconsistent track records of any player in baseball – Dave Ross. Take a look at his seasonal batting lines:

2003: .258/.336/.556, 140 PA
2004: .170/.253/.291, 190 PA
2005: .240/.279/.392, 138 PA
2006: .255/.353/.579, 296 PA
2007: .203/.271/.399, 348 PA
2008: .225/.369/.352, 182 PA

That’s a weird set of numbers. In 2003 and 2006, Ross was pretty close to the best hitting catcher alive. In 2004, 2005, and 2007, he was replacement level. In 2008, he abandoned his normal skillset, hitting for about half as much power as usual but walking twice as often. That made him a decent player, but nothing like what you’d expect from his historical record.

So, what do we make of Ross going forward?

His contact problems are always going to depress his batting avearage – when you’ve got a 30% K%, you’re just not going to hit for much of an average. Even with the inconsistency, we can be sure that Ross is going to strike out a lot and not hit for much of an average. His value lies entirely in his secondary skills.

As for the ridiculous walk rate he displayed last year, that seems to be a direct response to a change in approach. Over the last four years, Ross has cut back significantly in how often he swings, seeing his Swing% go from 50.96% in 2005 to 42.74% last year. Despite a big cutback in how often he swings, he’s still making contact at essentially the same rate, so he’s not just staring at hittable strikes. It seems like that his 12% BB% over the last three years is more likely to continue than for him to revert back to his free swinging ways of 2003 to 2005.

So, what about the power? His ISO the last three years are .324, .197, and .126. Not exactly trending the right way, especially for a catcher who will be 32 next year. However, if we were looking for evidence that he wasn’t hitting the ball hard anymore, his career high 24.5% LD% in 2008 seems to rule that out. He just hit a lot of doubles instead of home runs last year, which naturally drove down his slugging marks. But considering that 12 of his 32 hits were still extra base knocks, it seems unlikely that his power is just disappearing at a rapid rate.

The Marcel projection system seems to be in agreement with these assessments, pegging Ross as a .232/.330/.421 hitter in 291 plate appearances next year. While the average is low, the secondary skills are still very strong, and project Ross as an above average hitting catcher for 2009. Not bad for a guy who was released outright in August.

There are quite a few teams who could use some added punch from behind the plate in 2009. If they can get over their obsession with batting average, Dave Ross could be a very low cost, short term option to provide some walks and power at a position where both are rare.

So, this time, I swear, the free agent bargain is actually a free agent.

Much like with Juan Rivera, this potential bargain was a tremendous player in 2006 – he hit .283/.323/.506, good for a 0.71 WPA/LI in 586 plate appearances. The power overcame the low on base rate, and he was a somewhat above average hitter. Meanwhile, John Dewan’s +/- system had him at +31 plays at his position, a staggering total that profiled him as an elite defender. As a third baseman, the combination of above average bat and remarkably awesome glove made him one of the game’s under-appeciated stars.

I’m guessing that practically no one has figured out that I’m talking about Joe Crede, because he certainly didn’t get much publicity for his outstanding 2006 at the time. As defensive performances get more notoriety, however, we can look back at that season and recognize it as one of the best under-the-radar seasons in recent history.

However, in 2007, Crede’s balky balk finally went out, and he had to undergo season ending back surgery. Those back problems cut his 2008 season short as well, and over the last two years, he’s been limited to just 551 plate appearances while playing through pain. His reliability is a real question, and the White Sox are expected to go another direction in their quest for a full time third baseman.

Now, I’m not a doctor, so don’t take this as any kind of endorsement of his future health. For all I know, his back problems could be career ending. However, if they’re not, and the medical people can figure out how to keep him on the field, his 2008 performance should assure potential GMs that Crede can still play.

The power is still there – he had a .212 ISO, built on 36 extra base hits in just 335 at-bats. His contact rate was identical to his 2006 performance, so it doesn’t appear he had to adjust his swing to compensate for the pain. Contact and power are the building blocks of a good hitter, and Crede’s abilities in those areas didn’t seem to suffer when he was on the field.

Defensively, it’s pretty much the same story. He’s played just over 1,200 innings at third base the last two years, basically a full season’s worth of games, and +/- has him at +24 plays over that time. Even as a step down from his +31 in 2006, it’s a great rating. You could conservatively drop his true talent level to +15, accounting for more aging, and he’d still be among the very best hot corner defenders around.

