SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
7/13 ML Gameday Thread: Happy Anderson Espinoza Day
|
Post by stevedillard on Jul 14, 2015 10:56:02 GMT -5
Does anyone know why Andrew Benintendi is not playing? Really have missed seeing his name in the box score. Promoted to AL All-Star team in Cincinnati.
|
|
|
Post by Chris Hatfield on Jul 14, 2015 10:57:30 GMT -5
Does anyone know why Andrew Benintendi is not playing? Really have missed seeing his name in the box score. Promoted to AL All-Star team in Cincinnati. Actually, close - he's there to receive the Golden Spikes Award.
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Jul 14, 2015 11:44:59 GMT -5
Unless you believe that players do not differ from one another defensively, there's nothing wrong here at all. [...] So what was the difference between them? Valentin was 20/25 stealing, and Nomar was 5/7. People who don't believe in WAR probably do believe that everyone is equally good at baserunning. Or that it doesn't matter as much as anyone who has actually looked into it thinks, because how could you possibly quantify that? You don't know whether you have Big Papi, who is guaranteed to hit a home run every time, behind them in the order, or Mike Napoli, who is going to strike out anyway, blah blah etc. I mean there are a lot of ways to be ignorant. Just recently someone on this forum postulated that only Clay Buchholz could singlehandedly lead this team to victory, including the days he doesn't start. People don't want to believe that baseball is a game of averages. It is not a matter of ignorance, it is just the opposite. I like facts, hard data, and measurements. I'm very scientific in my thinking and approach. The problem with WAR is that it pretends to be scientific when it fact it is based on a considerable amount of assumption, not facts, particularly when it comes to fielding. Even now there are questions about fielding metrics even though they have improved dramatically. But there were no reliable fielding metrics in 1995 based on actual measurements. Those ratings are very subjective. I don't doubt that Valentin was a better fielder than Garciaparra. I saw both of them play. But Garciaparra in his prime was quite good. The difference was not significant enough to begin to close the gap in offense. Garciaparra was an impact hitter, an intimidating hitter in his prime, one of the best in Sox history. Valentin had a good couple of seasons but never had the offensive impact of Garciaparra. And I don't think that difference in baserunning really has much significance. See, that is where WAR has a problem. Somebody gives it a greater amount of significance but not on any objective basis that I can see. Show me the hard evidence of its value and I'll change my mind. I know WAR is supposed to even out the differences among players and rate their overall value, but I question how it is done and the values assigned to each aspect of performance and how the data is assembled. Winning in baseball occurs by scoring more runs than the number of runs allowed. Good defense and good pitching can reduce the number of runs allowed. Good hitting and baserunning can increase the number of runs scored. A player's value certainly is a combination of his performance at bat, on the field, and on the bases, but how that value is apportioned among those, so far as I can tell, is subjective. How much are 35 HRs worth if the player is not good defensively? A lot may depend on the position he plays but also how are those differences calculated? What factors and assumptions go into the calculation. If you read fangraphs long piece on WAR you will see that there still is debate over this. I am not saying that WAR has no value, just that it should be seen as just one factor in valuing and rating a player, and not the most important one. I think Sandoval is significantly undervalued in WAR this year. His occasional defensive lapses have been given far more weight than they deserve. He isn't routinely doing a poor job defensively. In fact, generally he is pretty good, and sometimes makes some great plays. This is but one example of why I don't trust WAR.
|
|
|
Post by jimed14 on Jul 14, 2015 11:55:28 GMT -5
But you can't really measure facts when a player makes a bad read and doesn't catch the ball. You cannot measure facts when a player should have tagged up and advanced on a fly ball but didn't.
Just because they are estimated instead of counted doesn't mean they don't significantly affect games.
|
|
|
Post by jmei on Jul 14, 2015 12:08:06 GMT -5
A player's value certainly is a combination of his performance at bat, on the field, and on the bases, but how that value is apportioned among those, so far as I can tell, is subjective. Nope, it's all very objective, and based on linear weights. It's really easy-- if you misplay a fly ball, how many runs, on average, would it cost your team? Same goes if you fail to go first-to-third or get caught on the basepaths. Add that all up, and you have WAR. Brief discussion of linear weights: www.fangraphs.com/library/principles/linear-weights/www.bleedcubbieblue.com/2010/1/17/1255925/uzr-error-fail-or-win-a-lesson-onwww.fangraphs.com/blogs/ultimate-base-running-primer/
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Jul 14, 2015 12:38:03 GMT -5
And a thing about baserunning and fielding. Look at fangraphs ratings of Mike Napoli. He has negatives for both baserunning and fielding, consistently.
Almost everyone on the Red Sox will say that Mike Napoli is the best baserunner on the team, not the fastest, but the smartest. In spring training he is used to teach baserunning to young players.
