Friday, February 12, 2010

Larry Walker vs. Albert Belle

For your baseball pleasure, my take on two terrific sluggers whose careers were both cut short by injuries. I give my readers, as the title subtly suggests, Larry Walker vs. Albert Belle.

First, some basic stats. I went in knowing that Walker would have a big edge in the traditional counting stats because his career was longer. As it turns out, I was correct about that. My thanks to Fangraphs for some of the stats I list, and to the REALLY cool comparison tool on Baseball Reference

Walker does indeed have the edge in the counting stats, as I had guessed, like the standard HR, doubles, runs, RBIs, etc., including a big edge in total bases, 3904 to 3300. This seems almost entirely due to having a longer career, 17 seasons (or parts thereof) to 12 for Belle - but that's still a big difference. Walker also has the edge in career OBP, with a fantastic career OBP of .400 (including a terrific EIGHT years in a row over .400), while Belle sits at merely .369 (including only four years over .400, with another year at .399). Right off the bat, Walker has a significant lead.

But Belle has a counter-punch or two to make. Here is a list of Bells's adjusted OPS (adjusted for park effects but not for position or league differences – 100 is average) by year, starting in 1989:

85
42
134
122
145
193
177
158
116
171
142
109

Career average OPS+ for Belle is 143.

For Walker, coincidentally also starting in 1989:

26
112
127
142
120
151
130
116
178
158
163
110
160
150
121
153
166
143
130

Career average OPS+ for Walker of 140.

Belle has the edge in career OPS+, AND played in the tougher league, making his higher OPS+ number slightly more impressive. Belle benefits by not having significant decline years, however - that props up his average stats while harming his counting stats. Furthermore, in a somewhat silly comparison but I think a useful one on some levels, there is the Black Ink Test on Baseball Reference (number of times they lead the league in a category, as follows: Four Points for home runs, runs batted in or batting average; Three Points for runs scored, hits or slugging percentage; Two Points for doubles, walks or stolen bases; One Point for games, at bats or triples) and the Gray Ink Test (number of times in the top-10 of a category). Belle has Walker beat in both. UPDATE - As pointed out by frequent commenter Dr. Brainsmart, it is worth noting that Walker had to contend with Barry Bonds (one of the top 5 hitters ever, no matter your opinion on how he got there) in his league so the league leader stat may be even less significant than the small weight I previously gave it.

Walker's WPA (win probably added) was 48.88 over his career, while Belle's was only 26.04, again likely due to playing time. Score a jab for Walker. And with a nice uppercut for Walker, according to this site: his career WAR (wins above replacement) is 67.1, ranking him 67th in baseball history (among position players). Belle is way down the list with 37.1 career WAR, ranking him 318th among hitters.

I also don’t have access to advanced defensive metrics like UZR, as that also only goes back to 2002, or +/- because that's a proprietary stat and I'm not a member of that site! I think it’s fair to say that Walker was a better defender, as for much of his career he considered a very good right fielder, while Belle was considered an average (at best) left fielder – obviously an easier position to play.

Walker gets a better "speed score" from Fangraphs as well, although I'm not sure what that encompasses. Let's just say that I'm willing to bet Walker was more valuable on the bases than was Belle.

Overall, it's pretty clear that Walker had the better career. But I also think it's fair to say that Belle had the higher peak and just didn't sustain his excellence as long. Belle's career wasn't as long but he was seemingly more durable during the years he played – his Games Played is consistently higher than Walker's. So while Walker has the better case for the HOF (but will almost certainly not make it due to the Coors Field stigma and his injuries keeping down his counting stats), it's a closer call as to which player you'd rather have for their 5-year peak – I might say Belle on that question.

Of course, if there was a statistic for which player I'd rather my hypothetical daughter date then Walker would win in a landslide. Belle was, by all reports, a pretty big jerk, while Walker was seemingly affable and goofy in a friendly sort of way.

11 comments:

  1. I can’t find a lot to argue with here, which is sort of a pity, and even a little bit confusing. But it does bring to mind an interesting question (or, at least, a question that interests me): how much of a factor should career longevity be when comparing players, or simply when talking about a player’s greatness?

    Belle vs. Walker is a great example. Belle looks to have slightly better peak years, at least offensively, but Walker played a lot longer. One could argue that Belle would have put up similar or better numbers than Walker if his career had lasted as long, but one could also argue that if my aunt had testicles, she’d be my uncle. The fact is that Belle’s career didn’t last that long, and Walker’s did. And while Walker wasn’t exactly a model of durability, he did play (parts of) 17 seasons in the majors, and it's dismissive to simply say "well, he got those stats only because he played so long" (for the record, I'm not suggesting that Squid Bandit said this). I think it's better to say that he earned those stats because he was good enough to play for that long.

