SubscribeLast Player Picked is a site exploring various aspects of fantasy baseball analysis, especially relating to drafting and player valuation strategies.
Much of what is discussed here will involve the Price Guide, an online tool I developed to build dollar values or player rankings customized for any league configuration.
Last year, I highlighted 17 pitchers who showed signs of becoming very good pitchers in 2009: They struck guys out and limited the walks and homeruns. Despite these indicators of success, they were still priced cheaply enough to provide a good fantasy return on investment.
As we saw last week, 12 of those 17 pitchers improved on their 2008 marks in 2009. Felix Hernandez, Adam Wainwright, Matt Cain, Randy Wolf, and Jon Lester surged to greatness. Even players who the projections panned (John Danks, Wandy Rodriguez, Jorge De La Rosa) made a fantasy impact in 2009.
So let’s try the same thing for 2010. If 2009 is any guide, these are players whose skills may be a better indicator of future fantasy success than the projections on the Price Guide.
We’ll first eliminate the top tier of pitchers who meet the LIMA qualifications — LIMA is about building a cheap staff and Roy Halladay and Zack Greinke aren’t coming cheap. It’s interesting that several names from last year’s list (the aforementioned Hernandez, Wainwright, Lester, and Cain) have graduated into the top fantasy tier.
We’re left with the following 18 pitchers with K/BB >= 2, K/9 >= 6, HR/9 <= 1, and IP >= 100 in 2009:
Very Good and Probably Not Cheap
Josh Johnson
Jair Jurrjens
Wandy Rodriguez
Clayton Kershaw
Tommy Hanson
Known Commodities Off a Down Year
Roy Oswalt
Chad Billingsley
Jake Peavy
John Lackey
True LIMA Pitchers
Gavin Floyd
Ryan Dempster
Kevin Correia
Hiroki Kuroda
Jeff Niemann
Colorado Rockies
Ubaldo Jimenez
Jorge De La Rosa
Jose Contreras
Jason Hammel
(That’s right, the Colorado Rockies get their own tier of LIMA pitchers. Despite solid peripheral stats, the projections uniformly hate them all, presumably because of their home ballpark. Of course, the projections already missed once on De La Rosa last year, so take that how you will.)
We’ll break down those categories one at a time throughout the coming week.
Yesterday, I asked if there were any rule changes that would make for a more realistic fantasy game without compromising fun.
For example, many leagues have tried adding extra categories — holds, errors, etc. — in an attempt to create a more realistic game. I’d argue that part of what makes fantasy fun is its simplicity, and going to 6×6 or 7×7 creates a more complicated game. Plus, going beyond 4×4 or 5×5 has a chance to completely alter the game’s balance.
I have one easy proposal for fantasy leagues, one simple change that makes the game more realistic, but does not change the balance of the categories or the ease of gameplay: Replace AVG with OBP.
In real baseball, walks are valuable. It doesn’t matter how a player gets on base, whether he gets a or a walk or is hit by pitch or reaches on an error. A walk is essentially worth the same as a single, but fantasy in its current forms completely misses the value of walks.
Since on-base percentage is functionally equivalent to batting average, I think the best thing to do is to replace BA with OBP. That lets the league keep the same balance of categories. But now your categories are just a little bit more better at matching values in real life.
Right now, I play in some some traditional leagues, but I also play in leagues that have made the switch to OBP. I can definitely say it’s the best change we have made. It’s small enough not to completely distort the game everyone is used to, but it also brings the fantasy game one step closer to the real thing.
With any game — not just fantasy — there is a precarious balance between being fun and being realistic.
Think about sports video games: Most games have two settings — the actual game and an “arcade” style game. The arcade game is much simpler — the plays are standardized, the buttons are limited. If you’ve never played the game before, you are much more likely to have fun with the arcade style, just because it is more straightforward. The trade-off, of course, is that you sacrifice some realism from the main game to make it easier to play.
With fantasy sports, I think in general there is a tendency to want to make things more realistic. Leagues are always wanting to add more categories because there is some element of real baseball that is not being perfectly modeled by the fantasy game. We know that real teams roster middle-relievers and not just closers, so we add holds. We know we are leaving out defense…maybe adding errors will help?
