Brian Kenny MLB network...picking Keith Hernandez over Eddie Murray in top ten first baseman. Hmmm.
1948_Swell_Robinson
Posts: 1,923 ✭✭✭✭✭
I'm watching MLB network do their "Top Ten" players for each position during the division era. It comes to first baseman and Kenny omits Eddie Murray from the list and has Keith Hernandez instead. Bob Costas had Murray as #5 on his list. Not going to really argue the spot...but since these two players are good examples of how a shorter career can give a false perception of the player with the shorter career being better, due to his rate stats staying higher, I thought I would highlight that, along with some of the points Kenny made to support his selection.
First thing is that Kenny used Hernandez's defense as his 'sabermetric' based type analysis to give Hernandez the edge over Murray, which is quite puzzling considering that the typical sabermetric view is that first baseman are not much better than statues, and that he himself makes that a point later on. I do give Kenny a tip of the cap for recognizing that first basemen are not statues at least.
But he basically said that Hernandez's defense was "game changing," because it forced teams to "not bunt" because of Hernandez's arm. That was his main basis for putting Hernandez ahead of Murray. I am puzzled at that since his stance on bunting is usually that it is "not worth doing" since it is viewed as sub optimal in the sabermetric community. Wouldn't that make it bad for Hernandez's impact then since teams are NOT bunting on him? I'm not sure which of his stances to believe on bunting, but I will say that Murray was no slouch on defense either and his arm was as good as Hernandez's. I'm not going to bother looking, but I am quite certain teams still bunted on Hernandez, and if his 'presence' in the handful of situations that caused a team to decide "not to bunt", did happen, I'll take Kenny's word at it and award Hernandez 'one run saved'...as that is what a sabermatrician would give generously as the value of executing several bunts.
Certainly at the heart of the sabermetric community is Hernandez's lifetime 128 OPS+ compared to Murray's lifetime 129 OPS+, where they may be viewed as 'equal' enough offensively, that Hernandez's bunting prowess of saving a run a year could give him the edge over Murray.
What many people mistake when looking at rate stats is looking at career length and players who play every single day vs every players who don't.. Ken Phelps had a tremendous OPS+(133) higher than both Hernandez and Murray, but he only played primarily vs RH pitchers and didn't play very long(2,287 career plate appearances).
The thing about Hernandez is that he was not good enough to play full time as a MLB player very long. Once he turned 35 he was no longer viable to have as a full time player, and by age 36 he simply was not good enough to play. I will give him his due for playing mostly full time when he was good enough, sitting against the occasional lefty to optimize his percentages.
Once Hernandez was fired, he wasn't any better of a MLB player than the guy sitting on his couch watching....so I ask then, how can a guy who is not good enough to play be considered "Better" than a guy who IS more than good enough to actually still be employed as a full time MLB player?
Hernandez retired with a 128 OPS+ over 8,553 career plate appearances. Excellent for sure.
Murray bettered that. He had a 139 OPS+ over 8,480 career plate appearances. Even MORE excellent than Hernandez.
So how can Hernandez be viewed equal to Murray offensively. You see above that Murray is 139 to 128...clearly better.
I will tell you why, it is because analysts do not consider career length when looking at percentage based stats. You see, Hernandez was fired at that point...relegated to being an observer like everyone else. Yet Murray was still one of the very few elite humans that were employable as a MLB player.
Murray CONTINUED to play. He had already bettered Hernandez's career. He was about to embark on a 'second' career as an old man. Murray was able to amass 4,337 more plate appearances at the MLB level with a 111 OPS+. Do you know what they call humans who can produce a 111 OPS+ in MLB for 4,337 FULL TIME plate appearances?
Millionaires.
When you add Murray's 'second' career to his first, it averages out to a 129 OPS+ over 12,817 plate appearances....far more valuable than the same production over only 8,000 plate appearances.
He bettered Hernandez in their equal first careers...and then dwarfed Hernandez in the 'second' career because Hernandez was not even good enough to play. Unless one views Ken Phelps superior to both of them, Murray is vastly ahead of Hernandez.