In terms of on field skills, Crede projects as something close to a league average hitter with defense that’s worth +1.5 to +2.0 wins above an average third baseman. Over a full season, that would make Crede a +3.5 to +4 win player. If he was completely healthy, we’d expect him to get something like $15 million a year in a long term deal.

But he’s not healthy, and so he’s not going to get anything close to that. It’d be shocking if he got anything beyond a one year offer with a team option for 2010. But if he can stay on the field, even for 100-120 games a year, he’s got the abilities to play at an all-star level. Right now, he’s Milton Bradley with less attitude problems, and there are probably quite a few teams who wish they would have taken a chance on Bradley last winter.

It all depends on his health, because there shouldn’t be too many questions about Joe Crede’s abilities to help a winning team when he’s on the field.

Today, we take a look at the third undervalued free agent (my bad, he’s signed for ’09) this winter, who will be able to help a team win some games without requiring significant dollars or a long term commitment. The next name on the list is Endy Chavez, the polar opposite of the outfielder we profiled yesterday.

See, Chavez isn’t much of a hitter. He hit .267/.308/.330 last year, and for his career, he’s got a .680 OPS and a -4.95 WPA/LI in 2,274 PA. That makes him about one win below an average hitter per full season. Considering he’s already 30, it’s unlikely he’ll be improving with the bat much. He is what he is – a slap hitter without much power who doesn’t draw walks.

But he also might just be the best defensive outfielder alive today. He’s certainly in the discussion, at least. John Dewan’s +/- system ranked him as +19 during his time between LF and RF this year – in 600 innings! In what accounts to about a half season of playing time, Dewan’s system judged Chavez to be almost 20 plays better than an average defensive corner outfielder who plays an entire season. That’s just a huge, huge difference.

The reason, of course, is that Chavez isn’t really a corner outfielder. He’s a CF who has spent time in the corners because of Carlos Beltran’s presence on the Mets roster, and on pretty much any other team in baseball, Chavez is a CF. In fact, over the last three years, Chavez has accumulated about 360 innings in CF, and +/- gives him a +10 ranking in that time. +10 plays compared to other center fielders in about 1/4 of a season’s worth of playing time.

It doesn’t matter which system you look at – the results are all the same. Chavez is projected to be +10 to +20 runs better than an average CF over a full season, and something like +25 to +30 runs over a corner outfielder given regular playing time. The guy can cover ridiculous amounts of ground.

That defense counteracts all kinds of bad hitting. If a team sees him as a CF, and gives him 600 plate appearances, he’ll probably be a -1.5 win hitter, a +3 win defender, and a +0.5 win baserunner. Add it all up, and you’ve got a +2 win player, or approximately a league average center fielder.

Endy Chavez – league average player at a premium position. You won’t find too many teams who think he’s that good, but the defense is more valuable than almost all of them realize. He could hit .230 and still be a useful major league player, and right now, he’s established a track record of good enough hitting to be worth an everyday job. Look for some smart team this winter to snatch him up (in trade!) and turn him into the outfield version of Adam Everett.

This afternoon, we looked at the first of several potential free agent bargains this winter – Jeremy Affeldt. This afternoon, we take a look at another guy with some potential for positive reward who won’t cost an arm and a leg. That guy is Juan Rivera.

Two years ago, Rivera hit .310/.362/.525 for the Angels, racking up a 2.45 WPA/LI mark that made him a very solid contributor to Anaheim’s offense. Then, injuries struck, and the last two years have been something of a wash. He managed just 324 plate appearances between the two seasons, and the missed time cost him his shot at a regular gig in LA. Now 30 and hitting free agency, it’s unlikely that a team is going to be giving him a multi-year contract without proving he can play everyday again, but if we look at his skills, he’s still a pretty solid hitter.

