Friday, July 31, 2009

More on the Rolen deal

Photo of Scott Rolen originally taken by Googi...Scott Rolen's a fine player. But was the cost too great? Image via Wikipedia

Through a strange twist of fate, I suddenly have a bit of time to spend working through today's big deal in a more careful way than I did earlier today (I have no real opinion on the Hairston deal as of now). Let's look at the players one by one.

Scott Rolen, RHB 3B
Year Age Team PA %K %BB %LD BABIP AVG OBP SLG ISO wOBA RAA WAR
2006 31.2 STL 594 12% 9% 20% 0.301 0.296 0.369 0.518 0.223 0.380 22.1 4.0
2007 32.2 STL 441 13% 8% 20% 0.287 0.265 0.331 0.398 0.133 0.327 -3.6 0.9
2008 33.2 TOR 467 15% 10% 21% 0.292 0.262 0.349 0.431 0.169 0.346 5.5 2.3
2009 34.2 TOR 373 11% 7% 25% 0.341 0.320 0.370 0.476 0.157 0.368 11.1 2.5
2009 34.2 TOR (ROS)
174 13% 8% --
0.308 0.284 0.351 0.439 0.155 0.349 2.2 0.9
2009 34.2 TOR (ZiPS)
547 12% 7% --
0.331 0.308 0.364 0.465 0.156 0.362 13.4 3.4
Rolen's had a bit of a resurgence this year, though as noted elsewhere it seems likely to be, in part, based on a "lucky" BABIP (though his LD rate is also up...though that too is also likely unsustainable). His updated ZIPS, based on stats to date and rest of season performance, put his value this season at 3.4 WAR (0.9 WAR the rest of the way). He's a nice all-around hitter, with average walk rates, nice power, and good contact ability. A big part of his problem has been injuries--he hasn't hit 500 PA's since 2006 (a full season is 700 PA's!). A weighted 3-year average of the 07, 08, and projected 09 stats puts his expected "true talent" hitting value at 2.4 WAR.

Fielding-wise, he's not what he once was, but still is a plus defender. A weighted 3-year average of his bUZR puts him at ~8 runs above average. His Fan's Scouting Report in 2008 was outstanding, rating him best in all of baseball at his position. I'll put him as a +1 WAA fielder, with the understanding that injuries can pull this value down.

So, in terms of win value (i.e. total value) the rest of this season (60/162=37% games remaining), we have:
+0.9 offense (from ZiPS Rest of Season)
+1.0 WAA * 0.37 = 0.4 fielding
+0.25 WAA * 0.37 = 0.09 position adjustment
---------------------------------------------
1.4 WAR over the remainder of this season, which is worth around $6.3M.

He's due to make $11M this season, of which I'm estimating 37% should be paid by the Reds, which is $4M. That's a surplus value of ~$2 million this season.

Next year:
2.4 WAR offense
1.0 WAA fielding
0.25 WAA position
-0.5 WAA aging
----------------
3.2 WAR Total

That's worth between $14.5 M and $16 M, depending on how the free agent market goes this offseason. He'll make $11 million, so that's ~$4 million in surplus value. Overall, we have a "property" that can reasonably be expected to provide ~$6 million in surplus value over the rest of his contract. He also fills several holes on the Reds--he's a good RHB, and solves (at least until the end of July 2010) the defensive problem at 3B. He's a good pickup. The question, of course, is the cost...

Edwin Encarnacion, RHB 3B
Year Age Team PA %K %BB %LD BABIP AVG OBP SLG ISO wOBA RAA WAR
2006 23.5 CIN 467 17% 9% 21% 0.307 0.276 0.355 0.473 0.197 0.361 9.7 2.4
2007 24.5 CIN 560 15% 7% 19% 0.322 0.289 0.354 0.438 0.149 0.353 8.3 2.5
2008 25.5 CIN 582 18% 10% 16% 0.264 0.251 0.340 0.466 0.215 0.352 9.6 2.7
2009 26.5 CIN 165 23% 15% 14% 0.250 0.209 0.333 0.374 0.165 0.322 -1.7 0.3
2009 26.5 CIN 181 19% 10% 0% 0.288 0.259 0.348 0.449 0.190 0.354 3.1 0.8
2009 26.5 CIN 346 21% 12% 0% 0.271 0.236 0.341 0.414 0.178 0.339 1.4 1.1
Eddie's had a horrible season thus far with the stick. But prior to this year, he had posted three consecutive seasons with offensive WAR's in the mid-2's. I tend to think that this year is an aberration. But to be objective, a 3-year weighted average puts his "true talent" offensive performance at almost exactly 2 WAR.