From every observation of mine, and from all the reports of his performance in the field that I ever have read, Napoli is a superior-fielding firstbaseman. He is so good that his performance has generated much comment over time. Also, this year when it has been down, it really has been noticed.
Now about Garciaparra's baserunning. In 2000 he mostly was hitting cleanup from what I can tell from random boxscores. He had some sluggers hitting behind him. It is not the usual practice in baseball to have runners trying to steal bases in front of the sluggers, the best hitters. Usually the most bases being stolen are done by the leadoff hitter, or sometimes, by hitters way down in the order. But middle of the lineup guys usually don't steal many bases. There are exceptions.
So the fact that he stole so many fewer bases than Valentin basically is irrelevant and should not be a factor in his WAR. What might be a factor is what kind of a baserunner was he? I don't remember, but I don't think he was slow, or dumb. In the stats on Baseball-Reference.com there are stats for taking extra bases, etc., but I don't see how they are useful without more data about opportunities and situations, which, of course, are very hard to quantify. Probably if one went through every play by play of every game, it might be possible to quantify enough to come up with a reasonable metric. Short of that, about the only things that have any relevance are basestealing percentages and maybe being caught off base.
|
|
|
Post by jmei on Jul 14, 2015 12:48:20 GMT -5
And a thing about baserunning and fielding. Look at fangraphs ratings of Mike Napoli. He has negatives for both baserunning and fielding, consistently. Almost everyone on the Red Sox will say that Mike Napoli is the best baserunner on the team, not the fastest, but the smartest. In spring training he is used to teach baserunning to young players. From every observation of mine, and from all the reports of his performance in the field that I ever have read, Napoli is a superior-fielding firstbaseman. He is so good that his performance has generated much comment over time. Also, this year when it has been down, it really has been noticed. Napoli's fielding at 1B has been above-average for all three of his years in a Red Sox uniform. He had UZR/150s of 13.3 in 2013 (tops in the league), 7.3 last year (well above-average), and 2.9 this year (when, as mentioned, his defense has been noticeably worse). The Fangraphs "Def" column, which is what I think you're looking at, includes the positional adjustment, which is a little confusing, but still makes sense insofar as even an elite-fielding 1B is a worse overall defensive player than an average-fielding SS. The unspoken thing about Napoli's baserunning instincts is that it allows him to somewhat overcome well below-average foot speed. He makes the most of his physical tools, but that doesn't mean he's a good baserunner compared to the rest of the league. The fact that Fangraphs' baserunning stat rates him as slightly below-average as a baserunner despite his well below-average speed reflects that reality.
|
|
|
Post by redsox04071318champs on Jul 14, 2015 13:03:28 GMT -5
Unless you believe that players do not differ from one another defensively, there's nothing wrong here at all. [...] So what was the difference between them? Valentin was 20/25 stealing, and Nomar was 5/7. People who don't believe in WAR probably do believe that everyone is equally good at baserunning. Or that it doesn't matter as much as anyone who has actually looked into it thinks, because how could you possibly quantify that? You don't know whether you have Big Papi, who is guaranteed to hit a home run every time, behind them in the order, or Mike Napoli, who is going to strike out anyway, blah blah etc. I mean there are a lot of ways to be ignorant. Just recently someone on this forum postulated that only Clay Buchholz could singlehandedly lead this team to victory, including the days he doesn't start. People don't want to believe that baseball is a game of averages. I read that post and sincerely doubt that the poster meant that Buchholz was the sole reason why the Sox were playing better baseball. Buchholz pitching consistently well, keeping the opponents off the scoreboard on the day he pitches, obviously increases the Red Sox chance to win that day, even if he doesn't swing a bat. That's pretty obvious. As far as the day he doesn't start, a well pitched game can help rest an overworked Red Sox bullpen the day before and the day after increasing the chances to win the day before or after. When Pedro pitched, the Sox knew that they could use up the pen the day before or the day after because the odds were pretty against them using the pen heavily on the day Pedro was slated to go. That might not be measurable the way you like, but there's still an impact. Obviously Buchholz pitching well and then going down with an injury doesn't help the cause. I doubt whoever takes his spot in the rotation will pitch as well as he had been pitching. Now what this has to do with Anderson Espinoza, I have no idea. Other than having used Pedro in this post and hoping that one day Espinoza will be the best Sox pitcher they've had since Pedro...
|
|
|
Post by okin15 on Jul 14, 2015 13:14:12 GMT -5
danr, I don't think anyone is holding up WAR as the holy grail be all end all of evaluating a player, and I think you could certainly argue Garciaparra's single season superiority to Valentin's but WAR is absolutely the best method we have publicly available of gauging a player's value over a season or multiple seasons and taking into account everything that player does on the field.
