Rant & Rave Saturday

This article has 41 Comments

1. WAR – A stat that is not precise, rather an estimation of value based on imperfect data and assumptions. lol

   A 6 WAR player is worth 5-7? That’s like a 18% rate of error. So a guy with a 3 WAR could be worth 2 WAR and a guy with 1 WAR can be worth 2 WAR. Yeah, that’s useful and it is exactly what I was saying.

   I really like the rotation with Bueller, Kershaw, Urias, May, Maeda with Gonso and Strip backing them up. But, is it enough? Hill just completed a 3 year deal in which he pitched in his age 37-39 year seasons. I don’t have a problem giving Ryu a contract to pitch in his age 33-36 seasons as he will complete his contract younger that Hill began his. On the other hand, I think Urias, May and Gonso all have very high ceilings and velocity seems to play up in the post season. I’m really on the fence here.

   I think either Lindor or Betts will be a Dodger next year and that they’re staying “in” on Donaldson as a fall back. This off-season isn’t done yet. I like Clevenger a lot, but I think our three kids all have the stuff that Clevenger has. But, will they be able to be similarly good next year?

   I agree that RVS will continue to improve our hitters. We had a great approach for most of season last year, but seemed to tail off towards the end. I also like Bard coming back as the bullpen coach. The big unknown is how Prior will do the with the pitching staff. I’m glad Honey will be around to continue to work with him.

   It’s hard to deny that AF is doing a good job. But, you can’t give him an “A” without a Championship especially when his hire keeps messing it up for him. But, there’s more than one right way to build a baseball team and to run a baseball team.

   An analogy – many cars come with chips these days that can be tuned for economy or performance. If your goal is to get good gas mileage during a cross country trip, economy would be the proper tune. But, if your goal is to get from point A to point B as fast as possible, you want to performance tune. Let’s see if AF can figure out how to keep his cross-country tune during the regular season, but get the performance tune for the post-season.

   1. I think one needs to look at WAR in a relative sense not absolute. The important thing is the trend. Clearly a 6 WAR player is better than a 2 WAR player by a fair amount.

      As you point out there is an error band associated with any annotation of WAR to a player. If you did an uncertainty analysis to the given WAR projection you could aplply an uncertainty band to the number. So if there is an 18% error a 6 WAR player could be either would fall into a 5-7 range as has been stated. Assuming the error band was linear ( and for this exercise appropriate). A 1 WAR player would also have an uncertainty band of 18% so he’d be worth 0.82 – 1.18.

      1. Making another assumption just makes the stat even weaker. The more assumption, the less precise. Using your assumption, the 10 WAR player could be a 5 WAR player? Does that make sense to you?

         1. No, using your example a 10 WAR player and an error band of 18% you would have a range from 82 – 11.8. All numerical analyses need to have uncertainty applied to them. I refer you to the work of Doctor Hugh Coleman of the University of Alabama, Huntsville. Given 2 sets of data and associated curve fits one would think they could determine the better of the two correlations. However, once uncertainty is applied, the results are less clear. Using uncertainty analysis does not make things less clear it does the opposite.

            I have no idea of the level of detail that Dodgers analytic team is using but given the level of expertise I’d be surprised if uncertainty was not part of the analysis.

      2. Terrific explanation. Thank you

   2. But the fact remains that there is NO other single stat that pretty much embodies what a player is other than WAR and to ignore it is to deny reality. Look at the TOP WAR Players:

      http://www.espn.com/mlb/war/leaders

      We may disagree about the “order” they are ranked, but do you see any players in the TOP 20 who obviously should not be there?

      Nick Castellanos had a 3.6 offensive WAR, but when Defensive WAR was factored in, his WAR was 1.5. I don’t care which version of WAR you use, the Rankings are very close.

      Tell me what other SINGLE STAT you can use to evaluate a player that is even close to WAR?

      1. Thanks for providing that list because it helps prove my point. I guess Mike Minor would have gotten 40M per year since he is clearly better than Garrett Cole. Almost 1 WAR better. And yes, Eduardo Rodriquez is NOT a top 20 player, Fangraphs has him at 85.

