A couple of Howardian thoughts
Wow, this is fun. Lots of analytic discussion about the Phillies’ contract extension with Ryan Howard, ranging from Fangraphs’ Matthew Carruth saying “…I was tempted to fill this entire article with just me laughing” to BPro’s Matt Swartz saying “For Phillies phans, this is actually a great sign.” Two very analytic guys with two very different conclusions.
Because Matt Swartz’s article requires a BPro subscription, let me recap his very thorough analysis:
{exp:list_maker}The starting point for free agent salaries is about $5 million per win above replacement (something I intuitively agree with). Other folks appear to be using a figure close to $4 million.
Salary inflation may be higher over the next six years than it’s been in the recent past (probably true, but hard to predict).
The Phillies know Howard better than we do, and have a better feel for how he’s likely to age (a good point, but nothing I’d hang my hat on). {/exp:list_maker}By the way, Matt’s work on free agents who stay with their team vs. those who are signed by other teams is top-notch, though I think he’s overdrawn the conclusion in this case.
Still, I appreciate the fact that Matt has come to a different conclusion than the rest of the screamin’ blogosphere. He’s provided something to ponder.
Matt uses WARP3, other people use WAR. When I used to analyze salaries, I used Win Shares Above Bench. Honestly, I don’t think the exact system matters a whole lot. Back in 2007, I found that teams were paying free agents $5.3 million per win above replacement. Even with low inflation, I’ll bet the figure is at least $5.5 million now.
So, if you assume $5.5 million per win in 2010 and a salary inflation rate of five percent, at what level of production will Howard’s contract break even? The answer is that he needs to average between three and four Wins Above Replacement per year for the length of the contract for the Phils to break even. Last year, he produced 4.9 WAR. Can he average three-to-four WAR over the next six years of his career? To me, that is the fundamental question. Personally, I’d put the probability at less than 50%, higher than 25%. I have nothing to back that up, other than my prodigious gut. And I don’t mean prodigious in a good way.
Remember the Paul Konerko brouhaha in 2006? Konerko was a free agent after the White Sox won the World Series. The Angels were going to sign him, but the Sox stepped in at the last minute and signed him to a five-year deal worth $60 million. Konerko was 30 at the time, the same age Howard is now. Konerko was coming off a 4.3 WAR year, though he had averaged between two and three a year before that. A lot of us didn’t like the deal at the time because Konerko was kind of like, well, Ryan Howard today—just not so much. Didn’t seem like he was going to age well. We understood why the Sox did it, however. Something about keeping the fan base engaged. Sound familiar?
Konerko has averaged about 2.4 WAR each year since and he’s off to a good start in this, the final year of his deal. At roughly $5 million per free agent win, the deal has been a solid, breakeven kind of deal. Does that mean the Howard deal is likely to turn out well? Not at all, but it’s not out of the question.
By the way, I should add that I think Josh Leavitt’s post from earlier today is critical, too.
Rally has Konerko at 2.4, 1.5, 0.5, and 1.8 WAR over the last 4 years, despite having him as an essentially average defender. He gets dinged on the little things- baserunning, reaching on errors and grounding into double plays. Rally has Howard at 2.4, 5.8, 2.6, 2.8 and 4.8 over his career so far, despite a positive TZ for his career.
You are using Fangraphs’ figures which peg a higher replacement level and use a different run estimator.
So, in order for the contract to be halfway reasonable, Howard has to beat his 50th percentile and you have to prefer the Fangraphs’ figures.
I’m glad Matt was able to provide a coherent argument for the deal. Most of the articles criticizing the deal were written with the assumption that the Phillies were just being colossally stupid, and I honestly have no idea how someone can be that confident in challenging an extremely successful franchise. It’s nice to see Matt thinking from the Phillies point of view and trying to justify the deal, even if, as you point out Dave, he might be overreaching on some of his conclusions.
Mike, I don’t really “prefer” one method over the other. As I said, I think you would reach the same general conclusion using just about any win generator. This is all guesswork, really, which is why you shouldn’t draw hard conclusions from any one approach. Don’t take these systems too seriously.
I think a key difference between Howard and Konerko is the timing. Konerko was a FA at the time who looked very likely to leave the chisox. Howard wont be a FA for another year and a half. A lot of the fervor out there amongst the blogosphere is about the timing of this deal, especially considering that when is due to become a FA is the same time as prinec and adgon, and of course albert if StL decides to sit on their hands (which they wont i assume). I think the money is a bit high, who knows I guess, but I think its the timing and the position that it puts phil in that is the most curious.