Also, I would certainly take Valentin's 95 over Nomar's best season. Unfortunately, they were both aided by the same trainer, the one who helped both of them (in the words of a Globe article I've never been able to find) add strength without costing them agility, which a) only lasted a few years before father time very quickly caught up to each, and b) seemed nearly impossible at the time, and in my eyes, suspect many years later. I note how both had to move off of SS very early in their respective careers.
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Jul 14, 2015 13:16:29 GMT -5
Those are good reads and I appreciate it. I don't have much of a problem with offensive WAR. It makes sense and the data is pretty solid. I question the fielding and baserunning data because there are many variables that are not calculated - and maybe can't be.
I understand the use of averaging and linear weights. I love the chart showing the percentages of average outcomes. Some of that I knew, but I think mostly from experience and observation.
There are things in fielding and baserunning that are validly measured and certainly are factors that should be used in valuing a player. They include things like errors, caught stealing, getting picked off or thrown out. However, because most fielding and baserunning events occur when the ball is in play, there literally are infinite variables affecting the outcomes, many of them beyond the player's control. Some of them include how hard a ball has been hit, where it has been hit, conditions of the field, where the fielder was positioned (which often is not his decision), what the visibility situation is, and others.
Whether a player tries to score usually is at least partially the responsibility of the third base coach. Whether a runner can steal a base is affected by the pitcher's move and time to the plate, the catcher's throwing ability, and timing of his throw, and the ability of the fielder to properly handle the throw and the tag. Whether a runner can successfully take an extra base depends on the fielding and throwing ability of the defense as much, or more as on his baserunning ability, and those skills obviously vary infinitely. I know averages are being used but I question whether the averages really are accurate. Fangraphs thinks more data would not change them significantly. Right now I think it makes sense to lessen the weight given them. In fact, I would not combine the metrics into one WAR rating.
So, my argument is that the weight given to fielding and baserunning does not have as solid a basis in actual fact and performance as does hitting. And I think baserunning probably has considerably less than fielding.
|
|
|
Post by jmei on Jul 14, 2015 13:28:34 GMT -5
However, because most fielding and baserunning events occur when the ball is in play, there literally are infinite variables affecting the outcomes, many of them beyond the player's control. Some of them include how hard a ball has been hit, where it has been hit, conditions of the field, where the fielder was positioned (which often is not his decision), what the visibility situation is, and others. Whether a player tries to score usually is at least partially the responsibility of the third base coach. Whether a runner can steal a base is affected by the pitcher's move and time to the plate, the catcher's throwing ability, and timing of his throw, and the ability of the fielder to properly handle the throw and the tag. Whether a runner can successfully take an extra base depends on the fielding and throwing ability of the defense as much, or more as on his baserunning ability, and those skills obviously vary infinitely. I know averages are being used but I question whether the averages really are accurate. Fangraphs thinks more data would not change them significantly. Right now I think it makes sense to lessen the weight given them. In fact, I would not combine the metrics into one WAR rating. So, my argument is that the weight given to fielding and baserunning does not have as solid a basis in actual fact and performance as does hitting. And I think baserunning probably has considerably less than fielding. None of those variables are unique to fielding and baserunning. Replace the fielding and baserunning references to hitting ones and your statements remain true. For instance: - However, because most [hitting] events occur when the ball is in play, there literally are infinite variables affecting the outcomes, many of them beyond the player's control. Some of them include how hard a ball has been hit, where it has been hit, conditions of the field, where the fielder was positioned (which often is not his decision), what the visibility situation is, and others.
- Whether a [hitter] tries to [sacrifice] usually is at least partially the responsibility of the [head] coach.
- Whether a [hitter can get a hit] is affected by the pitcher's [pitching ability], the catcher's [game calling and framing] ability [...] and the ability of the fielder to properly [field his position].
- Whether a [hitter gets a hit] depends on the fielding and throwing ability of the defense as much, or more as on his [hitting] ability, and those skills obviously vary infinitely.
It's fair to say that baserunning and fielding metrics have higher error bars because of smaller sample sizes and sketchier inputs (particularly on the fielding side). But you're exaggerating the magnitude of the difference.