Or maybe we just need one stat category: How about just using WAR for both hitters and pitchers? Using that one stat is probably a more realistic reflection of value than any combination of others.
The downside of those more realistic 9×9 and 10×10 leagues is that you lose the simplicity. And the downside of the uber-stat like WAR is that you lose the fun. Fantasy is about balancing realism and fun.
Fantasy baseball in its most common forms (4×4 and 5×5) is a pretty fun game. It also does a decent job of reflecting reality: Albert Pujols and Joe Mauer are valuable both in real life and in fantasy. Carlos Silva is worthless in both.
But I’m curious: Are there any rule changes that can improve realism without detracting from the simplicity of the current game? Are there any changes that would make fantasy baseball more fun without compromising its realism?
Yesterday, I went through the results of how last year’s LIMA pitchers did relative to expectations. (In case you aren’t familiar with it, LIMA is a strategy for drafting bargain starting pitchers with good strikeout, walk, and homeruns rates. Getting cheap — but potentially very good — pitching lets you focus most of your draft attention on loading up on hitting.)
On the whole, the LIMA pitchers did extremely well. Players like Felix Hernandez, Jon Lester, and Matt Cain took a huge step forward to become elite pitchers. Guys like Jonathan Sanchez and and Randy Wolf went from marginal to good. And, of course, a couple of guys (like Joba Chamberlain and Gil Meche) flopped completely.
So what did we learn from the whole thing? Here are a few lessons that I took from these results:
LIMA isn’t dead, yet.
As I’m re-reading my thoughts from last year, it’s clear that I was skeptical of the value of LIMA in the 21st century. LIMA originated in a world of 4×4 fantasy leagues, and today’s 5×5 dominated fantasy scene already appreciates high-strikeout pitchers. LIMA doesn’t know about BABIP or GB/FB rates that can also be indicators of future success. Projections are designed to consider strikeout, walk, and homerun rates and so the elements of LIMA should already be present in the dollar values.
But how do you explain two-thirds of LIMA pitchers outperforming their projections? Now, 12 out of 17 isn’t a great sample size, but that does seem to be a tangible effect. These pitchers went from a $4 average value in 2008 to an $11 value in 2009 (higher even than the $7 average projection).
A strategy that can identify a class of pitchers that will beat their projections by an average of $4 is a winning strategy. The overperformance of these starters certainly makes me inclined to continue bidding even past the projected dollar amount on similar pitchers in 2010.
Watch out for players on the border.
After a couple of years of striking out around 5 batters per nine, Paul Maholm edged into the LIMA qualifications for 2009 with a K/9 of just over 6 in the previous year. Of the players I looked at, he had the lowest strikeout rate and was in the bottom half for K/BB and HR/9.
In 2009, he promptly returned to his 5 K/9 ways. Hindsight tells us that Maholm was really the mediocre pitcher we saw in 2006-2008 and not the one who snuck into the LIMA qualifications in 2009. Apparently borderline LIMA players are not the best bets for success.
Of course, removing the bottom tier of LIMA qualifiers might have also dumped Wainwright (6.20 K/9) and Garza (6.23 K/9), both of whom improved in 2009, so this rule doesn’t work every time.
Watch out for players switching roles.
In 2008, Joba Chamberlain topped all of the LIMA categories, leading LIMA-inventor Ron Shandler to peg Joba as the #1 pitcher for 2009. The projections loved him as well, expecting him to double his value from the year before.
However, far from being the best fantasy pitcher of 2009, Chamberlain ended up in negative territory. What happened?
Most of the decline may have just been injuries or simple bad luck. But Joba’s failure also serves as a reminder that pitching relief is easier than starting and facing the same lineup multiple times a night. A RP who has LIMA skills may not be a LIMA starter.
We’ll keep those lessons in mind as we look at who fits the LIMA strategy for 2010.
Last year, I did a series of articles looking at pitchers who met the qualifications for the LIMA Plan. For those who may be unfamiliar with the strategy, LIMA stands for Low Investment Mound Aces, and it’s a strategy to get SP with good skills on the cheap so that you can spend more on hitters. You target cheap players who have a K/BB over 2.00, a K/9 above 6.00, and a HR/9 below 1.00.