But, you don’t get the same result using any win generator. If you look at Howard’s career (which is generous to him), he’s averaged 3.7 WAR per season according to Rally’s number.
Matt at BP said:
“Obviously, if Howard ages gracefully and produces well into his thirties, the Phillies are probably going to be getting a great bargain”
Wrong. If he ages very gracefully and performs as well at ages 31-34 as he did at 26-29, the Phillies will be breaking even.
But Mike, if you use different win generators, you also need to use different monetary values. Can’t mix the two. That’s why I stick with WSAB, which, in the case of Howard and Konerko, closely resembles their Fangraphs WAR. You can’t just insert a different win generator and say it’s still worth $4 million or $5 million or whatever. It will change.
Excellent point, Beau.
I know the “LOL” criticism of the deal may be not be too pretty, but I tend to agree with it, and find those filling the Phillies side to just being contrarian for the heck of it.
Look, Howard is a really good player, but he’s also thirty years old and has skills that deteriorate quickly. As he loses some bat speed and power, teams will be less afraid to pitch to him and he’ll walk less. He doesn’t hit lefties. It’s not out of the realm of possibility to see him hitting .260/.330/.530 pretty soon…in fact, he’s on his way to it this season.
I think Beau’s point is spot on. One of the things that should happen when you sign a contract like this before you need to is what I call a “risk discount.” It’s something we see when a team buys out a players arbitration years. The player takes a little less for the security of having a contract over a period of time during which they could develop an Albert Belle hip injury, or blow out their knee, or get in a car accident. The team pays a little bit less than if the player were a FA today in return for taking on that very same risk. In this case, Howard got essentially what he would have gotten if he were a FA today and Philly took on all the risk but received no discount in return.
I should add that when Matt said it was a “great sign for Phillies fans,” he meant that it was a great sign that the Phillies are spending on the future, not that signing Howard in particular was necessarily a “great sign.” Apologies if my quote was misleading.
While I think it’s appropriate to hand out some “teams might know their guys better than we do” coupons, aren’t the Phillies *notoriously* anti-analysis? The teams I’d give the most leeway to are the ones performing their own number or biomechanical wizardry.
Yes, if the Eagles made this signing, or the Phillies under Gillick, I would assume that they knew something that I don’t. Amaro hasn’t done anything to deserve that kind of credit.
And of course, even if they know for a certainty that Howard is going to provide $125 million in value, there is still no reason to actually pay him $125 million, because no one else could reasonably conclude that he will be worth that amount.
Am I the only “sabermetrician” who doesn’t feel like he has anything intelligent to say about the deal? I don’t know the future. Maybe I’m more of a baseball physicist and not a “sabermetrician”, although I thought sabermetrics was about objective knowledge, too.
I agree that you wouldn’t call this sabermetrics. It’s more about baseball economics and management, which is just as fascinating to some of us.
I don’t begrudge anyone the discussion. I find it interesting, too. But I don’t like what I’m hearing out in the general baseball discussion world where the sentiment is that sabermetrics proves this is a bad deal.
Sabermetrics tells us what reasonable aging expectations are for groups of players similar to Ryan Howard. Sabermetrics tells us what teams have paid for wins above replacement value in the past.
Sabermetrics does not tell us how Ryan Howard specifically will age, and it does not tell us what teams as a whole will pay for wins in the future, or what the Phillies’ situation with respect to wins and playoffs will be over the next seven years. Those things are subjective judgments and guesses, and I hate to hear them proclaimed with as much certainty as the former things that are based upon objective facts. Not everyone is doing that, of course.
I meant to say “the Phillies’ situation with respect to wins and payroll” in the comment above, but perhaps their situation with respect to the playoffs is just as important.
Agreed, Mike. That’s why I posted my own opinion as a probability based on my gut (prodigious or otherwise). Matt did the same at BPro, though he used more data to inform his gut.
Of course, some of the people with strong opinions about the deal are also leaders in the sabermetric field, and/or post on sabermetric sites. I don’t begrudge them their strong opinions, and I also understand why the “general media” would associate their strong opinions with sabermetrics. Not sure there’s any way out.