Sabermetric community tries to define defense value, but they typically fail...so I will leave that as a topic for another day.
There are other measurements other than OPS+. The traditional measurements such as career RBI leave no doubt, and those have merit as well, but are just icing on the cake. One measurement that measures the impact on winning is Win Probability Added.
Murray WPA is 51.7 for his career. That is 51.7 wins above an AVERAGE player.
Hernandez WPA is 38.6 for his career.
That is vs AVERAGE player. If you make that vs replacement player, then that gap widens even more. You can't just find league average players on a whim. They cost a bundle of money.
There is no amount of first base defense that can close that gap. Plus Murray was exceptional defensively as well.
There really is no valid reason to select Hernandez over Murray for any list. Costas got it right. Kenny...a little too contradictory in your philosophies...and just wrong in every measurable way. Hernandez DOES have some merit for the HOF, but that is another topic. Hernandez was outstanding, but just not as good as the real game changer.
I do want to talk about his "game changing" comment, because one of those players really was game changing, and it was Murray. From 1982-1985, Cal Ripken played every game, and Murray batted behind him almost every game. Cal Ripken was good in case nobody noticed. He won an MVP(one that when looking deeper really belonged to Murray. That is topic for another day).
Do you know how many Intentional Walks Cal Ripken received in all those games when Murray was hitting behind him?
ZERO. A big fat zero. That is game changing. In his prime, Murray wasn't given the chance to beef his numbers because they simply pitched around him in the most optimal times to put crooked numbers on the board, both in standard and sabermtric(which is also a topic for another day).
Comments
I would have Hernandez behind Murray and ahead of Perez.
m
Fellas, leave the tight pants to the ladies. If I can count the coins in your pockets you better use them to call a tailor. Stay thirsty my friends......
I think the best way to evaluate a "better than" comparison is to take into account career totals and peak numbers. Lots of ways to do that, and there is no right way, but one way I like is to line up each players seasons from best to worst and give the most weight to the top season, a little less weight to the 2nd best season, and so on. I mean when we think of how great a player was, what they did in their 12th best season really doesn't enter into it except to the degree that it added to their career totals, and any players place in history is going to rise or fall on their best seasons. Harold Baines may have made the HOF, but he never had even a single great season, and history will not remember him at all.
For the three players you mentioned, their career Win Shares are Murray-437, Perez-346, and Hernandez-311, and these are their Win Shares in all of their seasons, sorted from best to worst:
If you apply a weight of 15 to their best season, 14 to their 2nd-best, etc., and a weight of 1 to every season 15 or beyond (and divide the sum total by 7.5) you get
Murray 446
Perez 402
Hernandez 402
Perez did play a lot longer than Hernandez but he was playing at replacement level and those years really don't (and shouldn't) add much to his legacy. I'd still give the nod to Perez but it's really a coin flip between him and Hernandez. Murray, conversely, is clearly better than either of the others and anyone not listing Murray in their top 10 first basemen should not be allowed to discuss baseball in public. I think Perez suffers in our memories because he didn't really get famous until the BRM years, and his career was solidly on the downside by then. All of his great seasons came between 1967 and 1973; in the Reds WS-winning years he was the worst regular player on the team (but still good).
For context for anyone not familiar with Win Shares, 30 marks a season that will get MVP attention, 25+ is great, 20+ is very good, 15+ is good, 10+ is average, and < 10 (for a full-timer) is pretty bad. Eddie Murray was very good (20+ Win Shares) for longer than all but the very top tier HOFers, and longer than even some of those.
I always thought that OPS+ points should be deducted if a player doesn't hit the threshold of what an average amount of games played are.
It's great that you had a good two months, but not better than a guy who plays everyday and has slightly lower numbers.
Kind of like the guy who makes 200 widgets at work on Monday and then takes the rest of the week of and he's considered "better" than the guy who shows up every day and averaged 150 a day. Who do you want on your team?
By the way, we fired the guy for not showing up.
or not showing up.