         There is no single stat you can use to accurately evaluate a player, including WAR. If Fangraphs, Baseball America, and ESPN can’t agree on which player is better, how can you possibly justify using that stat to make an argument that one player is better than another? In my reality, I can recognize this. Your reality is clearly different.

         1. The question raised from Marks list is why is Minor listed ahead of Cole? Looking at basic stats (whip,batting ave, era) Cole was better. So why the lower WAR? Without knowing all the details I’d even wonder if it’s an error (does he even belong on this list).

         2. Are you making a serious argument, or just arguing for the hell of it 59inarow?

            1. There is no argument… except in his mind.

            2. Hey Bluto, I was making serious arguments, but was using sarcasm for effect.

               I’m glad we didn’t sign Garrett Cole because Mike Minor’s in his walk year. We should just wait until the trade deadline and trade for him. He’s way better anyways, because his WAR was higher last year.

               WAR was mentioned 50 times so far today so at least there was something to talk about.

            3. What is the serious argument?

               That WAR isn’t perfect?

      2. Harper isn’t even in the top 50! Why is he considered to be sooo good! Max Muncy however, IS in the top 50. Just saying!

2. So, I’m not burdened with the knowledge many of you have. I do love the game and the Dodgers (I do have an opinion, however). What this lay person sees when I read MLB about trades ….Lux, May and Ruiz for Lindor?? Cra cra! Anyone who would give up that haul for 2 years shouldn’t be a Dodger executive! Let’s hope not anyway. Let the kids play, release our hounds!!

3. AF is not trading Lux for anybody for 1-2 years of control nor should he. Boston is gonna have a tough time dumping Price and if they determine they can’t win next year and they can’t then they have to deal Betts to reduce payroll. We have the depth of minor league talent to get a deal done without giving up our top prospects and the financial room to take Betts on. I can’t believe there are too many teams that can match that..Same for Lindor. I’d do a depth offer outside our top 4 guys and if it doesn’t fly move on.

4. Okay, I’m not a fan of patience.

   Hit it, let it fly and don’t look back. But I do understand that sometimes you need to pace the market. The trade market has been slow to wake up, maybe that’s why we need a deadline at the winter meetings. Hey, trades are fun and it gives the MLB writers and analysts more to talk about than some team just gave some player a $200 or $300 million contract.

   While everybody can sit on the porch and chew, spit and argue until the cows come home, now the discussion comes down to the validity of WAR. Of course this is just a comfort zone to place value on a baseball player. Did we really need defensive analytics to tell us that Cody Bellinger was an incredibly good defensive right fielder? Or Mike Trout really hits a baseball hard? Or Gerrit Cole had a great 2019 season?

   Did Bill James actually make Baseball a better game to watch with all the data? Did the Boston Red Sox actually win the World Series because of the analytics and data that James generated? Did they use that same data to sign Carl Crawford? Did the Cubs use data to sign Jason Heyward? Actually they did. How did that work out? Does WAR explain Howie Kendrick?

   Did the Dodgers sign AJ Pollock because of WAR or other data? No, they didn’t. They had watched him come up with big hits and good defensive plays for years. The same reason I wanted them to sign him. He terrorized the Dodgers. As for the injuries, AF said they were freakish in nature, not a reoccurring type of injury, like a back or balky shoulder.

   I like the signing, some don’t. I consider the second half, but also remember how good AJ played against the Dodgers. Then I look at the big hits, homeruns, key hits with runners on base in the second half.

   I’m not sure what to make of the Donaldson rumors. Are the Dodgers still in, were they ever really in on the third baseman. Makes no sense. A 34 year old on a three or four year deal to replace a 35 year old, who has clutch hitting skills.