    I think that definitely counts for something . . . but I’m not sure how much. If I had to pick between Belle and Walker, it’d depend on the situation: I’d probably choose the ca. 1995 Belle if I needed a guy for one plate appearance, and the ca. 1986 Belle if I needed a guy to obliterate a clubhouse toilet with a baseball bat. But I’d have to say that Walker’s similar (if slightly weaker) average stats, plus fragile longevity—even without my bias towards Canadians and disappointed hockey players—gives him the edge in my book.

    Maybe longevity shouldn’t be a huge factor—Jamie Moyer was born before Sandy Koufax, I think, and Moyer is still playing while Koufax has been retired since 1966 . . . but few people are going to say Moyer’s longevity tips the scales and makes him a better pitcher. Phil Niekro, another durable guy, has almost twice as many wins as Koufax, but only a true Niekro-philiac would believe that this makes Niekro twice as good as Koufax.

    So what does this mean? I don't know. Where does Gordie Howe (32 seasons of pro hockey) compare to Wayne Gretzky (a mere 20 years, but almost 55% more points), when durability is factored in? Probably more or less where he does already . . . but I think it makes for a good, if ultimately pointless, argument.

    ReplyDelete
  2. Yeah....not sure what you're saying there, Some Guy. Did you pick the wrong day to stop sniffing glue? I explicitly stated in my post that I Walker was better over his career, and I would certainly draft him over Belle knowing how their careers turned out.

    And you're right, I didn't say "he only got those stats because he played so long". I said that the difference in their counting stats was likely because of the difference in their career length. And that extra playing time EARNED by his talent does matter.

    To put this in a Dodger-centric way (as I love to do): everyone would take five years of brilliance within eight years of very goodness (Koufax) over twenty years of average (Don Sutton). It's harder when it's a closer call, like Walker and Belle. Another guy I might look into was Juan Gonzalez, but my gut tells me he was not as good a hitter as Walker or Belle ever were.

    A few other things:

    1. Gordie Howe couldn't hold Gretzky's jock strap (he'd probably beat The Great One up with it and leave him to bleed, but Gretzky was much better at the non-fighting aspects of hockey); and

    2. How long have you been saving that "Niekro-philiac" comment to whip out at the right moment?

    3. Larry Walker, while certainly awesome in 1997, wasn't the true MVP of the NL.

    ReplyDelete
  3. Squidbandit wrote:

    "Belle has the edge in career OPS+, and played in a tougher park for hitters and in the tougher league, making his higher OPS+ number that much more impressive."

    But, you had just written that OPS+ is adjusted for the park... how can you turn around and say that Belle's small edge there looks "much more impressive" due to each player's home park?

    Furthermore, OPS+ IS adjusted for league factors as well. Considering this, along with your correct conclusion that Belle's lack of late-career decline (because his late career was spent limping in and out of various bars and courtrooms) helps his averages when compared to a longer career such as Walker's, makes the small difference in career OPS+ between the two players not really that impressive at all.

    The "Black Ink/Grey Ink" thing is, as you said, interesting but of questionable usefulness - it doesn't say as much about a player as it does about those he played against. Walker was in the NL while Barry Bonds was in his natural prime (not to be confused with his second, "bonus" prime), and there's black ink all over Bonds' stats. Is it really fair to say that Belle's better than Walker because Walker wasn't better than Bonds?

    I also find it interesting that you cite the Black/Grey ink numbers from BaseballReference.com which favor Belle, yet ignore the "Hall of Fame Monitor/HOF Standards" numbers listed right below them, numbers in which Walker exceeds Belle, quite easily in the Standards case.

    As hitters, they were close, and I suppose due to Walker's durability issues, you could effectively argue that Belle had a better 5-year peak at the plate than did Walker. But as an all-around, player, there is absolutely no comparison between the two. Walker was one of the finest five-tool baseball players of his era, while Belle was merely a slugger.

    Finally, I'd be interested to know who was the "true" MVP of the NL in 1997. Walker was Ruthian in 1997. He had an absolutely monster year, one of the best seasons in baseball history... and before blaming it on Coors, look at the splits for '97. He hit over .340 on the road, and he actually had MORE power on the road than at Coors that year. Piazza? He had an incredible season in 1997, too, but there wasn't one major stat where he exceeded Walker.

    ReplyDelete
  4. I didn't mention a lof of the career-oriented stats, as I determined that they all favored Walker. Indeed, my conclusion was that Walker "clearly" had a better career. But if you think there is no point in comparing the two then you missed the point of the post.

    OPS+ does NOT take into account league differences. It does take into effect park differences, and my bad for making that error. I will edit that right now.