Rivera’s signature skill has always been his power, and that hasn’t disappeared. Even with the health problems and limited playing time, he posted a .191 ISO last year. When you look a little deeper, there’s even more reasons for optimism. During the first three months of the season, Rivera managed all of 66 plate appearances in 30 games. He barely made it on the field from April through June, and he wasn’t giving the Angels any reason to put him in the line-up more often, struggling to a .177/.227/.226 line. He had three extra base hits, all doubles, and simply wasn’t hitting. However, Gary Matthews Jr’s struggles created an opportunity for him to play a bit more often, from July on, he showed that the old Juan Rivera still existed.

In the final three months of the season, Rivera hit .268/.299/.505, and while the OBP isn’t exactly impressive, the 22 extra base hits (12 of which were home runs) show that Rivera’s still got some juice in his swing. His overly aggressive approach at the plate will always make him a low on base guy, but that kind of power is still valuable. Even with the .299 OBP, Rivera’s WPA/LI over the final three months was -.07, making him essentially a league average hitter.

Now, if that was Rivera’s ceiling, he’d be a nifty platoon corner OF/DH type, and that would be the end of it. But Rivera was a league average hitter while hitting .270. As a guy with both good contact and power, Rivera’s got the kind of skills that could allow him to hit .290 to .310. If you don’t strike out, and you hit the ball over the wall with some frequency, it’s pretty hard to post a low batting average, thanks to the sheer quantity of chances you have for hits. Rivera managed to hit .269 over the final three months despite a .242 batting average on balls in play. That’s extraordinarily low, even for a guy hitting a ton of flyballs.

If we assume that Rivera’s true talent BABIP is more in the .280 range (it’s .292 for his career, but we’ll knock it down a bit for age and injuries), and the rest of his skills remain in tact, he’s a .290 to .300 hitter. Even with his aggressive approach, .300/.340/.500 isn’t out of the question.

Rivera may never get back to his 2006 prime, but he’s certainly better than he’s shown the last two years, and there’s no reason to think he’s washed up at age 30. For a team looking for a right-handed power bat who makes good contact and offers some upside without requiring a long term, big money deal, Rivera’s a good bet. He’ll never be a star, but he could be the kind of useful role player that is picked up cheaply that championship teams need.

As we head towards free agency, the focus is always on the big name guys – who is going to sign Mark Teixeira or CC Sabathia? Which team is going to give Manny Ramirez a deal that pays him into his 40s? What kind of contracts will older-but-still-effective guys like Randy Johnson and Jamie Moyer get?

But every year, there’s always a guy signs as something of an afterthought and turns into an important player, with the rest of baseball looking back and wondering why they weren’t more interested when he was available. This week, we’ll profile a group of players that should pay nice dividends for the team that recognizes that they’re worth more than they’re going to get this winter.

Starting off, let’s take a look at one of the more underrated pitchers in the game – Jeremy Affeldt.

A former top prospect as a starter, Affeldt didn’t live up to expectations in Kansas City and has bounced around the bullpens of the Royals, Rockies, and in 2008, the Reds. His career has seen him struggle with command problems, and as such, he’d never been trusted with any kind of meaningful role.

However, in 2008, he took a pretty big step forward. His average fastball velocity went from 92.3 MPH a year ago to 94.6 MPH, and his CB saw a similar improvement (going from 76.9 to 79.1). Adding several ticks to both pitches moves him into the power pitcher category, and it showed in his results.

His swinging strike rate (data from statcorner.com) jumped from 7.6% last year to 10.8% this year, putting him in the company of guys like Grant Balfour, Joakim Soria, and Kerry Wood. The swinging strikes translated directly to his strikeout rate, which jumped to 9.19 K/9. More swings and misses means less called balls as well, and Affeldt significantly improved his walk rate as well, down to a respectable 2.87 BB/9.

In a lot of cases, a pitcher will add velocity by switching to a four seam fastball and pitching up in the zone, but that’s not the case here – Affeldt continued to pitch down in the zone, posting a career high 54.4% GB%. Because he keeps the ball down, he was able to sustain a fairly low HR rate even though his HR/FB went through the roof (as it can with every reliever, thanks to the small yearly sample of batters and its variance).

Based on 2008, Affeldt has shown an entirely new skillset as compared to his major league career, adding a couple of ticks to both his fastball and curve, which have transformed him into one of the game’s best left-handed relievers. Because his curve is a legitimate second pitch, he shows almost no platoon split, and is equally effective against both LHB and RHBs. His stuff is clearly closer-worthy, and while he has no experience as the 9th inning guy, some smart team would be wise to give him a two or three year contract for setup man money and watch him blossom into one of the game’s best relief aces.