You are discounting Perez's contribution during his old man years and relegating it to basically zero. First, he was not merely hanging on. Second, he was a better hitter than half the guys in the league even as an old man.
Third, does he get any credit for his contribution as a mentor to younger players in the clubhouse when he was an old man and could still hit better than most of them? Leading both by EXAMPLE and words, It carries a little more weight as a mentor when you are an old dude and can still kick the young dudes butt. What was Hernandez doing from age 36-43? Nothing. How is that more valuable than what Perez did?
Perez could have retired at age 36 like Hernandez and end with a lifetime 128 OPS+. The same as Hernandez.
Then they would be equal as hitters. Perez actually more valuable because walks from a middle of the order hitter carry a tad less weight. Not a ton less weight, but certainly a little less. It won't make Joe Carter as good as Schmidt...but in a case here with Perez and Hernandez where they are basically equal in OPS+, it puts Perez as the better since Hernandez's OPS+ is carried more with walks, and Perez's is carried more with Home Runs....which is preferable for a guy hitting 3rd-5th.
There is also value in being the anchor in the lineup like Perez was, and certainly like Murray was.
But how does Perez get discounted because he was good enough to contribute after age 36, even if it was at a lower rate than his prime years?
From age 37-44 Perez accumulated 2,592 more plate appearances at a rate of a 103 OPS+. That is NOT merely hanging on. Finding a human that can hit at league average at the MLB level...that is GOLD. That is rare. Finding one that also has vast experience, knowledge, respect, and admiration from his teammates is a triumph.
Perez is superior to Hernandez and in the HOF rightfully so.
I agree. And I don't think of it as "discounting" Perez's old man years with this analysis, what I'm doing is combining career value and peak value into a single number. For HOF purposes it is undeniable that voters have always placed at least as much importance on how a player did in his best years as they do in the career totals.
But, Perez spent his old man years as a replacement level player, not a good player. The one good year he had, 1985, he was Ken Phelps, sitting against RHP and batting against LHP. That season does need to be discounted in some way. Since he played so little it fell to the bottom of his Win Shares list so I don't think any additional discounting is needed. But as offended as you are by hitters who sit out the tough pitchers, you can penalize Perez more than I did if you'd like.
I always thought Perez was under appreciated. The Reds trading him after the 1976 season is one reason the Reds didn't threepeat.
http://www.unisquare.com/store/brick/
Ralph
Costas has Jim Thome and Eddie Murray ahead of Willie Stargell.
I can't go along with that.
They're all very close. Stargell's peak is easily the best of the three, and if that's what matters most to you then it ends there. But Stargell's career was the shortest, and he platooned and rested a lot to get it as long as it was. Murray was the only one of the three who played first base well, which is worth something, and also the best baserunner, which is worth another something, and maybe enough to get him to the top in a very close call like this one.
I'd rank then Murray, Stargell, Thome but it wouldn't take much to convince me to flip Stargell and Murray. And either way, the gap down to Thome is very thin.
Dallas, where is the WinShares resource? Are they still being updated or is it just the ones from when they came out? Either way, where can those be obtained?
When comparing Perez's old man years where he was better than league average hitter for over 2,500, plate appearances, I most certainly will not be discount them more when comparing it to a player who was not good enough to play and sat on his couch for 3,000 at bats. If you would like me to, I will, but in order to make it a valid comparison I'm going to have to add 3,000 worth of plate appearances to Hernandez's line at an OPS of .530, and that would be generous since an OPS of .530 is better than someone not good enough to play who is sitting on his couch watching. Nice try, but sorry.
However, if I am comparing Perez to Murray, yes, I would discount him for his part time status in some of those years since Murray was full time status for nearly every single plate appearance of his career.