    No major stat where Piazza exceeded Walker? Are you talking counting numbers? Probably true. Good thing we have all sorts of advanced stats that take into account everything, right? In 1997 Piazza had many more Runs Created, a higher RAR (runs above replacement) and a higher WAR (wins above replacement) than did Walker. And WAR takes defense into account, and Piazza was a fairly poor defensive player. Piazza was more valuable overall, and particularly moreso with the bat, than was Walker.

    ReplyDelete
  5. You're either wrong or confused about OPS+, it takes the player's OPB and SLG and divides them by the average OPB and SLG for the league, and then works in park factor. Meaning that a guy with a 100 OPS+ in either league is the average of that league.

    Now, I guess you may be arguing that the average AL hitter is better than the average NL hitter, and thus Belle's 143 is actually more than three better than Walker's 140. And if that's the case, you're correct in that OPS+ does not attempt to equilibriate between the AL and NL, but there's a reason it's done the way it's done. Any time you see a higher league average OPB or SLG average in the AL over the NL, you only need to consider that pitchers don't hit in the AL, inflating all league-wide AL averages in relation to the NL... which is exactly why OPS+ compares each player to his own league rather than to both leagues. Doing it any other way is a distinct disadvantage to NL hitters, just as comparing pitching stats across the two leagues to the other league's average pitching stats would be a major disadvantage to AL pitchers.

    The fact that you can cherry pick three or four stats in which Piazza bested Walker in 1997 is hardly an argument that he was the MVP of the NL that season, because there are three or four more that go Walker's way (on top of every single "traditional" stat on the list). And one of the stats you quoted is wrong... according to baseballreference.com, Walker led the league in 1997 in the Runs Created with 187, to Piazza's 150.

    Using the "black ink" thing for 1997, Walker led the league in 13 categories tracked by Baseball Reference.com. Piazza has black ink on one stat - OPS+. While I don't doubt that there are other stats out there (Piazza's 9.3 war over Walker's 9.0)

    ReplyDelete
  6. If MVP stands for "Mantastically Vile Porn-stache," then Piazza was definitely robbed in 1997, and, for that matter, in pretty much every year he played.

    ReplyDelete
  7. ...whoops, my cut-but-forget-to-paste editing technique caused the conclusion of my Walker/Piazza argument to vanish.

    (continued) While I don't doubt that there are a few stats in which Piazza finished ahead of Walker, to pick and choose a couple of them such as WoRP or RAR at the expense of all other stats at our disposal is just faulty logic, and the exact opposite of what SABR-type numbers are all about.

    All Value Over Replacement Player stats are useful at comparing a player's production in a season to that of other players at his position, but are not as applicable when comparing players at different positions, especially when one of those positions is catcher. When we see that Piazza's Wins over Replacement was 9.3 and Walker's was 9.0, that doesn't really tell us that Piazza was more productive (or more valuable) than Walker, only that Piazza was better than the average catcher to a slightly larger extent than Walker was better than the average RF.

    When we consider that the average catcher is a far less productive player than the average RF (a fact mathematically accounted for in VORP equations), we see that even though Piazza's VORP stats slightly exceeded Walker's, Walker could have been (and was, as nearly every other stat demonstrates) a more productive player overall. For those who equate production with value, that weighs heavily in Walker's favor for MVP.

    If you want to show that Piazza was the Most Valuable Catcher in 1997, you have nearly infinite stats to back you up, because he had an amazing year; little question it was the best-ever offensive season for a catcher. But MVC doesn't equal MVP.

    It seems you've found the two or three stats in which Piazza led Walker, taken them largely out of the context in which they were intended to be used, and then claimed that those two or three stats which Piazza has the edge are the Only Ones That Matter, and therefore Piazza was the true MVP. But your misunderstanding and misuse of those stats only bolsters the case for Walker.

    ReplyDelete
  8. I must be confused - I see a greater level above average over a higher average level and I see better performance.

    Your argument is logically faulty. The fact that pitchers bat in the NL lowers the average, so good hitters should be even MORE above average in that league. Not less, like you said. You are confusing the reason for using league average. It's not to normalize for the pitchers (I think I read somewhere that pitcher ABs are not included in the NL, but I can't find it now and could be wrong on that), it's to normalize across eras in baseball. So one can compare Barry Bonds with Babe Ruth, or Alex Rodriguez to Honus Wagner

    WAR includes positional value for sure, and that is one factor that made Piazza more valuable that year. But it does NOT only demonstrate how good that player was within his position. For one who is arguing that I'm misunderstanding the statistics, maybe you need to do more research on what the stats mean. What it means is that Piazza was worth more wins to his team than was Walker.

    I don't think any one stat is the Only One That Matters. But yes, I believe WAR is much more valuable than how many home runs someone hit.

    Regardless, they both had terrific years, each worthy of praise. Thanks, as always, for your bright and sunny commentary Dr. Brainsmart!

