BBA Ballot: National League Rookie of the Year


As part of the always amazing Baseball Bloggers Alliance, part of our job was to get a yearly ballot in for all the major awards and stack them up against the BBWAA to see how well we do. We already got through the Managers of the Year for both the NL and AL, and congratulations go to Colorado Rockies skipper Jim Tracy and Los Angeles Angels (of Anaheim) manager Mike Scioscia for taking home the inaugural BBA Managers of the Year crown for their respective leagues.

I didn’t have too much fun doing that one, but I’m going to love this one. The NL rookie race is as tight as it can possibly be. There are a lot of candidates and only three spots to fill, so it’s going to be tough to figure it out. To be able to tell how much each player contributed, I’m going to use a homebrewed version of WAR for each of these rookies.

For position players, it will be the typical WAR equation, except that I’ll be rebuilding the event weights for each player as per the directions shown here by Tom Tango. I removed pitcher hitting from the equation and took all major league numbers as a whole rather splitting the leagues. My weights did not come out exact (my league average wOBA as calculated by the weights was 0.003 points short of the league average OBP), but I’m OK with this, as the run difference between measuring from a .334 wOBA and a .337 wOBA average is 1.5 runs over 600 PA; if the distances are that close, I’ll think consider additional concerns. In addition, the individual weights pass the eyeball test for the In any case, to be safe, I’ll be comparing to the average wOBA that I got from using the component league stats rather than using the average OBP. Park adjustments will be done using terpsfans’ regressed park factors from 2008.

For baserunning, as always, I’ll use Baseball Prospectus’ Equivalent Baserunning Runs (EqBRR), meaning I did not take into account any baserunning outs in the wOBA weights. For defense, I’ll unfortunately have to use only UZR, which restricts my defensive scope a good deal; in the past I’ve used UZR, TotalZone, and Fans Scouting Report data. I don’t know if I can convert this, but I doubt it would make a whole lot of difference, as I’d still be weighting UZR 75%. If TotalZone numbers come out or if someone can make me privy to plus/minus numbers for the year, this may change. Position and defensive adjustment values are ripped directly from FanGraphs.

For pitching, it will be a bit easier. I took the average of StatCorner’s and FanGraphs’ tRA, then took that value and averaged it with FanGraphs’ FIP/0.92 (to put FIP on the runs allowed scale) and used that for runs. I used Pythagenpat for wins, replacement level pitcher at 38% win percentage.

As for the results, Marlins fans, you aren’t going to like it.

3. Tommy Hanson, RHP, Atlanta Braves

Hanson came out third in WAR with 2.9 WAR despite pitching only 127 innings. It was an impressive feat to pick up as many wins and be in the discussion as he was. Hanson had a 3.50 FIP and an average tRA of 3.78, mainly on the back of an excellent 22% strikeout rate. As far as pitch values are concerned, it’s amazing that a rookie, especially one of Hanson’s age, owns two plus breaking pitches; his slider this year was worth 1.79 runs per 100 pitches and his curveball worth 2.36 runs in the same number of pitches. He’s been hyped like crazy before his arrival and has definitely lived up to the name, so it’s no surprise that he got on here. Had he pitched 167 innings like Randy Wells or J.A. Happ did, he would likely have a sizable lead in this race and would probably be my winner.

2. Randy Wells, RHP, Chicago Cubs

Wells has been a revelation for Chicago, even though most of Chicago’s season was lost in the spree of injuries and underachievement. Wells’ 12-10 record and 3.02 ERA are decently impressive, though I suppose they fall short of Hanson’s respective numbers (11-4, 2.89). While Wells wasn’t the lights out pitcher that Hanson looked like he was, he was still more than efficient, striking out 15% of his batters faced while only walking 6.6% and allowing just 14 home runs. These numbers combined to give him a solid 3.88 FIP. His 47% ground ball rate also helped in the tRA department, though slight difference in the classification of line drives and fly balls (one of the primary problems with using tRA raw instead of regressed tRA*) led to drastically different tRA totals. It actually made me consider using the average tRA over the tRA from one source.

In any case, despite Wells’ age (he’s 26), he definitely deserves credit for his excellent season. It doesn’t shine like some of the other rookie years, but 3.2 WAR is 3.2 WAR, and Wells has done with longevity, being among the rookie leaders in innings pitched and pitching at the decently high level he has.

1. Andrew McCutchen, CF, Pittsburgh Pirates

I’ll admit that I had McCutchen in mind when I started this analysis. McCutchen posted a 3.3 WAR season which admittedly isn’t light-years ahead of Wells’ 3.2 WAR. If I could, I’d give both players first place votes, but since I cannot, I’ll go with the player with the highest total in my system and the one I would prefer to see win Rookie of the Year. McCutchen has been everything the Pirates wanted him to be: fast, patient, decent pop with the bat, good in the field. Well, he hasn’t been exactly everything, as he posted -0.8 runs in UZR, which is about average. But you can’t doubt the offensive production McCutchen brought: a .286/.365/.471 slash line, good for a .366 wOBA in my book, plus 22/27 steals (though he only put up 1.4 additional runs in baserunning), done at a premium position.

null

I know Rum Bunter's going to love this; Cutch is looking for his approval!