Somehow it is gone, but I highlighted how in fact Steve Garvey WAS superior in hitting RH more than Teance was, even more so considering that Garvey did that over full time at bats and Tenace not. Also, walks are indeed a tad less valuable for middle of the order hitters, so that is another knock on Tenace in comparison
I believe they are still calculated but you have to pay for access. Theoretically, you could calculate them yourself because the formulas are in James' book Win Shares but you'd need to collect a whole lot of data and have a whole lot of time. It is not possible to calculate Win Shares for just one player, the process starts by calculating league averages for multiple metrics, proceeds through allocating Win Shares to a team's offense and defense, allocating those Win Shares to each position on the team, and finally allocating those Win Shares to each player who played that position. At each step the allocation process is insanely complicated, but each team's total Win Shares always equals 3 times the number of games they won. But most of the insanely complicated part deals with defensive Win Shares. The offensive Win Shares are still very complicated, but the foundation of them is the Batter Runs stat that I have referenced several times. For non-skill position players, playing a comparable number of games and PAs, comparing Batter Runs is a reasonable proxy for comparing Win Shares.
I just get them from Win Shares (the book) which runs through the year 2000.
Dallas, I enjoy your posts. I agree with almost everything you say. However, sabermetrics is wrong on several things, and this is one of them. WinShares will fall to the same fate even if it not a rate stat, because as you see below:
I posted this elsewhere but put it in my own thread.
1978 Gene Tenace had 25 Batter Runs
1980 Steve Garvey had 19.4 Batter Runs
On the surface it looks like Tenace was better. However, look at their at their PA vs RH and LH
Player.......RH......LH
Tenace....343.....172
Garvey....592.....112
Does anyone need to look further to see why that lead in batter runs is wrong? I don' think so. Just do the same exercises I did above with OPS, it comes out to the same conclusion. Garvey was a better hitter.
Ken Phelps had 3.3 Batter Runs in 1985
Eddie Murray had 3.4 Batter Runs in 1991
Plate Appearances:
Player.......RH.....LH
Phelps.....129.....11
Murray....360....279
Yet they are equal in your methods??? Get outtta here. Are you kidding?
Steve Garvey Batter Runs
1985 he had 5.5.
1986 he had NEGATIVE 10. Negative 10.
In 1985 Garvey played EVERY game and only had 5.5 batter runs. He was penalized by playing every game. So if I somehow made him a platoon player, his batter runs actually GO UP. They go up if he plays less because I will not play him against Nolan Ryan and aviod those nasty 0 for 4's that make your batter runs go down. So the bench guy goes 0 for 4 instead.
Same for 1986. Garvey's OPS vs LH in 1986 .838! Again, if I play him in just half of his at bats he had vs. RH pitchers, then his batter runs climb out of the negative for that year, and that would be equal to Tenace situation. If I ONLY play him vs LH, then his batter runs climb way up, and that would be Ken Phelps.
In 1986, Tim Pyznarski played for Garvey in the few games Garvey missed. HIs OPS was .581 and OPS+ was 64. So in essence, by Garvey playing every day vs the tough RH pitchers, he is saving his team from having to play Tim Pyznarski....even though by Garvey playing it is dragging down Garvey's OPS and Batter runs into a slight negative, making him look worse of a htter than Tenace who did the opposite.
In 1978, the primary First Baseman for the Padres in lieu of Tenace was Broderick Perkins and his OPS at 1B was .605.
In 1978, the primary catcher for the Padres in lieu of Tenace was Rick Sweet and his OPS at C was .603.
You see, when a guy like Garvey plays EVERY DAY, your team does not suffer by having to play Pyznarski, Perkins, or Sweet. So when Garvey goes 0 for 4 vs Ryan, he is just saving the inevitable of the reserve going 0 for 4.
When Tenace doesn't play in the tough matchup or at an old age, the team suffers with those scrubs, and Tenace avoids the 0 for 4 on his batting record, thus keeping his OPS and Batter runs safe for when a more opportune time to play occurs.
Granted, a team could have a viable option, but that costs MONEY. It also keeps that viable option from playing in someone else's stead.
Garvey playing EVERY DAY helps WINNING, despite the fact that it may knock his Batter RUns into the negative.
What was Gene Tenace doing in 1985 and 1986? He wasn't good enough to play in MLB.
Garvey 1985/86 NEGATIVE 4.5 Batter runs as a complete FULL TIME player.