There’s a good case to be made that Jeremy Affeldt is the best reliever available this winter, while most of baseball has no idea who he is.

As Eric and I have been illustrating over the last few days, there’s some really cool stuff in the new team pages that you guys should be checking out. Here’s one more example – a graph of all team’s Clutch scores, which measure the portion of their WPA that came from their performance in high leverage situations. Essentially, this is how team’s performed when it mattered.

Notice how most teams aren’t that far from zero? 23 of the 30 teams had scores between +3 and -3, meaning that they won three more or less games thanks to their high leverage hitting. It helps, but clearly, it’s not the thing that makes or breaks a ballclub.

But look at those two big bars on the far left hand side, especially the biggest one – that’s a clutch score of +7.32. That’s the 2008 Los Angeles Angels of Anaheim, who added almost seven and a half wins to their total by performing well in high leverage situations. That’s 2.6 standard deviations from the mean – a true outlier if ever there was one. A huge portion of Anaheim’s success in 2008 was directly due to clutch hitting.

Guess what – they led the league in 2007, too, posting a +5.19 clutch score. That’s 12.5 wins in two years thanks to timely hitting. Yes, it’s probably random variation, but it’s interesting random variation.

Just for kicks, here are their clutch scores and MLB ranks for the entire decade, which happens to be the length of Mike Scioscia’s managerial reign.

2000: -2.08 (21st)
2001: -5.41 (25th)
2002: 4.54 (3rd)
2003: 0.13 (11th)
2004: 1.33 (6th)
2005: 4.28 (2nd)
2006: -1.85 (21st)
2007: 5.19 (1st)
2008: 7.32 (1st)

This probably doesn’t mean anything, and I certainly wouldn’t advise betting on the Angels to be among the league leaders in clutch hitting again next year. But a +20 clutch score over the last seven years sure is eye opening. If you keep looking at those rosters and trying to figure out how they won so many games, here you go.

When you think of the standard cliches about “Moneyball” teams that have hired statistically inclined GMs over the last few years, what’s one of the first one’s that comes to mind? Patient hitters who draw a lot of walks has to be at the top of the list, right?

Well, with the new team pages here on Fangraphs, I decided to look at the teams rankings in our Plate Discipline stats, just to see if the stereotype holds. I figured the easiest way to judge approach was by O-Swing%, or the amount of times a team swings at pitches outside the strike zone, since laying off bad pitches is one of the fundamentals of the “sabermetric hitting approach”. The results were… well, just look for yourself. Sorted by lowest O-Swing%:

1. Tampa, 22.63%
2. Boston, 22.85%
3. Oakland, 23.88%
4. Cleveland, 23.25%
5. Arizona, 23.50%
6. Yankees, 23.62%
7. Toronto, 23.64%
8. Colorado, 23.81%

and the bottom eight:

23. San Diego, 26.83%
24. White Sox, 27.44%
25. Dodgers, 27.66%
26. Kansas City, 27.88%
27. Houston, 27.97%
28. San Francisco, 28.02%
29. Seattle, 28.49%
30. Anaheim, 28.87%

Among the top eight, the only team that you could argue doesn’t employ a sabermetric GM would be the Yankees, but they’ve still shown a clear affinity over the years for patient hitters. In fact, the five teams with the lowest O-Swing% are basically The Big Five stat-friendly organizations in baseball right now. Likewise, the five teams with the lowest O-Swing% are five franchises that have essentially looked at statistical analysis and said “nah, that’s not for us.”

The correlation between a team’s statistical bent and the players rate of chasing pitches out of the strike zone is pretty strong. Now, some of that is due at least in part to the selection processes of the players by the GM (Oakland is way more likely to give a job to Jack Cust than Seattle), but teams like Tampa and Arizona are mostly built upon position players developed from within. It seems clear that the emphasis of plate discipline as an approach can manifest itself in on field results. Perhaps the Royals, Astros, Giants, Mariners, and Angels might want to consider preaching a bit more patience at the plate – it’s not like their offenses are much to write home about anyway.