    ReplyDelete
  9. "Your argument is logically faulty. The fact that pitchers bat in the NL lowers the average, so good hitters should be even MORE above average in that league. Not less, like you said."

    I believe it's you, sir squidbandit, that has a logic issue with this one, although I see what you're getting at. The OPS+ stat is adjusted for league... which means that when Belle's is 143 and Walker's is 140, it tells us that Belle was slightly better when compared to the average AL hitter than Walker is when compared to the average NL hitter. I think we're square on that.

    But where you run into trouble is when you then argue that Belle be even better than the gap the OPS+ numbers suggest due to the AL's higher average. This makes sense at first, but it is precisely what's already taken into consideration in the OPS+ rating. The average for the AL is higher, but that does not tell us that the individual player averages are higher in the AL... just that there are more high averages.

    OPS+ compares a player to a league-average player, which is not the same thing as comparing him to the average of the league. So I think you're double-dipping when you say that an AL player's OPS+ rating is of a higher value than an NL player's, when the leagues are already normalized in the stat to begin with.

    As far as Piazza and Walker's 1997 WoR, I did not mean to imply that the stat is not useful for comparing one position to another. They are weighted to compensate for positional differences... I only meant that because an attempt is made to mathematically equate a catcher's value and a RF's value, they cannot be the precise and objective number many claim them to be.

    One big reason I doubt the RaR and WaR advantages you cited for Pertboy is that when looking at both guys' offensive outputs that season, I have a very hard time believing that Piazza's bat outdid Walker's to such an extent -- even after adjusting for park factors -- that it would totally overcome Walker's obvious superiority as a defensive player and a baserunner. A close look at what's factored into the stat bears this out.

    I'm assuming you got the WaR and RaR stats from Baseballprojections.com. This site has what I consider to be a major problem in their computation of these two stats - it gives Piazza a +8 rating for "position adjustment," defined as "how players perform at multiple positions." It gives Walker a -7 for the same stat. The only other "position" Piazza played in 1997 was seven games at DH... hardly worth giving him bonus points for. Walker played two other real "positions" (CF and 1B) that season, in addition to one game as DH. That glaring problem alone could have completely altered the final WaR numbers.

    ...which is exactly my point about them. Any stat you find, even "advanced" stats, are subject to bias and/or mistakes. Looking at overall stats, its tough to argue against Walker's MVP. You've done so, but considering that the there appear to be major issues with the way the RAR and WAR stats were computed for the two, as well as the fact that the runs created stat you quoted was simply wrong... I think your argument is more emotion than proof.

    ReplyDelete
  10. ... Ok, so I've gained a better understanding of the "position adjustment" number. Piazza didn't get a "bonus" of 8 for DHing, that number's actually lower than it would have been had he played every game at catcher... he actually got dinged for being a DH. That number is an attempt to equalize a player's defensive worth based on how many games he played various positions.

    Since catchers are typically the best defensive player on a team, a catcher's defensive stats are weighted positively, while a RF's are weighted negatively, because the RF is often a butcher defensively. Like Brad Hawpe, for example.

    Then they factor this general "positional value" number in with a player's actual defensive rating, in an attempt to balance how a player actually played his position with how the "average" player would have done.

    But even though I misunderstood the number, it doesn't alter my problem with the way the rating is computed: it mathematically devalues RFs because most RFs are lousy defenders, while giving a higher value to all catchers because catchers are typically good defensively.

    The first problem here is that it's not Walker's fault he was a better athlete than your typical RF, but yet he still gets dinged for the generally sub-average play of his contemporary RFs. Similarly, Piazza gets bonus points simply for being a catcher, despite the fact that he barely reached "average" on his best day. I realize what the creators of this stat were trying to do here, but in a so-called "objective" stat, it's just ridiculous to throw in something so subjective.

    The second problem is that the site lists Piazza's total zone rating as zero - not only for 1997, but for every year of his career except 2004. For anybody who actually saw the guy play defense, there's definitely something fishy about that. This seems to indicate that the resultant RaR and WaR numbers were boosted simply by the fact he played catcher, without ever being balanced by anything indicating how well he actually played it.

    Perhaps it's just a typo on this web site. But if not, it really raises questions about whether these particular stats, for this particular player, really did factor defensive play into the final result.

    ReplyDelete
  11. Dr. Brainsmart is close, but still not there. The bonus has nothing to do with your assertion that catcher is the best defensive player on the team (he's almost certainly not, by the way - the SS is). It has to do with positional scarcity. Basically, the argument goes, Piazza plus the average RF is worth more to the team than Walker plus the average C. Thus, Piazza gets a positional bonus for being such a ginormously plus bat at a terrible-hitting position.

    ReplyDelete