Tenace 1985/86 Zero batter runs while he sat on his couch.
Batter Runs is telling us that a guy sitting on his couch at home was a better hitter than a guy who was just a tick below a league average MLB hitter. Keep in mind, that he really wasn't below league average, because he played vs every tough RH pitcher. If I give Garvey the Teance or Phelps treatment in those years then he is above league average hitter or WAAAY above in the case of Ken Phelps treatment.
I really appreciate the fact that since being exposed to dallas' opinions, I have learned a great deal about how I evaluate sports players.
One of the things I read is that some of the numbers he relies on are not good in using to compare players.
For example; when a batter with the highest batting average doesn't get enough times at bat he doesn't get the credit for the Batting Championship". Sometimes the wrong guy gets to win, sometimes it's fair and sometimes it's absolutely correct.
There should be an OPS-C stat "minus couch" if a guy plays more than the league average games in a year he should be rewarded and if he plays less he should be punished. This would be a MUCH better indication of his contributions to the team.
I like the OPS-C "minus couch" stat. But remember, even being five percent lower than league average is still vastly better than the couch. There is a point though where if a guy is soooo bad in MLB that the couch may be preferable, because you do have to pay that guy for being so bad. Garvey was never that player though, as outlined above.
Some of that may go on Garvey's manager. Garvey may have been able to play a few more years if a smarter organization used him more like a Cliff Johnson at that point of his career.
As it stands, there are still very few players in MLB who were able to handle full time at bats in MLB like he did, and those very few players are called Hall of Famers.
Pete Rose's last 2,500 plate appearances he had an OPS+ of 86. When you consider he was a first baseman and only hit 5 home runs in that time, then the couch is beginning to become a viable option, because power from that spot is important, and power/Total Bases from a middle of the order hitter is more important than walks from those guys, even though the walks are still of value, just a tad lower from middle of the order guys. Though I know Rose wasn't middle of the order hitter.
There is no exact line to draw and say, this is where the couch is. Garvey never crossed that line until his last year where he only had 76 at bats and the couch became the right call....unless he was salvaged by becoming a Cliff Johnson type player.
Thank you, and if you can peel away all the pomposity and masturbatory rehash, this has always been my primary goal. It is impossible to have a meaningful discussion about baseball players if people are speaking different languages. Bill James and the other stat geeks aren't always right, but they are always speaking the same language and they are using words that the old-time fans simply don't know. And whether they ultimately agree with James, et. al. and their conclusions or not, people who don't understand the words the stat geeks are using will always reach wrong conclusions themselves.
For Tenace, and this will be my last word at least for awhile (hold the applause), there are several key concepts that simply have to be taken into account to make a meaningful evaluation of his career:
1. he played in the 1970's, the first half of which, at least, was comparable to the 1910's in terms of runs scored per game. There was not one deadball era, there were two. A run produced in the 1970's is simply worth more than a run produced in the 1980's and very much more than a run produced today.
2. Tenace played in the worst hitter's park in MLB history when he was in Oakland, and then got traded to the worst hitter's park in the NL when he went to San Diego. A run produced in Oakland or San Diego when Tenace played was simply worth more than a run produced anywhere else. To say "but Tenace mostly walked, so he shouldn't get a park adjustment" is to wildly miss the point of ballpark adjustments. In Oakland, 3 runs was probably enough to win a game, and a walk from Tenace was a good start on creating one of those few elusive runs. If walking was easy - the implication of many posts I read here - then everyone would do it. But it isn't easy, and it's damn near impossible if you worry more about "making contact" then you do about helping your team win the game.
3. Tenace was a catcher. Yes, he "only" played 900 or so games at the position, but that's still 165th all time, and it was close to top 50 when he retired. And in his best years on Oakland he wasn't rested when he wasn't catching, he was generally moved to 1B. That shortened his career, as it shortened the career of everyone who ever played the position. All catchers take days off, and all catchers who hit well and have managers who aren't brain dead take those rest days against RHP. Yes, take platooning into account and adjust for it, but it is mostly a red herring in Tenace's case because it had very little effect on his career numbers. When Tenace retired the only catcher with more career batter runs was Joe Torre; he has since also been passed by Piazza and Mauer, but he will forever remain above Bench, Berra, Cochrane, and Dickey.