Andrew McCutchen, congratulations, you’re the Marlin Maniac’s NL Rookie of the Year!

Honorable Mentions: Chris Coghlan (LF, Florida Marlins), Garrett Jones (RF/1B, Pittsburgh Pirates), J.A. Happ (LHP, Philadelphia Phillies)

And now comes the part where I explain why I of all people would leave Chris Coghlan off my ballot. It isn’t because I don’t want to be seen as a homer (even though I can be at times). For what it’s worth, Coghlan was fourth on my ballot, but the cutoff is three. In my mind, Coghlan was second behind McCutchen, but a distant second, and this is as a result of one thing: defense.

Let’s face it, Coghlan wasn’t pretty in the outfield. This was his first time out there, and the Marlins were foolish really to throw a player out to a position to which he was unaccustomed and expect good results. Coghlan tallied -9.8 runs at the position by the end of the year. My eyes generally agreed, though I passed on evaluating Coghlan because I needed to see more. In short, he hurt the team with his play out there and he hurt his value by playing poorly in a replaceable position.

This doesn’t mean I am discounting his offense, but it does mean that his value was diminished. To those who say UZR is unreliable in such a small sample, I say that this is the best we’ve got and that, even considering UZR’s measurement error of +/- 5 runs for a given season, we’d still be talking about an even race between McCutchen and Coghlan on defense before adjusting for position. Of the five position players I measured, he was the worst by far defensively, which evened out a lot of his 20 runs above average offensively, tops among players those same players.

(P.S. Those who claim the amazing finish to Coghlan’s season gave him Rookie of the Year seem to forget that he also started the year like Jeremy Hermida. You just can’t start counting only the good times.)

Once again, we come to the adage that “defense counts.” I was prepared to consider different approaches if the gap was within something like 0.3-0.5 WAR, but Coghlan was only 0.5 WAR away from the third place Hanson, and I figured it would not be worth the effort. Coghlan will get a lot of votes regardless, and I’m sure he’ll finish in the running, though ultimately Hanson and Happ may overtake him, but if you buy into defensive analysis at all, you have to figure Coghlan hurt the team with his fielding.

This does not take away from his season and his future with the team. He solidified himself as a good player and someone we’ll be counting on next year. Here’s hoping wherever he plays, he shows improvement.

  • Juan C. Rodriguez

    Hey Michael,
    Long time no talk. I agree with you on Coghlan’s defense. Not entirely sure how much better or worse of a second baseman he is than a left fielder, but I think the Marlins would be wise to bring him back to the infield.

    That said, don’t you think the historic nature of his offensive season justifies forgiving some of those defensive sins? If we condemn him for doing a lousy job in a position he never played, shouldn’t we also laud him for excelling in a spot in the batting order he never hit?

    Right or wrong, the writers only look at the offensive numbers (my paper is among those that doesn’t let its writers vote BTW). Just go back to the Braun-Tulowitzki debate for evidence of that.

    Keep up the good work.

  • Michael Jong

    Juan,

    For the award, I chose to run with it in terms of production. I agree that he had an amazing end of the year, but in my opinion, it shouldn’t count any more than the beginning of the year. He should get credit and blame for everything good and bad that he’s done. I loved his last two or three months, but there’s no reason to count them more than their plate appearances count, in my book.

    And you’re right, the BBWAA only votes offense, and it’s sort of a shame. My gut tells me it’ll turn out 1) McCutchen, 2) Coghlan, 3) Hanson/Happ, which I’d be perfectly happy with. It just wouldn’t take into account the fact that Coghlan wasn’t good this year at an “easy” position, while McCutchen was a good deal better at a much better position. I’m all for getting our players more recognition though, and being close or winning the Rookie of the Year definitely catches some eyes.

  • fetch

    Honestly I’d be hard pressed not to go with Hanson. Honestly, I wonder if the writers try to peg who will be a future star when they vote and not go with who had the best season. For example, if you pretended that McCutchen didn’t play this year, Wells would have Hanson beat in WAR and would merit the award, but Hanson looks like an ace in the making so to make sure they look “right” in the future the writers would vote Hanson.

  • Michael Jong

    Fetch,

    I absolutely agree. I think the writers do indeed look for future stars in their voting, so they have a tendency to go after the hyped prospects. Luckily for McCutchen, he was somewhat hyped (though not at the same level as Hanson), and so he’ll get some of that as well. It’s a shame that Wells probably won’t get a whole lot of credit even though he certainly deserves it.

  • Simon Cowbell

    “For the award, I chose to run with it in terms of production. I agree that he had an amazing end of the year, but in my opinion, it shouldn’t count any more than the beginning of the year. He should get credit and blame for everything good and bad that he’s done. I loved his last two or three months, but there’s no reason to count them more than their plate appearances count, in my book.”

    What about the 0 plate appearances McCutchen had while Coghlan piled up his first 72 (Coghlan finished with 565 while McCutchen had 493) at .321 BA/.850 OPS (remember, we are combining the good four months of numbers with the one bad month… that’s what you wrote)? I imagine you used some stat-geek rationalization to say what McCutchen WOULD have done in those three weeks he wasn’t in the majors.