The problem with Eddie is that his fielding really hurts his value. A 3-year weighted average of his bUZR puts him as a -10 run fielder. His Fan's Scouting Report was more "positive," rating him just in the bottom third of 3B's defensively. Consensus seems to be he has a solid glove, but an awful arm--seems right to me. FSR data would put him at about a -5 run fielder. I'll split the difference and call him a -0.75 win fielder.

So, over the rest of this year, we have:
0.8 WAR offense
-0.75 * 0.37 = -0.3 WAA fielding
0.25 * 0.37 = 0.09 WAA position adjustment
----------------------------
0.6 WAR, which is worth ~$2.7 million on the free agent market. He makes $2 million this season, and 37% of that is $0.7 M. That makes his surplus value over the rest of this year ~$2 million.

Next year, we have:
2 WAR offense
-0.75 WAA fielding
0.25 WAR position adjustment
-0.25 Aging (I'm not taking off the full 0.5 WAR because he's still young; injury risk and my lack of regression still warrants a deduction)
-------------------------------
1.3 WAR, which is worth between $6-$6.5 million next year. He'll make $4.75 M, so that makes for about a $1.5 million surplus value--entirely due to the fact that he's still in his arbitration years (otherwise, his salary would be better).

So, overall, I'd rate Encarnacion as being worth about $3.5 million in surplus value. You can guestimate that he might be worth another $0.75 million surplus in his third year of arbitration (arbitration players don't make what free agents do), which would bring his value to $4.25 million.


Zach Stewart, RHP
Year Age Team GS% IP K/9 BB/9 HR/9 BABIP ERA FIP
2008 21.8 Reds (A) 0% 16.3 7.2 1.7 0.0 0.232 0.55 2.34
2008 21.8 Reds (A+) 0% 16.6 12.4 5.9 0.0 0.400 1.62 2.60
2009 22.8 Reds (A+) 100% 42.3 6.8 1.7 0.2 0.345 2.13 2.63
2009 22.8 Reds (AA) 100% 37.0 7.5 2.4 0.2 0.276 1.46 2.77
2009 22.8 Reds (AAA) 0% 12.3 11.7 5.8 0.0 0.369 0.73 2.55
Stewart was the Reds' 3rd-round selection (their second pick, as they did not have a 2nd-round selection) in the 2008 draft. He's done very well thus far and is skyrocketing through the minors. Scouting reports on him are strong, and he's put up extremely good ERA's thus far. His strikeout and walk rates have been a bit all over the place, but somehow it has all balanced out thus far in a very consistent FIP. Sickels rated him a B- pitcher prior to this season, and the consensus view is that he's a legitimate "B" pitcher now. "B" prospect pitchers are worth about $7 million in surplus value, based on their typical success rate and expected 6 years of below-market pay.

Josh Roenicke, RHP
Year Age Team GS% IP K/9 BB/9 HR/9 BABIP ERA FIP eW% WAR
2006 23.9 Reds (R) 0% 15.6 13.8 6.9 0.6 0.308 6.32 3.26 63% ---
2006 23.9 Reds (R) 0% 7.6 10.6 3.5 0.0 0.388 1.17 2.42 73% ---
2007 24.9 Reds (A+) 0% 27.6 13.3 4.9 0.3 0.373 3.25 2.44 73% ---
2007 24.9 Reds (AA) 0% 19.0 7.1 2.8 0.0 0.237 0.95 2.57 71% ---
2008 25.9 Reds (AA) 0% 22.0 11.5 4.9 0.8 0.358 3.27 3.47 59% ---
2008 25.9 Reds (AAA) 0% 39.0 9.9 3.2 0.5 0.323 2.54 2.74 68% ---
2008 25.9 CIN 0% 3.0 18.0 6.0 0.0 0.709 9.00 2.13 75% 0.1
2009 26.9 Reds (AAA) 0% 28.0 10.3 1.9 0.0 0.390 2.57 1.77 80% ---
2009 26.9 CIN 0% 13.3 9.5 2.7 0.0 0.355 2.70 1.85 79% 0.5
2009 26.9 ZiPS (RoS) 0% 17.0 7.9 4.2 1.1 0.313 4.24 4.23 52% 0.1
2009 26.9 ZiPS (Update) 0% 30.9 8.6 3.6 0.6 0.331 3.56 3.19 63% 0.5
Roenicke's been all over the place the last few years. But his profile is fairly consistent: hard thrower, very good strikeout numbers, and high-but-not-outrageous walk totals. He's in his first full season, so I think it's probably best to go with Sickels' pre-season rating for him and use that to assign value to him. Sickels gave him a B-, which I'll guess makes him worth about $4 million in surplus value, given his fairly advanced age (he's the oldest player in this deal aside from Rolen!). That's spread out over the first six years of his career.