   Then there is this: David Vassegh, the Dodger Talk host and a regular on Sportsnet, the Dodgers TV show, said that Mookie Betts isn’t going to be traded, according to his sources in Boston. Lindor? He said Ryu really likes LA and wants to return to the Dodgers, but the potential length of the deal may prevent that from happening. There is concern about Ryu’s injury history and the team would prefer a higher annual dollar amount and less years. He could end up in Anaheim, he suggested.

   He also mentioned Nick Castellanos as a potential target, which should make Mark happy, but said the Dodgers would need to clean up an overcrowded outfield to make that feasible. He suggested the way the Dodgers position outfielders would help with defensive deficiencies.

   Now Vassegh has some insight into the Dodgers so take it for what it’s worth.

   1. Good post sbuffalo.

5. I would like to see in the next MLBPA players agreement use of a version of WAR as the basis of players salaries including rookies.

   Maybe a team could offer a free agent $20M if his WAR falls between 4 and 5 while another team would only offer $18M. Non free agents would earn a salary based on their end of year WAR.

   That pay plan shares the risk of performance and nonperformance equally between players and owners.

   1. I’d like to see no regulation on salary.

      1. Me too

6. Bums, teams pay for future projected performance not past performance. If that was the case then the Albert Pujos’ contract would have been justified.

   1. I was saying to pay for WAR they earned during the year and not past years.

7. https://www.google.co.uk/amp/s/www.12up.com/amp/posts/irving-piken-oldest-living-man-united-states-brooklyn-dodgers-fan-01dwj3f6zyef

   1. Terrific post Watford. Thank you and Merry Christmas.

      1. Thanks DBM – back at you.

   2. Wow, Watford! Great find. Thanks for posting.

8. Perhaps the hitting coaches can work with Seager , Bellinger, and Pollock to train them to hit the pitching that they may face in the playoffs. In other words guys that can throw secondary pitches for strikes early in the count. And to somehow figure out that once you are behind in the count you will be getting the slider in the dirt, not the fastball.

9. ”Many forms of government have been tried and will be tried in this world of sin and woe.
   No one pretends that democracy is perfect or all wise. Indeed, it has been said that democracy is
   the worst form of government, except for all the others that have been tried“ – Winston Churchill

   I’m perpetually amazed at fans opinions regarding WAR. That if it does not perfectly quantify everything it is therefore useless. As if MLB executives blindly make decisions on a single stat spit out by some magic black box with an algorithm designed entirely to ruin baseball.

   Statistics are information. Nothing more. WAR is designed to aggregate a large number of statistics into a uniform standard number. Kind of like when you got a report card as a kid. It did not contain 15 pages of every quiz, test, homework assignment and project completed in every single class. Those grades were weighted, averaged and compiled into a single letter grade. And all those letter grades were used to calculate GPA. Is GPA a perfect reflection of a students abilities? It’s not. Nor was is ever meant to be. But the college admission department at a university isn’t going to read your crappy book report from your freshman year about The Grapes of Wrath. They’re going to start with your GPA, SAT scores and work from there.

   A statistic like WAR, BABIP and the like provide additional information. Nothing more. They are a means to BEGIN the process of digging deeper, not the end.

   AD (or any other executive) would be an absolute fool to not embrace any and all statistical data that can provide additional insight into any aspect of the game. Not to do so would be insane.

   But no, I hear fans complain about WAR and advanced stats as though AD just let’s some algorithm make decisions for him. “Uh oh, computer says Puig sucks now. Better make a trade”

   Because that’s how it works. Just like when your check engine light goes on. You must head directly to the junk yard and dispose of that car. Immediately buy a new one. But wait, you say check engine lights don’t mean your car is in danger of an imminent, massive explosion? Well then, they must be completely ignored because they are useless.

   That’s kinda the same logic.

   More information is never a bad thing.

10. When my check engine light went on I always thought it was a bad bulb. I guess that’s why they had to tow my car away when I donated it last month!

    1. If it was a bad bulb, wouldn’t the light be off? Maybe my check engine light is on. But the bulb is bad!

       I’m buying a horse.