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Jul 14, 2015 13:31:59 GMT -5
danr, I don't think anyone is holding up WAR as the holy grail be all end all of evaluating a player, and I think you could certainly argue Garciaparra's single season superiority to Valentin's but WAR is absolutely the best method we have publicly available of gauging a player's value over a season or multiple seasons and taking into account everything that player does on the field. Also, I would certainly take Valentin's 95 over Nomar's best season. Unfortunately, they were both aided by the same trainer, the one who helped both of them (in the words of a Globe article I've never been able to find) add strength without costing them agility, which a) only lasted a few years before father time very quickly caught up to each, and b) seemed nearly impossible at the time, and in my eyes, suspect many years later. I note how both had to move off of SS very early in their respective careers. Differences of opinion are fair. Garciaparra did push Valentin to 3B and I know Valentin had some injury problems. Much of Garciaparra's later problems occurred because of a rare ailment he had that caused enlarged scar tissue. That meant that every time he suffered an injury, a greater amount of scar tissue than normal would build up. It began with that wrist injury suffered from an Oriole pitched ball. His wrists were the key to his great hitting and his wrist never was the same again, and thus neither was his swing. I don't look at all players the same and thus WAR ratings don't mean much to me. I look at players in the context of the team and the position. What I want in a shortstop is a whole lot different from what I want in a leftfielder. Or to put it another way, if I have a great hitting shortstop, I can take a chance on a weak hitting but great fielding second baseman. So I look at the individual skills separately and try to balance a team so that overall there is both good fielding and good hitting, but not necessarily both in every position. If my leftfielder hits 40 HRs, but isn't too good defensively, I'd make up for it by having a great fielding CF and shortstop. I write about Sandoval a lot because the hatred of him is irrational. He's not a great 3B. He's not a great hitter. He's better than what the Sox had but not as good as a number of other 3Bs. But in any case, he is not to blame for the Sox failures this year. He has had only one period when he slumped and it was when he was hurt. The Sox would have been worse off without him. He's probably paid more than he should be, but I really don't care about that. I think many players are greatly overpaid, but considering the revenues of baseball, not unfairly overpaid.
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Jul 14, 2015 13:35:27 GMT -5
None of those variables are unique to fielding and baserunning. Replace the fielding and baserunning references to hitting ones and your statements remain true. For instance: - However, because most [hitting] events occur when the ball is in play, there literally are infinite variables affecting the outcomes, many of them beyond the player's control. Some of them include how hard a ball has been hit, where it has been hit, conditions of the field, where the fielder was positioned (which often is not his decision), what the visibility situation is, and others.
- Whether a [hitter] tries to [sacrifice] usually is at least partially the responsibility of the [head] coach.
- Whether a [hitter can get a hit] is affected by the pitcher's [pitching ability], the catcher's [game calling and framing] ability [...] and the ability of the fielder to properly [field his position].
- Whether a [hitter gets a hit] depends on the fielding and throwing ability of the defense as much, or more as on his [hitting] ability, and those skills obviously vary infinitely.
It's fair to say that baserunning and fielding metrics have higher error bars because of smaller sample sizes and sketchier inputs (particularly on the fielding side). But you're exaggerating the magnitude of the difference. ------ That's fair comment and well-argued. Good argument usually involves some exaggeration.
|
|
|
Post by okin15 on Jul 14, 2015 13:39:43 GMT -5
I am having the longest, most serious case of Deja Vu I've ever had. Didn't danr and JMei have this EXACT same conversation (like verbatim with some of the posts copied) a few weeks ago? Bewildered.
|
|
|
Post by ethanbein on Jul 14, 2015 13:39:53 GMT -5
Relevant comment thread on Tom Tango's blog, on uncertainty in Offense vs. Defense. Basically, on any given plate appearance, there are factors external to the player (i.e. quality of opportunity), but they are small compared to what happens. If a player hits a home run, that's an excellent outcome, regardless of if it was a meatball right down the middle from Craig Breslow. For fielding on the other hand, the "what actually happened" is that a fielder made or didn't make a play. The differences in those are dwarfed by quality of opportunity. There are huge differences in difficulty among balls in play, and there just isn't an equivalent on offense of that. That's not a reason to say WAR is stupid or that defensive metrics are useless, but I think that, especially when we're talking about pre-2001 seasons where defense is not based on location data, only play-by-play, it's completely conceivable that one player who has a whole extra WAR over somebody else actually had a worse season. Now, Valentin had several good years defensively, but this was his best by the numbers. I think it's fair to regress his numbers down a little bit. I would probably still think that Valentin had a better season, by the numbers, but danr isn't making an outrageous case, even if some of his WAR criticisms are misguided.
|
|
|
Post by jchang on Jul 14, 2015 15:40:29 GMT -5
not to belabor the point, f/b WAR represents the best efforts of f and b respectively to build a composite metric of the stats to correlate with runs, and implicitly wins. The correlation is not perfect, having some degree of error as stated or simply compare team WAR with record. Now if you think you have a quantitative method for translating the myriad of stats into a better representation of wins, then present it to your favorite GM and get yourself hired, like the chubby kid in MoneyBall (movie version). (Another neat trick would be to predict a player's future stats).
|
|
|