After throwing out the obviously good pitchers (Lincecum, Halladay, Sabathia, Santana, etc.) who wouldn’t come cheap, I found 17 LIMA pitchers for 2009, and I broke them down into categories based on the 2009 projections. Here they are again, with their 2008 actual values, 2009 projected values, and 2009 actual values:
Already Good, Could Be Great
| Player | 08 Act. | 09 Proj | 09 Act |
|---|---|---|---|
| Josh Beckett | $10 | $22 | $18 |
| Matt Cain | $2 | $18 | $21 |
| A.J. Burnett | $9 | $19 | $6 |
| Joba Chamberlain | $8 | $17 | -$5 |
| Felix Hernandez | $4 | $16 | $32 |
True LIMA Pitchers Projected to Improve
| Player | 08 Act. | 09 Proj | 09 Act |
|---|---|---|---|
| Jonathan Sanchez | -$6 | $3 | $4 |
| Kevin Millwood | -$14 | -$4 | $7 |
| Andy Pettitte | -$2 | $3 | $5 |
LIMA Pitchers with No Projected Improvement
| Player | 08 Act. | 09 Proj | 09 Act |
|---|---|---|---|
| Adam Wainwright | $10 | $9 | $28 |
| Matt Garza | $7 | $8 | $9 |
| Gil Meche | $6 | $5 | -$8 |
| Paul Maholm | $4 | $1 | -$4 |
| Randy Wolf | $0 | -$2 | $20 |
| Wandy Rodriguez | $6 | $2 | $19 |
| Jorge De La Rosa | -$4 | -$7 | $8 |
LIMA Pitchers Projected to do Worse
| Player | 08 Act. | 09 Proj | 09 Act |
|---|---|---|---|
| Jon Lester | $13 | $2 | $20 |
| John Danks | $13 | $1 | $10 |
With hindsight, we might split them up like this, instead:
Better than Expected
| Player | 09 Proj | 09 Act |
|---|---|---|
| Randy Wolf | -$2 | $20 |
| Adam Wainwright | $9 | $28 |
| Jon Lester | $2 | $20 |
| Wandy Rodriguez | $2 | $19 |
| Felix Hernandez | $16 | $32 |
| Jorge De La Rosa | -$7 | $8 |
| Kevin Millwood | -$4 | $7 |
| John Danks | $1 | $10 |
Met Expectations
| Player | 09 Proj | 09 Act |
|---|---|---|
| Matt Cain | $18 | $21 |
| Andy Pettitte | $3 | $5 |
| Matt Garza | $8 | $9 |
| Jonathan Sanchez | $3 | $4 |
| Josh Beckett | $22 | $18 |
Worse than Expected
| Player | 09 Proj | 09 Act |
|---|---|---|
| Paul Maholm | $1 | -$4 |
| Gil Meche | $5 | -$8 |
| A.J. Burnett | $19 | $6 |
| Joba Chamberlain | $17 | -$5 |
On the whole, that seems like a very good performance. Almost half of the pitchers did significantly better than projected. Another five met their projections, which was still an improvement over what they had done in 2008. Only 5 out of 17 players did worse than projected, and really only 3 (Meche, Burnett, and Chamberlain) would count as disappointments.
Those results are both surprising and pleasing. I really expected the projections to recognize that pitchers with solid peripherals would improve, but they still undersold most of these guys. And it’s exciting to think of what can be applied for fantasy drafts in 2010.
Tommorrow: What lessons can we learn from 2009?
Did you know that you can add the Price Guide’s dollar values to your fantasy league’s website (Yahoo, ESPN, CBS Sportsline, or Allstar Stats)? Here’s what it looks like on a public CBS league:
These are dollar values that are customized to whatever settings your league is using and that only show up on your computer. During the preseason, you can use the projected stats on this site or your own customized projections. Once the season starts, you might instead show each player’s actual, year-to-date dollar value.
The dollar values are easy to add. Since CBS already has their leagues up, here is all you need to do to add them:
1. Download Mozilla Firefox and Greasemonkey. Firefox is an internet browser that works as an alternative to Internet Explorer. Greasemonkey is an add-on to Firefox that allows you to customize how pages appear.