       1. That reminded me of a friend who had trouble during her driving test because she could not demonstrate hand signals. The dmv inspector told her “you must know these signals if your car stalls and you lose power”. My friend said,” well if the power goes out and the car quits running, the electric windows will not go down so I can use my hand signals”. He flunked her anyways.

       2. I have found that a tap on the check engine light with a sharp object will turn it off.

          1. Works with left fielders too, sometimes. Often they are simply overweight and it just pisses them off.

11. Dallas Keuchel signs with White Sox for 3years, $55mil. I’d guess that’s a starting point for Ryu’s next deal

12. re: “All truth passes through three stages: First, it is ridiculed. Second, it is violently opposed. Third, it is accepted as self-evident.”

    Hopefully the same isn’t true with lies although it is trending.

13. There are probably only a few posters here (I’m not one of them) who understand how WAR is computed, the differences between WAR computation, and exactly how valid the calculation really is.

    I am troubled by certain aspects of WAR – for example, pitchers strikeouts are a big deal, so a low K out getter (like Tommy John, Claude Osteen,and just about every soft-tossing left-handed sinker pitcher you’ve ever watched) gets lower WAR than a K machine who is less successful at actually preventing runs. There is a certain modern bias that is reflected in WAR that troubles me.

    WAR is just OK as a starting point as far as i a concerned. I agree with JC – it’s data, but it’s not an end all or be all. Part of the modern preference for computer analysis over the eyeball test also troubles me. There are lots of ways to analyse players – I wouldn’t rely too heavily on WAR.

    1. Again, all I have said is that WAR is the single best stat… if you only have one.

       WAR evolves as the algorithm changes and improvements are made.

       It’s not perfect, but it is the same for everyone.

       WAR is just one of DOZENS of ways to analyze players.

    2. The “eye test” is limited. Vastly more so than statistics. The reason is actually simple. The limitations of our brains. We evolved abilities that computers today can’t remotely emulate. While our brains are massively inefficient at processing complex mathematical equations, which computers have done with ease since the 1950s, we can can make spatial estimates (grabbing a ball thrown to us, for instance) which even modern computers/robotics struggle at immensely. Our brains are incredibly efficient at making estimates for things that happen in a relatively short period of time. We can identify short term patterns and match them with past experience with incredible speed. Our ancestors could identify the sound of rustling in the bushes and the distinctive growl of a predator and take off running with astonishing speed. Individuals who were less adept at the skill didn’t reproduce. They were eaten.

       But this skill was only honed within the limits of a relevant timeframes. We are not very adept at identifying subtle patterns over longer periods of time. Simply because it didn’t offer a survival advantage.

       I remember distinctly hearing criticism of Ken Griffey Jr. He was “a bit lazy”. He “didn’t really go all out”. That was commentators eyes. They were bias towards seeing a player running all out and diving for a ball. That was “full effort”. Never mind that Jr didn’t dive for as many balls or look like he was going “all out” as often because he had a natural ability to correctly calculate a rout. He got to balls easily which a lesser CF would have to noticeably struggle. He caught balls on the run which most fielders would have to dive for.

       Over the course of 162 game season this is something difficult to detect with the eye. Only when one maps a stadium and compares exit velocity, launch angle and hang time can you really calculate how efficiently an outfielder gets to balls. It’s something that human perception isn’t very effective at. But it’s something computers excel at.

       You can’t replace scouts or the human eye. Computers can’t tell you a players character. They can’t tell you the look on their face under pressure. They can’t understand all the intangibles we all know exist in a very complex game. But they are very good at quantifying and comparing data sets over time and comparing them to other data sets (like outfield range).

       You can’t replace the human eye for everything. But the human eye has vast limitations. Which analytics has been designed to compliment. What you can never replace is a person skilled at understanding the strength and limitations of both. Which is where the magic happens.

       1. … and the attorneys on the blog can tell you that eyewitness testimony is some of the most unreliable.

       2. Good points, JC. Sounds like we need both the human eye and the War stats.

14. Is it possible that we need both the human eye and the stat WAR to truly evaluate a player?
    Both seem benificial.

Comments are closed.