Trade Tally
Rating strictly on surplus value, we have:

Update: Fay says they get cash to cover the rest of Rolen's 2009 salary. That would be ~$4M. I've updated the below bit to reflect that guesstimate.


Rolen: $6 million surplus
+$4 million cash(??)
vs.
Encarnacion: $4 million surplus
Stewart: $7 million surplus
Roenicke: $4 million surplus
Total: $15 million surplus

Assuming the cash sent to the Reds above is accurate, Toronto "wins" trade by ~$5 million, which is roughly the value of a bench player as a free agent.

Does this deal make sense for the Reds? You can make the argument in a lot of ways that it does. They upgrade third base by roughly 1.5 wins next season, get a good right-handed bat (though only modestly better than Eddie's), and continue to upgrade their fielding.

The problems from my perspective are two-fold. First, the Reds overpaid by roughly one Josh Roenicke (though we're admittedly within the margin of error here). Second, even if this was the market rate, they acquired a single aging, injury-prone player in exchange for cheaper, younger players. If the Reds were likely to contend this or next season, it makes sense to do. But who outside of Bob Castellini really thinks that the Reds are in that sort of position? Rolen doesn't improve them by more than 2 wins per season, tops. The Reds prior to this trade were a sub-0.500 club, and by at least one way of forecasting these things, on their way to finishing last in the division. The Reds sacrificed a small part of the future for the now, and the two wins Rolen provides don't help the now very much.

On the plus side, at least I don't have to fret about why the Reds aren't moving Edwin to the outfield anymore. ;)

14 comments:

  1. I think the reason Stewart's K and BB rates are all over the place is because he was used as a starter in A+ and AA ball this season, but he's been a reliever everywhere else. The rumor is that they put him back in the bullpen in AAA to manage his innings, but the intention was to convert him to a starter over the long-term. His numbers definitely look impressive as a starter.

    Now you have to wonder if trading Stewart means it might be too risky to trade Arroyo or Harang since they don't have the pitching depth to fill in. Travis Wood is the only true starting prospect in AA or above. And with Volquez having a setback yesterday, I'm not sure trading a young starting pitcher really made sense.

    ReplyDelete
  2. Good points on the k-rates, and all around as always.

    Since this season's more or less in the can, I have no problem with trading Arroyo given that he's so overpaid. Hopefully they could replace him at better value on the free agent market this fall. But I absolutely agree that this seriously cuts our pitching depth. It's a deal a competing team might make--but that's not what the Reds are.

    ReplyDelete
  3. Justin-

    I love your work, but is just adding up the value a fair way to analyze the trade? I think that you (and most of the saber-community) is significantly over-rating Roenicke and Stewart because the bias is to kill Jocketty for acquiring the proven veteran at the expense of young arms and cheap production (I know it was mine when I first looked at it). But the "straight" calculations are too linear and don't really accurately reflect the true merits/demerits of the deal.

    The variance on the predicted production of 2 minor league pitchers is much greater than the variance on guys like EE and Rolen, isn't it? A linear analysis equates Roenicke and Stewart’s future value with the same confidence interval as Rolen and EE. That’s not true though. Because of sample sizes, we can much more accurately predict EE and Rolen’s future value than Roenicke and Stewart. Indeed, minor league relievers’ statistics tend to have the least predictive value of any minor leaguers’ stats.

    On a related point, isn't it safe to say that any minor league system ought to be able to spit out a Josh Roenicke every year or so (e.g. Jared Burton, Nick Masset, Danny Lee)? If so, what is really the harm if you occasional give up one or two of these guys to upgrade from EE to Rolen. It seems to me that the absolute worst case scenario is that you blow a few extra million on that year’s free agent reliever out of the Rhodes / Weathers class.

    Lastly, I’d put the odds that Roenicke and/or Stewart are significantly effective pitchers in 2011-13 at roughly the same odds that Rolen continues to produce at an all-star caliber level for the next 1.5 years, and so the Reds offer him arb and get either 2 picks or one more positive value year out of the guy. On top of all that, Rolen may like it in Cincinnati and sign a below market deal to stay.