    Also, where is the analysis of how these guys hit with when some plate appearances DO count more than others… two outs and RISP…. Or when the games were close?

    Well, here they are:
    McCutchen Coghlan
    2 outs, RISP
    .222 BA, .822 OPS .308 BA, .873 OPS
    Late and Close
    .226 BA, .697 OPS .269 BA, .743 OPS
    Tie game
    .246 BA, .776 OPS .355 BA, .926 OPS

    And, in games where the margin was out of hand (at least four runs)?
    McCutchen Coghlan
    .314 BA, .926 OPS .284 BA, .767 OPS

    Finally, getting to this notion of scarcity of position. Yes, McCutchen is a good center fielder. Of course, Coghlan isn’t the butcher you made him out to be (those defensive stats more or less say the same thing about Mark Teixeira at first base for the Yankees).

    But what about batting leadoff. You act like that isn’t a position requiring an exceptional skill set compared with almost anywhere else in the order. Few major leaguers can handle it like Coghlan did in 2009.

    Oh, and the guy was fifth in the NL in batting average. That includes all those non-productive at-bats you were crying about.

    Very disappointed in your flimsy, incomplete breakdown on the Rookie of the Year.

  • Michael Jong

    Simon Cowbell,

    All right, let’s hit those points, shall we?

    0 PA: WAR takes into account playing time, in the replacement level adjustment, so there’s no need to consider. I don’t even have to extrapolate, Coghlan got two extra runs in the analysis for playing when McCutchen was in the minors.

    PA’s that matter: Looking at FanGraphs’ Clutch, which determines performance (in terms of wins) in situations where the leverage is 2 or greater (higher leverage means more important situations, such as the ones you’re mentioning).

    McCutchen: 0.57 Clutch wins
    Coghlan: -0.15 Clutch wins

    These numbers DO matter, but I did not include them because I like my analysis more context independent. If you wanted to add them, you’d have more of an advantage to McCutchen.

    Defense: Teixeira posted a UZR of -2.7 this year, indicating that he performed at something around average, and this point would not even be useful in arguing against the usefulness of defensive metrics. Coghlan had a UZR of -9.7. Plus/minus had him worse (from what I heard, -16 plays, which is -11 runs). McCutchen was at -0.8 runs at a better position. You’d have to dock McCutchen the full 5 run error bar and add on the full 5 run error bar to Coghlan in order to make it even. The difference is large enough that I can confidently say that Coghlan was significantly worse than McCutchen playing an easier position.

    If you’d prefer a subjective measurement, let’s use the Fans Scouting Report, which is based on fan voting.

    Coghlan: 3.18 out of 5 in left field (18 votes)
    McCutchen: 4.32 out of 5 in center field (13 votes)

    The fans agree, though not as violently as the metrics do.

    Even if they were both average or even, McCutchen playing center field gives him a 1 WAR edge over Coghlan due to scarcity of position. Coghlan was likely ahead half a win with playing time and hitting. McCutchen still takes it by half a win.

    Batting leadoff: McCutchen batted leadoff 108 times this season. Coghlan batted leadoff 106 times this season. Should McCutchen get extra credit for that too? I don’t really think so. And while leadoff is one of the most important lineup positions in baseball (tied with second and cleanup), both players did a very good job producing at the position.

    Batting average: So what? Batting average is useless compared to using OBP/SLG/OPS/wOBA. I already said Coghlan was about four runs better on offense than McCutchen; they’re close, but Coghlan has a definite edge. It’s on the defensive side and the fact that he plays a less valuable position where he hurts himself.

    Look, I’m not denigrating Coghlan’s numbers, but defense counts, and the metrics say Coghlan wasn’t good. If you cost the team runs, you cost the team runs. McCutchen’s batting was almost as good as Coghlan’s, and he did it at a more difficult/scarcer position and played his position better. It’s difficult to deny that.

    I hope that addressed your concerns, though I don’t mind continuing this via email. Anyone else care to jump in on either my defense or Simon’s?

  • Simon Cowbell

    I give you simple-to-understand situational numbers, and you spew car-mechanic byzantine sludge.

    Two outs, with men in scoring position…. EVERYONE gets that. Coghlan was dominant.

    In close games, In late-and-close games … EVERYONE gate that. Coghlan was dominant.

    In blowouts … EVERYONE gets that. McCutchen inflated his stats, where Coghlan’s were hurt.

    English, please.

  • Simon Cowbell

    Just was reading up on the FanGraphs Clutch wackiness.

    You ARE kidding, right?

    Barry Bonds was a -0.49 in that crap stat in 2001, when he hit 73 HRs in 476 at-bats.

  • Michael Jong

    Simon Cowbell,

    It’s not about “byzantine sludge,” but rather using a measure that more completely covers important situations. Leverage as defined here:

    “Leverage Index (LI) is a measure of the relative importance of any game situation. Created by Tangotiger based on the Win Probability Added (WPA) framework, it is scaled so that average is 1. An LI of 2.0 would be twice as crucial as a typical situation; likewise, an LI of 0.5 would be half as crucial as a typical situation. The highest possible LI is 10.9.”