2. Enter your CBS Sports league settings into the Price Guide. For a public league, you should only need to change the number of teams from 12 to 10. If you use “Standard Roto” (i.e. 4×4) rules, you will also want to remove batter runs and pitcher strikeouts. (If you have a Premium league at CBS, you may need to change more settings to match your league.)
3. Click the link on the results. At the top of the Price Guide results, you should see a link to add dollar values to a CBS Sportsline league. Click the link and Greasemonkey should recognize and install the script.
4. Check it out on your league homepage. Go to your league, and you should notice the dollar values showing up next to each player’s name.
While CBS provides player rankings that give you an idea of a player’s value, remember that these are not customized for your specific league. Also, it might be valuable to know places where the Price Guide’s projections diverge from CBS’s. Compare CBS’s 2B rankings to what the Price Guide thinks:
While much of those rankings line up, the Price Guide’s projections put much less weight in Aaron Hill’s and Ben Zobrist’s remarkable 2009 improvements, expecting them to settle somewhere in between 2008 and 2009.
At 3B it is clear that the Price Guide continues to put less emphasis on one year’s stats: The projections don’t discount David Wright’s mysteriously poor 2009 as much as CBS does, placing him at $28 and the 2nd best 3B.
While you certainly don’t have to agree with one or the other, I think those kinds of discrepancies can be valuable to realize. Also remember that these values will update as the Price Guide updates this spring. During the year, they will update daily based on the latest year-to-date stats.
By my calculations, this week’s rush of free agent signings (Sheets, Nady, Thome, Winn, Garland, etc.) has left only five free agents with positive fantasy value.
These are players who have two main questions marks surrounding them for 2010: The first question is what their situation will be like on the team where they sign. Will they be an everyday starter or just a bat off the bench? And the other concern is age. Most of these guys are well into their 30’s, and some decline is expected. There’s a possibility, however, that they’re in for a nosedive instead of a gentle decline.
Both of those risk factors make these five players worth watching as the season gets closer:
Johnny Damon ($14)
Now that it looks like the Yankees have moved on without him, I’m very concerned about Damon’s fantasy value. He managed a revival last year launching homeruns into the short RF porch at New Yankee Stadium, so leaving New York looks like bad news for him.
(CHONE and CAIRO both try to adjust for park, but Yankee Stadium as a whole was not that favorable for homeruns. Damon, however, seemed to particularly exploit its dimensions in a way that aren’t accounted for in the park factors. I think the HR’s projected are way too optimistic.)
Then there’s age: Damon is 36 years old. He’s a lock to sit out 15-20 games. Basically, I’m not betting that Damon manages to repeat anything like last year, and I’m not going anywhere near the $14 the Price Guide suggests for him.
Erik Bedard ($12)
The Price Guide has an unhealthy infatuation with pitchers who don’t pitch many innings, so of course Bedard with his projected 106 IP is a Price Guide favorite.
He is projected to strike out over 9 batters per nine with a solid ERA and WHIP, so there’s definitely something to like here. If he puts together most of a full season in 2010 he’ll be a steal at $12.
Jermaine Dye ($10)
Somehow Jermaine Dye is a player who had completely fallen off of my radar: I never would have guessed he was projected as a double-digit player, and I was even a little surprised that he was valued above replacement ($7) last year.
I think I must have been burned on him in 2007, when he followed up a 44 HR, .315 BA year in 2006 with 28 HR and a .254 average. Since then, he’s maintained his 25+ HR power, but the average has varied tremendously.
The Reds look like they could be a good fit for Dye, if he’s still up for playing OF everyday, and Great American Ballpark seems like a good fantasy location. I’m not finding a whole lot of teams that have a regular spot open for an OF/DH, though.
Orlando Cabrera ($10)
Cabrera has seemingly been a fantasy mediocrity for years. I think part of that reputation is that he doesn’t stand out in any category: He hits a few homeruns, but not too many. He steals a handful of bags, but is never among the league leaders. His batting average is decent — not a killer but nothing special, either.
If he can find an everyday job, then I expect Cabrera will put up another unremarkable (but still valuable) fantasy season.
Felipe Lopez ($6)
Felipe is looking for potentially his sixth team in five years. Despite being continually passed around by real teams, he’s the kind of player who can provide a cheap boost for a fantasy team if you stick with him.