    The more I look at this deal, the more I think Jocketty improved the 2010 Reds at very little cost to the budget, and the cost of MAYBE having to overpay for 1-2 middle relievers on the 2011-13 Reds. That seems like quite a nice trade.

    -CTM

    ReplyDelete
  4. CTM,

    Thanks for the detailed comment.

    With respect to the value placed on the pitchers--a single league-average season by a player making league minimum is worth roughly 8 million in surplus value. So, I'm rating Stewart as if he will do one of those seasons. That's because, on average, a player like him will do that. As you said, the variance is huge. Some will put together much more value, while some will provide zero. But the average performance is the equivalent just shy of one average season (as a starter).

    I think that's a pretty conservative projection for the guy, and one grounded in Victor Wang's data. It almost sounds like your alternative suggestion is to assume they'll provide zero value, which seems far to "conservative" to me.
    -j

    ReplyDelete
  5. J-

    First off, I want to reiterate how much I love your site. I’ve lurked around on it for years. Your work almost makes it fun being a Reds fan.

    My understanding of the Wang analysis is from his THT piece on the Santana trade last year (here: http://www.hardballtimes.com/main/article/the-bright-side-of-losing-santana).

    I think it’s a great bit of analysis, and furthers the discussion, but I still think that he’s overvaluing future pitching prospects. He more or less admits as much.

    Here’s my main point of contention (which is somewhat related and somewhat different than Wang’s): Roenicke and Stewart were traded precisely because they’re at their peak value right now.

    I didn’t see either of them on any major Reds Top 10 prospect list, which means it is unlikely that either pitcher was one of the Top 100 pitching prospects in baseball. I suspect that the list of pitchers outside the Top 100 list who have made meaningful contributions to MLB rosters is thin.

    Jocketty, to his credit (and I never thought I'd be typing that), recognized this and traded on the high. And even if they’re both in the B/B- class (I contend that they are not, and that their “upside” is a B/B-), there is still a 1-in-4 chance that both washout and never meaningfully contribute to a MLB roster.

    If Wang’s approach were sound, the best “surplus value” strategy that a team could employ would be to always trade their stars and always acquire prospects with 6 years of control. The problem there is that you’d always be 20 games out of first, with the best AA team in baseball. You’d be the Marlins, who every 7 years or so get lucky when the stars align and they have an incredible core of pre-arb MLB-ers. But it’s defeatist because the one certain thing we know about pre-arb guys is that they will reach arbitration and then they’ll have to get dealt for the next package of Andrew Millers and Anabel Sanchezes.

    Since that strategy would necessarily produce an undesirable Marlins-like outcome, the analysis must be wrong. The flaw is the “straight” analysis of the minor leaguers fails to add big enough cost for the uncertainty of the prospect. This is even more so with minor leaguers pitchers. The prospect-dealing club always has to overpay (on "dollar surplus" standpoint) to account for this extra risk.

    CTM

    ReplyDelete
  6. "If Wang’s approach were sound, the best “surplus value” strategy that a team could employ would be to always trade their stars and always acquire prospects with 6 years of control. The problem there is that you’d always be 20 games out of first, with the best AA team in baseball. You’d be the Marlins, who every 7 years or so get lucky when the stars align and they have an incredible core of pre-arb MLB-ers. But it’s defeatist because the one certain thing we know about pre-arb guys is that they will reach arbitration and then they’ll have to get dealt for the next package of Andrew Millers and Anabel Sanchezes."

    I agree with this conclusion. You have to spend more money or else be an outlier like the Marlins. Unless the Reds find 10 or 20 million somewhere, they'll be hard pressed to find success. The likelihood of them finding that money is very slim, which makes the prospect route seem more viable, but you can also end up wasting a 5 year cycle if your "small batch" of quality prospects don't work out and you don't have 100 mil. That's why Pittsburgh is all in for quantity and you can now look at them the very same as the Marlins and perhaps Rays.

    We'll see if they find success before the Reds - what havoc that would cause in Cincinnati!

    ReplyDelete
  7. Jon-

    I think that's right -- Pittsburgh will be an interesting study, even moreso than the Marlins because Pittsburgh hasn't really acquired the A Level talent (in trades) that the Marlins obtained when the dealt for HanRam and Maybin.

    That having been said it made/makes more sense for Huntington to have executed on the Marlins/Rays strategy because of where the Pirates were on the success cycle (i.e. off it).

    The Reds, on the other hand, probably have something like a 6:1 shot at the NL Central next year, and so making these small but real upgrades really make a lot of sense. If the Reds can get into the mid 80s wins next year, it likely leads to increasing attendance/viewership and Rolen pays for himself.