    More explanations here from Baseball-Reference, which uses the same sort of measurements.

    This definition makes intuitive sense: PA’s that are in more pressure packed situations are more “clutch,” correct? Leverage Index measures that. If you check the various game logs on FG or B-R, you can see that the high LI’s (LI > 2) correspond to important game situations. High LI encompasses more important situations than the stats you quoted; it also includes many of the situations you included.

    If you don’t like the Clutch metric, I’ll quote B-R’s splits.

    McCutchen’s batting line in high LI: .299/.372/.493 (78 PA)
    Coghlan’s batting line in high LI: .277/.411/.322 (75 PA)

    I’m not running the wOBA’s to tell how many runs each production value is worth, but I’d say Coghlan is a dog in this comparison.

    Not even using Clutch, I got values that say McCutchen’s performance was slightly better in important situations.

    The difference between high leverage situations and the ones you quoted? For example, if there’s an RISP and 2 outs in a 12-3 blowout, is that situation important? Also, is a bases loaded, 2 out situation in the third inning important? The categories you mentioned at times do not include situations that are important to the game, whereas the “high leverage” category covers all important events in the context of base/out state and inning/score.

    Anyone else interested in jumping in on this one? Do the leverage splits make sense?

  • Michael Jong

    Simon Cowbell,

    I wanted to answer the Bonds question separately. First off, note that clutch measures the performance as compared to that same performance in a neutral leverage. If there are two on and one out in the seventh inning, tie game in the eigth inning, and Bonds is at the plate and gets intentionally walked, do you think that walk is worth more or less than if it were a normal situation, say no one on, no outs in the second inning? I’d venture to say that that walk is worth less, and you probably would agree.

    That’s the sort of thing clutch may be measuring. Bonds walked 34 times, 13 times intentionally, in appearances with LI > 1.5, according to Baseball-Reference. Clutch measures appearances at LI > 2, which I’d venture to say would give him an even higher walk rate. So while those walks are generally valuable (worth 0.03 wins usually), they may be worth 0.01 wins in these situations. His home runs (eight of them) count too, and so do his outs (51 outs in those 104 PA).

    So even though he hit pretty well in these high leverage situations, the slash line doesn’t quite tell the whole story. To truly tell how well he played, you’d have to look at his game logs and each high leverage situation. Clutch basically does this for you.

    I explained it, and if you’d like to talk to me more about it, feel free to contact me, you can find my email around here somewhere, and I would be more than happy to talk to you about this. This is my opinion, using statistics that I understand as more comprehensive of the concept of important situations.

    Anyone else care to chime in as well? Join the conversation and get on the Maniac’s case if you don’t agree.

  • Simon Cowbell

    McCutchen hits .226 in late and close, and by the time you guys are done with him, he’s Mickey Mantle. Incredible.

    Uh-oh… I just invited a dissertation of how Mantle really wasn’t that good of a player according to VPWOM (Value playing without Maris).

  • Michael Jong

    Simon Cowbell,

    I’m not propping up McCutchen any more than how he played in the situations that are important, the ones you wanted to measure, but used multiple categories that don’t completely tell the entire story. Leverage is a great way to split up what you are looking for, the PA’s that matter the most. If you’d like to address the points I mentioned previously, be my guest.

  • Simon Cowbell

    I gave it to you.

    .226 late and close. How do you salt-water taffy that into something decent, much less better than Coghlan’s numbers in such situations?

    If you can’t simply explain the stats you are using, you shouldn’t use them.

    The deafening silence in here would seem to back my contention up.

    Coghlan is only the ninth rookie ever to hit at least .320 in a minimum of 550 plate appearances with at least a .850 OPS.

    Only the second in the past 70 years. Him, and a guy named Pujols.

    You FAR short-sold Coghlan’s offensive performance.

    From 1901 to 2009, During first season (BA>=.320, PA>=550 and OPS>=.850)

    2009 – Chris Coghlan
    2001 – Albert Pujols
    1939 – Ted Williams
    1936 – Joe Dimaggio
    1929 – Earl Averill, Johnny Frederick, Dale Alexander
    1928 – Del Bissonette
    1926 – Paul Waner

  • Michael Jong

    Simon,

    I can simply explain my stats. In fact, I just did. If you’re not listening to my point, you’re sort of just yelling at me and plugging your own ears. It’s not easy to have a discussion when you do that.

    I’ve been saying that the stats you’re quoting don’t include all important situations in a game, and in fact include some relatively non-important situations. Which situation, for example, is more important: 2nd and 3rd, 1 out, fourth inning, down one run; or no one on, 2 out, ninth inning, down one run. The higher leverage goes to first one by a good margin, but you’re quoting one stat that doesn’t even include it! That’s why I was quoting the high leverage batting line. It covers all situations that are important. Nothing simpler than that.