In previous years, Lopez has added some fantasy value by qualifying all over the diamond — playing some 2B, 3B, SS, and OF. This year he will initially only qualify at 2B, although (depending on where he signs) he may gain another position quickly.
A few other names worth keeping an eye on:
Darren Oliver
Orlando Hudson
Jarrod Washburn
Garret Anderson
Kevin Gregg
John Smoltz
Yesterday, I asked what you would be willing to pay for Stephen Strasburg if your draft were held today. Reassuringly, the comments I received echoed what I had been thinking, and they basically raised the same questions that I have.
The projections so far (CAIRO and ZiPS) have him as a late-round 4.00+ ERA pitcher for about half of a season. That would probably be worth a mid-single digit bid in most leagues.
But there are enough questions around Strasburg to make bidding much more complicated. I see at the following factors weighing in on Strasburg’s value:
How many innings will Strasburg pitch in 2010?
I don’t follow the Nationals too closely, but right now it looks like there’s at least a chance Strasburg could start the season in the majors. There aren’t a ton of great options available for Washington, so a good spring gives Strasburg a shot at spending April in D.C.
However, it has become common practice to cap a young player’s innings to prevent injuries, and so I’d say the most optimistic prediction for Strasburg would be 160 IP. The Nationals won’t have anything to play for in September, so I’d think they would shut him down.
Of course, the Nationals may also decide to start Strasburg off in the minors. In that case, we’re talking about something closer to 100 IP.
I’d say the risk of only getting 100-150 IP out of a SP is a factor that keeps Strasburg’s value low for 2010.
How high is the upside?
I mentioned yesterday that projecting Strasburg in 2010 is more complex than just picking a single number, like a 4.38 ERA. I’m making up some numbers here, but let’s throw out some quick probabilities for what Strasburg’s ERA might look like:
5% – 3.00
10% – 3.50
25% – 4.00
40% – 4.50
20% – 5.00
That gives, on average, a 4.30 ERA. It also gives Strasburg a 5% chance of becoming a top-tier pitcher immediately. There’s a 10% chance he winds up in the middle-tier for fantasy, and a 25% chance he’s worth a late-round grab. Most (60%) of the time, however, he’s waiver wire fodder.
I admittedly have no idea how realistic those probabilities are. However, I think the chance of him being either good or very good gives him a much higher ceiling than your typical 4.38 ERA pitcher. And I think that upside must slant his value a little bit higher.
How much risk can your fantasy team assume?
When drafting, I try to balance out upside picks with more conservative choices. If I already have, say, Roy Oswalt and Roy Halladay, then I have quite a bit of freedom to take some chances with the rest of my pitching staff.
The answer to this question will be mostly dependent on how your draft is progressing when Strasburg comes up. But I think it also matters regardless of context specifics. In general, I think pitching depth looks pretty good this year. It seems like it will be possible to put together a solid, reliable staff without spending too much money. The availability of quality SP might make me somewhat more inclined to take on a riskier proposition.
Put all of those factors together, and I’m guessing that I’d be willing to bid about $8 on Strasburg in a typical league. That’s not enough to ruin my season if things go wrong, but I think there’s a decent chance he’s worth it.
CAIRO’s latest update gives the Price Guide its first projection for the Washington Nationals’ Stephen Strasburg:
4 W, 4.38 ERA, 1.47 WHIP, 70 K, 76 IP
SG is going from college stats here, so this is still wildly speculative. The Price Guide indicates that a pitcher with those stats isn’t worth starting in most fantasy leagues, including NL-only leagues. It would be the kind of pitcher you draft for your bench, only to drop in April so you can pick up some guy who starts the season on an unexpected hot streak.
Dan Szymborski’s ZiPS gives a somewhat more optimistic look for Strasburg:
10 W, 4.18 ERA, 1.32 WHIP, 113 K, 114 IP
That’s a draftable pitcher for late in a fantasy baseball draft. The IP make it hard to find a comparable, but that’s roughly what you might expect from Rich Harden or Chris Young. (They have injury risk, but Strasburg has risk as well.) It’s also not too far off from what you might expect from A.J. Burnett (with significantly fewer IP).