    CTM

    ReplyDelete
  8. CTM,

    A few things.

    1. If you look at my analysis again, you'll note that I think it was a fairly minor overpay. I rate it as the equivalent of 1 WAR overpay, which is all the value I allowed for Josh Roenicke in surplus (which my feeling is a conservative estimate). As trades go, that's pretty close to the mark. In fact, I acknowledged that it's within the margin of error for this analysis, so you can argue that it's a market-value deal.

    To put it another way, EDE for Rolen would not be a fair swap. Rolen's a much better talent, even after accounting for the salary difference. So the prospects are needed to sweeten the pot, especially given the money that was also sent. The question is how specifically you quantify their value. Wang's work offers one of the first solid estimates of those players' values. It's early in this field, but his estimates make sense to me.

    2. Some of what you said implies you're just disagreeing on how we value Stewart. He was a B- at the beginning of the year by Sickels, and his performance this year justifies a "B" rating given his age, that he's starting now (or was, rather), and reached AAA. I have seen similar ratings by others (and I am not including him on the top-100 list, though some have). But it's fine to disagree on his value--doing so, however, has no bearing on the methodology being used, which you're simultaneously arguing against.

    3. Wang's (and others') methodology assumes that teams have a set payroll that they can afford to spend, AND that they will spend to that amount every year. Therefore, the question isn't how much you spend. The question is how you spend that money. When we say someone has surplus value, that means that if teams have that player on their roster, he'll contribute more than you could reasonably expect a free agent to contribute. So, if you have a team of all free agent-value players, and another team of surplus value players, both with the same payroll, the surplus value team will score (and prevent) a lot more runs.

    4. If you're acquiring AA prospects, you can expect that some of them, at some point in the future, will contribute a certain amount to the team. The estimates come from how much players of this sort will generally contribute when making slave-wages, plus (iirc) a discount for it taking a while to arrive. It's also worth noting that we typically assume that the money saved by the cheap player will be invested to make the big-league club better (i.e. teams spend to budget). This is why the Marlins aren't good--they have tons of surplus value, but even so they just don't invest enough into the team to make it a good team.

    What I'm not hearing from you folks is any kind of alternative way to value prospects. The only thing I'm hearing essentially sounds like assuming they have no value. And that's obviously not true, or teams wouldn't make trades for prospects. So how would you go about estimating their value? Half what we're using? 1/4th? What's the empirical justification? I'm happy to hear such arguments, but just complaining you think we're overvaluing them isn't advancing the conversation much.
    -j

    ReplyDelete
  9. "I suspect the list of pitchers outside the Top 100 list who have mande meaningful contributions is thin."

    Whoa! No way. Pitchers come out of nowhere all the time. Basically all pitchers in the Top 100 have similar expected future production, which sounds crazy, but with injury attrition and not living up to expectations, it's basically true. The value remains somewhat flat outside the Top 100. (The Top 100 is actually a pretty shallow group, as it's about three prospects per team. When you start including each team's Top 10 list, that's 300 prospects...)

    ReplyDelete
  10. Thanks Sky. I’m sure you’re right and I’ve overstated the paucity of Below 100 pitchers to make contributions in the bigs. I think my point is the same (as distilled by Brian in the next set of comments): Wouldn’t it be better to analyze these prospect-for-MLB-ers trades on a median value basis instead of an average value basis?
    CTM

    ReplyDelete
  11. ...and by "Brian", I of course, mean Brad.

    ReplyDelete
  12. Word is now leaking out that the Jays may also be paying $4 million of Rolen's salary in 2010. That seems to make this an awfully close trade, even without taking into account the intangibles of "leadership" and "experience" that Rolen may provide to the rest of the Reds position players. Thoughts?

    ReplyDelete
  13. spmancuso,

    If that's true, then yes, I'd say it's a dead-on market even deal. Neat. :)

    I'm still not convinced that it is the direction the Reds should be going, because again, it is a trade for the now rather than the future. But the Reds (and some commenters) obviously think they're closer to contention than I do.

    FWIW, my power rankings at BtB are almost certainly going to rank the Reds 30th out of 30 teams when I post them tomorrow. Barring a blow-out win of epic proportions tonight. I just hope the Yankees aren't in first (very close between them & the Rays).
    -j

    ReplyDelete
  14. According to Baseball America, Zach Stewart is the second-best prospect acquired at the deadline, behind former Sun Devil Brett Wallace.

    ReplyDelete