    Also, as “historic” an offensive performance as Coghlan’s was (and I’m not saying it wasn’t, as Coghs went on an absolute tear in the second half), it was still only worth a certain amount. You can make lists all you want, but I can tell the approximate amount of runs it was worth, and that’s far more important than the arbitrary baseline that you used to put Coghlan on a list with Pujols.

    Pujols hit .329/.403/.610, a .421 wOBA. That’s not nearly on the same scale. The reason they’re on the same list was because you built a baseline to make it that way. You can’t lump Pujols’ production with Coghlan’s as so similar as to be put on the same list.

    I’m explaining my point, it’s actually fairly simple. I’ve also addressed your points, I think you’ll agree with that. I believe the conversation is stalling on your end, when you continually quote your stat and refuse to listen to my end of the story. I’m certainly listening to yours, but the story “late and close” is telling does not include all important situations, whereas the “high leverage” does. Did you read the definition from I quoted?

  • Michael Jong

    Simon,

    Didn’t want to run the previous post too long. If you’d like, we can take context entirely into account and use WPA. B-R explains WPA, but essentially it’s the win probability added by a player’s events/actions. If there’s any clutch situations, you’re going to see a large jump in WPA. If there are any blowouts, WPA will not care about those events. For the clutch argument, it seems like that’s what you would like (you did quote how McCutchen was “padding his stats” in blowouts and such).

    McCutchen: 2.02 Wins
    Coghlan: 1.19 Wins

    This includes any clutch situations they faced and any non-clutch situations they faced. It’s based on all of their events and how they added or took away from the team’s chances of winning a game. I don’t like using it because I’d prefer production that is context-neutral, so it’s not dependent on teammates, but there’s a measure that’s entirely context dependent.

    Also, I’m not making any of these stats up. They’re well researched and featured on FanGraphs and on Baseball-Reference.

    I would like to let it be known once again that our organization voted McCutchen overall, so I clearly wasn’t the only person who felt this way. And McCutchen plays for Pittsburgh, which gets less attention than Florida because the Marlins actually win games, so there can’t be any argument of a lack of exposure.

  • http://www.2thinkgood.com Jorge Costales

    Count me as siding with Simon’s points made here. Further, Michael you wrote – “I loved his last two or three months, but there’s no reason to count them more than their [earlier?] plate appearances count.”

    Perhaps that would be a more valid point with veteran players. But as rookies, how they finish is the only indicator we have as to how they are adjusting and developing.

    Both players had 4 months of plate appearances ranging between 100 & 120 AB’s [only exception was Coughlin in July]. McCutchen’s September OBP & SLG numbers were both his lowest month of the year. Whereas Coughlin was one of the best hitters in baseball for August & September.

    McCutchen was good, but Coughlin was historic.

  • Michael Jong

    Jorge,

    I see your point, but my contention is that a two month span at the end of the season could just as easily be statistical fluctuation as it could be adjustment. It’s hard to tell the difference in such small sample sizes. And what I’ve heard a lot of is that Coghs’ second half was so “historic” (and it was awesome, believe me) that no other considerations should be made. There is little to no doubt in my mind that Coghlan was the best rookie hitter in the National League.

    However, the RoY isn’t awarded to the best hitter. And, to your point Jorge, it isn’t awarded to the “rookie hitter who adjusted the best” either. In my mind, it goes to “the most productive rookie over the entire season.” If you count defense, even if Coghlan wasn’t as bad as the stats say, McCutchen was still much better just because he plays center field. A left fielder needs to be a full win better just to make up the difference in positional scarcity. That offensive difference is the difference between Chris Coghlan’s offense this year and Nate McLouth’s offense this year. McCutchen was about four runs worse than Coghlan, according to wOBA.

  • http://www.2thinkgood.com Jorge Costales

    Not to beat a dead rookie, but if you saw my point re how the end of season performance was a more substantial factor for rookies than it would be for veterans, then what’s the point of mentioning that ‘its hard to tell the difference in such small sample sizes.’ The answer is that that is true, but it’s all you have in the case of rookies.

    That aside, you make a good and typically under-appreciated [by people like me] argument for not overlooking the value of defense. The development of defensive stats are a welcome addition to the baseball debate – but the ‘concreteness factor’ of the ‘he’s worth a run’ argument makes me want to pass on that Kool-Aid for now.

    Aside from the math working, stats have to make sense at a certain level of intuitiveness. It eventually made sense to fans that it matters more that you get on base [OBP] than batting average. Quantifying defensive value has a much bigger hurdle to climb, something that those knee-deep in UZR’s would do well to remember.

    Question: If your were a GM [for a normal team with budgetary constraints], which would you prefer?

  • Michael Jong

    Jorge,

    First off, I appreciate your tone in this discussion, as it mirrors the tone I would like to have when discussing anything with readers or anyone in general. Thanks, it is much easier to discuss this in this kind of tone.

    As for the concreteness factor, I agree that defensive stats are not as concrete as offensive stats over the course of a season. For the sake of this argument, I assumed that both players were average, though given the large difference in the numbers, I would venture to say they aren’t, and that the difference is significant. How much (in terms of runs, since we need to put a unit on it) do you think an average left fielder is worth compared to an average center fielder? Most of the research on defensive stats (using multiple years to study this effect) say the talent/difficulty gap is worth 10 runs, or about 1 win. This is more or less the difference between playing center field and catcher (catcher being more valuable). Does that sound reasonable to you?