With Strasburg, of course, you’re not simply drafting a projected statline. There’s an enormous amount of upside, much more so than your typical 4.38 or 4.18 ERA pitcher. That upside is obvious for those in keeper leagues, but is it worth anything in 2010?
What do you think? Let’s say you’re in a 12 team, standard redraft league and have $260 to spend. If your draft were held today, what would you pay for Stephen Strasburg?
And you can’t say, “I wouldn’t draft him because someone else would pay more.” How cheap would he have to come for you to bid on him?
I’ll give my answer tomorrow.
Here are CHONE’s 2010 projections for three very similar-looking starting pitchers:
Pitcher A – 3.96 ERA, 1.20 WHIP, 150 K, 168 IP
Pitcher B – 3.89 ERA, 1.23 WHIP, 165 K, 183 IP
Pitcher C – 3.85 ERA, 1.23 WHIP, 174 K, 187 IP
Pitcher C is World Series hero and fantasy mainstay Josh Beckett. Pitcher B is two-time Cy Young winner Johan Santana. Pitcher A, of course, is Colby Lewis.
Colby Lewis? Really?
Lewis has spent the last two years playing for the Hiroshima Toyo Carp in Japan’s NPB league. He was recently inked by the Texas Rangers to fill out their rotation. Before his time in NPB, he was a top prospect for Texas who managed some unimpressive MLB stats from 2004-2007. In Japan, however, he became a new pitcher:
2008 – 15 W, 2.68 ERA, 1.00 WHIP, 183 K, 173 IP
2009 – 11 W, 2.96 ERA, 0.92 WHIP, 186 K, 176.3 IP
CHONE takes those awesome Japanese stats and tempers them considerably; the sub-3.00 ERA and sub-1.00 WHIP are lost in the translation. Even taking quite a bit off of his NPB stats, Lewis looks like a very good player in MLB.
Before you get too excited abut Colby Lewis, though, I’ll make a few cautionary observations:
1. CHONE hates Santana as much as it loves Lewis.
The comparison of Colby Lewis to Johan Santana is not just a statement about Lewis, but also about Santana. CHONE is projecting a 3.89 ERA for Santana, and he hasn’t topped 3.33 in the past eight years.
Apparently there needs to be a separate post to examine what to expect from Johan Santana in 2010. So that comparison is more for shock-value than for balanced analysis. Guys like Matt Garza and James Shields are the sorts of guys we really expect to have a 3.90 ERA with lots of strikeouts, and they are probably a better benchmark for Lewis.
2. Not all projections love Colby Lewis.
ZiPS isn’t on this site for 2010, but it is considerably less optimistic for Lewis:
11 W, 4.39 ERA, 1.30 WHIP, 128 K, 176.3 IP
That’s still above replacement level, on par with end-game picks like Gavin Floyd and Andy Pettitte. ZiPS appears to be putting less weight on his Japanese stats and more on his undistinguished MLB-career.
This raises an important question: Which is more indicative — recent stats in a non-MLB context or MLB stats from three or four years ago? Colby Lewis in 2010 could be an important data point in this discussion.
3. Translating Japanese stats is an imperfect exercise.
While MLB-NPB translations are based on how the switch impacted other players, it is still hit-or-miss. The Japanese game is simply different than what is played in North America, and skills that were valuable in one league might not help in the other.
However, the track record for CHONE has been pretty good on recent imports Hiroki Kuroda, Kenshin Kawakami, and Koji Uehara. These were pitchers who looked like decent-not-great fantasy pitchers based on the projections, and who basically lived up to those expectations. The fact that a 2.90 pitcher like Lewis would be a 3.90 pitcher in MLB seems to indicate that the translation is being fairly conservative.
The bottom line: Lewis could come very cheapily in 2010 drafts, as long as the hype on him doesn’t build. Magazines that have an early print date probably won’t be aware of him, and that means that there are probably several people in your league who won’t be aware of him, either. There’s also no $51 million posting fee to create the buzz like what surrounded Daisuke Matsuzaka.
Even using the more pessimistic ZiPS, I feel pretty good about paying at least $5-6 for Colby Lewis. At that price, it’s not a big deal to cut him if he doesn’t work out. If it turns out CHONE is right, then that’s a cheap bet that could pay huge dividends.