There is no single correct way to take these three factors into account, but not taking each of them into account is always wrong. Any conclusion reached by an analysis that doesn't take each of these factors into account will always be wrong.
Oh, and 4. His first name is Fury. How awesome is that?
In the end I don't actually have a strong opinion on Tenace vs. Garvey, this really was just masturbatory rehash. But Tenace was a whole lot better than many or all of you are giving him credit for and Garvey was a whole lot worse. I do vote for Tenace being better than Garvey, but it's close. As I'll say for the 76th and final time, it really just comes down to how important you think it is that Garvey played more games.
Dallas, I appreciate you shedding light on guys like Tenace. He was much better than what people think when they scoff at his lifetime betting average. They tend to forget that Tenace hit home runs at an elite rate as well. I don't want to leave this thread thinking I truly believe that Tenace is Ken Phelps. I just liked my story about them coming off the 'part time player' bus together. Tenace was a full blown part time player for approximately half his career.
No need to rehash all I said about Garvey. That is all true. It is a rare commodity to find a human to provide league average MLB hitting in full time at bats. Currently, sabermetrics actually penalizes that, as shown above. So, much like you are championing Tenace for being better than his lifetime batting average shows, I am shedding some light on those sabermetrics misdeeds that are unfairly penalizing guys like Garvey for being better than 99 percent of the human race at baseball when Garvey's ability allows him to continue to play at an 'average' level while others are relegated to siting on the couch because they can't hit anymore.
The difference in our thinking on this is more philosophical than statistical. Yes, Garvey was good enough to play those last four years, but his WAR for those years was 0.0; he was not an average player, he was a replacement level player. And the Padres paid him $5 million to get production they could have gotten for league minimum. Note that I normally never refer to WAR because it's crap, but here it looks about right. Garvey's OPS+ was in the high 90's over this period, less than average but better than replacement, but everything OPS+ does not capture (baserunning, GIDP, and fielding) was a negative for Garvey in his old man years. If he was better than replacement level, it was not by much.
On your broader point, though, I agree. Garvey has a career OPS+ of 117, but it was 123 after 1983 and I think 123 better describes Garvey than 117. But, if you want to think of Garvey as a 123 hitter, then you also have to think of him as a guy who left baseball at 34 with 7,500 PA, not at 38 with 9,500 PA. It's complicated trying to untangle these things, and my preference is to simply ignore hanging around years; they neither add nor subtract from a player's legacy or HOF case.
For the true greats, I like to break their careers into two pieces and consider the player as the sum of the two. Example: Carl Yastrzemski could have retired at 34 with an OPS+ of 139, instead he retired at 43 with an OPS+ of 130. Is Yaz a 139 player or a 130 player? I'd say he's a 139 player with over 9,000 PA (comparable to, say, Duke Snider) PLUS a player with an OPS+ of 113 for about 5,000 PA (comparable to, say, Grady Sizemore). {Barely relevant side note: if you add up the careers of Sandy Koufax and Dave McNally you get very close to the career of Bert Blyleven.} In other words, Yaz had a HOF career at age 34, and what happened after that shouldn't take away from that. But what did happen after 34 was that Yaz had the career of Grady Sizemore; he added to his HOF case even as his career OPS+ went down. You can repeat this exercise for Garvey, but his second player is going to be closer to Ted Sizemore than Grady Sizemore and Garvey added so little to his career value or HOF case in those hanging around years I don't think it's really worth considering. Still, I agree that those years added something rather than subtracted anything.
dallasactuary,
I agree 100% the way you break down Yaz.
To be fair, I never said Garvey should be in the HOF. He either needed to have five more productive full time years, or had to have a "Yaz" type peak by age 34. He didn't do either.
562 complete games. There's "resting" and there's playing the full game at catcher only 63% of the time.