    If it does, than we can use that as the defensive valuation. Offense is significantly easier to quantify. Using wOBA, after park correction, Coghlan was worth 4 runs more than McCutchen. Does that sound reasonable?

    If it does, and we then considered playing time (about two runs more for Coghlan, since playing time is measured as production above a bench player), we’d still have Coghlan behind.

    I think the problem here isn’t defense evaluation necessarily. A lot of us Marlins fans really like Coghlan, myself included. What we’re missing is that McCutchen was not so far behind offensively. There seems to be an idea that Coghlan was leagues better than McCutchen. Not even using wOBA, McCutchen had an OPS of .836, Coghlan had an OPS of .850. What do you estimate the run difference between those two is? Two runs? It’s not a big difference, even after you correct for park.

    I’ll be writing a piece on this to break down the argument between the two. I want to be fair to everyone involved in the discussion and make myself clear on my opinion, which is based on context neutral stats converted to production.

    I’ll answer your question, even though it has nothing to do with my RoY ballot (remember, I’m voting based on “production throughout the season,” not potential). I think Chris Coghlan could be something like 5-6 runs better on offense. I don’t think Chris Coghlan is a good defensive outfielder (he might be average at some point in the next few years, if he sticks at the position), while I think McCutchen is an average to good defensive center fielder. If they both have the same playing time, I would choose McCutchen is I were a GM, because CF is harder to fill than LF. If Coghlan is a slightly below average 2B, the argument is much better for Coghlan. But as a below average or even average left fielder, he would need more bat to make up for the position scarcity/difficulty.

  • http://www.2thinkgood.com Jorge Costales

    Michael – I return the compliment about this being an enjoyable discussion.

    I don’t dispute the reasonableness of how you present your case for McCutchen, my reluctance goes to the validity of the stats themselves – the process by which we arrive at the 10 runs/1 win.

    I’ll confirm your suspicions about us Marlins fans thinking Coghlan had much better offensive numbers. I was very surprised to see how close they were. I was all set to brag that Coghlan’s OPS was even >1 in August, before I saw that McCutchen was as well.

    One last shot/angle. I remember a while back from Bill James, after Orel Hershiser had something like a 29 inning scoreless streak earlier in his career – not his record setting streak in 1988. The question was whether Hershiser might just be a fluke. James response was that a ‘fluke’ could never have accomplished what Hershiser did, it was too difficult, he was the real deal.

    So let’s talk probabilities. I submit that there is a higher probability that McCutchen’s career will not see him develop into a quality MLB player as opposed to Coghlan. Rookies who average .388 over a 2 month span — sorry, but especially over the final 2 months of the season — are the real deal.

    Having said that, the correct answer was still McCutchen. See the GM question was a bit of a trap for extra credit, since Scot Boras is Coghlan’s agent.

    I look forward to reading your additional posts on this.

  • Pingback: Coghlan vs. McCutchen: Battle for Rookie Supremacy | Marlin Maniac | A Florida Marlins Blog

  • Simon Cowbell

    Your organization???

    McCutchen hit .226 late and close.

    How does that translate into a good clutch number.

    Please explain this.

  • Simon Cowbell

    Do you even know what numbers are going into your esoteric stats, or are you just taking someone’s word for it?

    How does a guy hit .226 late and close turn into a clutch hitter?

  • Michael Jong

    Simon,

    The organization refers to the Baseball Bloggers Alliance, for whom the ballot was cast.

    OK, let me try to explain this again.

    Late and close encompasses events that are both highly important and not highly important. It doesn’t encompass events that are important and occur early in the game. Does that sound reasonable to you? Can we agree that not all “late and close” batting events are of high importance, and not all important events occur after the 7th inning?

    Let’s cite an example: on 6/12/09, McCutchen singled with no one on and 1 out in the 7th inning with the Pirates down 3-1. The leverage index in that situation was 1.14. The average leverage index is 1. The change in game state added a 5.3% chance that the Pirates would win that game. This situation was more important than the average situation, yet the single only changed the Pirates’ odds of winning from 11.7% to 17.0%. But it’s included in “late and close.”

    Here’s another example. On 9/9/09, McCutchen doubled with runners on second and third with two out in the 2nd inning. The LI for that situation was 3.23. McCutchen added 18.5% to the Pirates’ chances of winning. That was an important situation that happened before the seventh inning.

    Not all situations are like that of course, but it is important to make the distinction that, by using “late and close,” you’re forgoing other important situations and sometimes including situations that aren’t as high impact. By using the split for “high leverage,” you guarantee that all plays included are high impact. It’s as simple as that.

    Simon, I ask of you to take the time and read a bit about leverage. Here’s a great explanation on the topic by Tom Tango. Leverage index and win expectancy is a great tool for establishing what is an important event. Since this is what we’ve been talking about, I encourage to check it out. If it’s not your cup of tea, I can understand, but I hope you find it interesting and of use to what you are talking about.

    If I’ve been light on explaining what the stats I’m using are, then it is indeed my fault for not explaining them well. I hope you read the explanation by Tango and check out the chart he provides at the end of the article. It’s good stuff, and he does an excellent job of explaining it.

  • Michael Jong

    Jorge,

    Given what I’ve seen of the two players, my opinion would be that the odds are more even than you’re giving, but you are weighing the most recent PA’s more, while I’m thinking that due to the .365 BABIP on the season, those PA’s aren’t as likely to stick. We can agree to disagree there.

    The key I think in what you’re saying is that this discussion is no longer about “what did they do this year” and more about “what we can expect in the future.” I’d be happy to have that discussion outside the purview of awarding a Rookie of the Year. For the RoY, though, I’m measuring production, and by my account, at the very most (average defense by both players) they’re even. My evaluation of Coghlan’s defense tips the scales in my opinion to McCutchen for this year.

    On your issue regarding the 10 runs/1 win, we can keep the denomination in runs, it’s not a big deal. Wins are calculated using Pythagorean expectation and generally seem to hold up until you get to the extreme run environments. Really, it ranges around 10 runs to a win; in a league average 4.45 run environment, adding ten more runs allowed translates to 1.042 wins, for example. You can just stick the numbers above average into Pythag for a replacement team and figure it out from there too.

  • Simon Cowbell

    Michael….

    I’m trying to figure out where all these key situations are popping up if McCutchen sucked Late And Close, sucked with 2 outs and RISP and sucked when the score was tied (at any point in the game).

    I think you are taking someone’s word for his great stats you are adhering to.

  • Simon Cowbell

    How do the funky stats address one guy’s team being in a playoff race, and the others playing spring training games from June through October?

    What is the adjustment there?

  • Michael Jong

    Simon,

    It sounds like you really didn’t bother to read anything I linked. Fair enough, it seems like we’ve reached some sort of impasse, whereupon you quote your stat, and I quote and explain mine.

    Truth be told, you’re sort of “taking someone’s word” on McCutchen’s production in “late and close” as well, since you haven’t gone through each individual PA and calculated it. You got it from a source, right? So did I, I got it from Baseball-Reference. Let’s not use that line, please. I’m not going to plow through play-by-play data to go find each high leverage PA and calculate it for you, just like you won’t go through every single late and close PA and calculate his line for me.

    McCutchen in Late and Close: .697 OPS
    McCutchen 2 out RISP: .822 OPS

    The late and close number isn’t good, the 2 out RISP is far from sucking. I won’t even talk about your tie game split, as any score splits incorporate so many low importance at-bats that it won’t be accurate. For example, a game is tied in the first PA of the ball game. Is that really an important thing to take into consideration? Third inning, one out, no one on, tie game, is that really the high-impact, “clutch” situation you want to quote?

    The only thing you have a leg to stand on is in late and close. I’m quoting a much more inclusive statistic that incorporates far more important situations. I’m not going to run down every high leverage situation for you, but you may check it out yourself here; the LI’s and events are right there for you to use if you’d like. For me, I’ll “trust” that B-R calculated these numbers correctly.

    Your point about the playoffs is also unimportant to me. I don’t think a player should be penalized for bad teammates. That’s entirely context driven and an unfair way to measure players. Is Ryan Zimmerman a worse player for playing for the Nationals?

    I think I know your response is going to involve quoting late and close again (and batting average only, of course, conveniently leaving out that, for example, McCutchen slugged better than Coghlan in those instances) and not respond to any of my actual points by reading into what I discussed. At this point, Simon, I’d say we should respectfully agree to disagree. You can read in my latest post discussing Coghlan vs. McCutchen in full, and if you don’t agree with me, it’s your opinion, and you’re welcome to have it. I tried to sway you, but it appears you won’t be swayed, and that’s fine. I hope it doesn’t deter from reading any of my other content in the future.

  • Simon Cowbell

    “I think I know your response is going to involve quoting late and close again (and batting average only, of course, conveniently leaving out that, for example, McCutchen slugged better than Coghlan in those instances)”

    Good to know that in your view that .697 is better than .743.

    “Here’s another example. On 9/9/09, McCutchen doubled with runners on second and third with two out in the 2nd inning. The LI for that situation was 3.23. McCutchen added 18.5% to the Pirates’ chances of winning. That was an important situation that happened before the seventh inning.”

    You have to make up your mind… either at-bats early in a game are important, as you state here, or they aren’t as you say here:

    “For example, a game is tied in the first PA of the ball game. Is that really an important thing to take into consideration? Third inning, one out, no one on, tie game, is that really the high-impact, “clutch” situation you want to quote?”

    Frankly, ANY tie score IS an important plate appearance in a game. Sorry to rain on your dismissiveness. Getting a lead is a big moment. Ask any MLB pitcher or manager.

    Coghlan’s OPS beat McCutchen’s by 46 points in L&C, 51 in 2 out/RISP, and 150 in tie-score situations.

    Yet, you are married to stats that involved subjective poppycock as “Runs allowed” as one of its formula components.

    The stats I quote are facts. Yours …. it’s like quoting figure skating judges. They interpret what they see and assign a subjective score to it.

    Runs allowed?

  • Simon Cowbell

    And regarding your “other” material…. your hard-headedness reminds me of the trogs with whom I have argued re: how John Smoltz had no business winning the 1996 Cy over Brown, or how Livan had no business winning the 1997 WS MVP.

    Utter insanity.

    I guess the next time Juan mentions something on this blog that intrigues me or raises my ire, I suppose I may come back.

    But, you need to do better.

    Thanks for the conversation, though, Michael.

  • Simon Cowbell

    Luckily, I didn’t even get into the ridiculousness of having Hanson and his puny 128 innings ranked ahead of Coghlan. Or Wells, with a good ERA, but, again, limited innings and a woodwork 1.278 WHP and 12 wins.

    This was not the year for part-rime pitchers with merely “good” ERAs to be considered for this award.

  • Michael Jong

    Simon,

    Some at-bats are more important than others, a fact that you mentioned yourself. The first at-bat in a game is less important than an at-bat with a leverage index above 2, for example. Now you’re telling me they aren’t? I’m not sure what you’re getting at.

    Regardless, thanks for the input Simon, and I hope next time you drop by to Marlin Maniac, we can have a more civilized discussion.

    On your final note, here are my opinions: Smoltz was a slightly better pitcher and pitched more innings, so he probably deserved his award. An MVP for a series is fairly irrelevant, but I’d venture to say that Alou was just as good if not better than Livan in the ’97 WS. Just my thoughts.

  • Simon Cowbell

    Smoltz was a SLIGHTLY better pitcher than Brown??? Brown’s ERA was MORE THAN a full run better.

    The only category that Smoltz dominated Brown in that year was run support.

    Talk about decimating one’s credibility.

    Or READING ability… YOU were the one pumping up an earkly-game stat for McCutchen as being important… then pooh-poohing the idea of early-game stats when it reflected well on Coghlan.

    Hell, I QUOTED you as such in the 8:54 a.m. post.

    I find your arrogance (ignoring the quotes of yours that showed your hypocrisy, for one) to be as uncivilized as any words I have used in this discussion.

  • Michael Jong

    Simon,

    I agree, ERA did say that Smoltz was a good deal worse than Brown. I took a look at the two pitcher’s FIP values, whcih evaluates them on the basis of K’s, HR’s, and BB’s and puts them in a scale of ERA, and Smoltz’s was slightly below Brown’s that season. Also, Brown pitched 20 less innings. By that measure, I went with Smoltz.

    Of course, both were pretty deserving candidates, I’d say. If you go by Rally’s WAR totals over at his Baseball Projection site, you get a Brown with a WAR of 7.5, while Smoltz had a WAR of 6.1. In that case, I would say Brown won. I’m not really sure which WAR method of measuring pitchers should be used.

    Hey, I agree with you that we shouldn’t just be looking at wins in this case. Brown had a ridiculously good season, and if we use Rally’s method of measuring WAR (taking earned runs and removing prorated defensive contribution) he is the clear winner. Having not checked that prior to saying my comment, I’d say I would retract my statement and lean towards the Brown argument. I only gave the question a cursory glance at first using FIP, though FIP says Smoltz’s defense-independent performance was better that year.

    As for your increasingly ad hominem attacks regarding me, you didn’t look closely enough at what I am saying. Important situations in games aren’t measured just by the base/out state or inning/score situation, but rather all of those aspects combined. Leverage tells you what situations are more important.

    Example:

    First inning, 1 out, no one on, tie game
    vs.
    First inning, 1 out, bases loaded, tie game

    Which one is more important? I think we can all agree it’s the second one. The second situation has a leverage in the 3′s, the other one has a leverage in the 0.8′s. Is one more “clutch” than the other? I think we can both say yes to that.

    So why use the “tie game” split when it includes both situations and, as you’re using it, describes them as the same level of clutch? They clearly don’t. That’s why I’ve been offering High Leverage splits as the way to determine clutch performance, because that split yields all clutch situations, regardless of the individual base/out states or inning/score states.

    Finally, if you’re taking it as arrogance, I apologize. I’ve tried to be as patient in my explanations as possible. You have increasingly been hostile in your tone, and it really has made it difficult to have an appropriate discussion. I’ve answered all your points every time you’ve posted, and from what I can tell, you’re simply restating your points while mostly ignoring mine. I’ve pointed out different sources for my thinking, which I’m sure you’ve also ignored.

    That being said, here’s a final source. Tom Tango has a leverage index chart for each base/out/inning/score state possible in regulation. You can look at it here, then you can tell me whether you don’t buy how the leverage split is done. The way B-R does it, high leverage is considered LI > 1.5. You can find all the types of situations that have an LI > 1.5 and see if they jive with how you think. If you don’t buy it, then again, we have an impasse and we’ll have to respectfully disagree on this clutch definition. Please check out the chart and tell me what you think.

  • Pingback: 2009: A Look Back At The Year In Sports(blog) | FanSided

  • Pingback: History: Coghlan vs. McCutchen 1.5 years later « Marlin Maniac | A Florida Marlins Blog