Page 2 of 4 FirstFirst 1234 LastLast
Results 21 to 40 of 78

Thread: FG top 100

  1. #21
    It's OVER 5,000! zbhargrove's Avatar
    Join Date
    Aug 2013
    Location
    Bismarck, ND
    Posts
    5,625
    Thanks Thanks Given 
    291
    Thanks Thanks Received 
    1,264
    Thanked in
    869 Posts
    Quote Originally Posted by Metaphysicist View Post
    Incoming crotchety old man rant....

    Putting an 80 on Franco is asinine and makes me question either Longenhagen's judgment or methodology or both. These grades must be completely untethered from the traditional 20-80 OFP grading process, which makes them misleading at best. Like a perfect prospect, with 60-70/80s across the board, he could win a batting title, he might win a home run crown, he'll win gold gloves... you know, like Ken Griffey coming up, that guy would still probably get like a 70 because you know, prospects.

    An 80 is what you would give to Mike Trout the year before his first MVP, you know, when he was already a 10 WAR player. It's what FG themselves gave Aaron Judge after he put up 8 WAR as a rookie (they had him at 55 coming into the year)... because that was already his present value (link).

    Franco is a wonderful prospect, but Longenhagen's own writeup implies Franco may never develop power, doesn't have a single present tool higher than 60, only projects to have one tool over 60, doesn't project as a plus defender, has Cameron Maybin-esque groundball rates, and is still in A ball. Like, his hit tool can be so good that ultimately none of that matters, but those caveats are the exact reasons you don't give someone an 80. The idea that his median outcome is perennial MVP candidate, as his completely made-up "variance" chart shows, is laughable on its face. It reminds me of when BP projected (link) Matt Wieters to hit .311/.395/.544 as a rookie. Like, stop for a second and think about what your model is telling you.

    From FG's own site:



    So Franco, in A ball, is better than Mike Trout was when he debuted. He is equal or better than Mike Trout putting up 10 WAR as a rookie. Laughable. Just makes this entire project look unserious.
    It is interesting... in one of Kiley's last chats before he left for ESPN, he got a question if Wander could be an 80 FV prospect. He responded that he didn't think anyone could be considered an 80 FV prospect which seems to gel what Kiley said in your post. So then Kiley is out and Wander becomes an 80 FV... lol... I don't know if I can take Longenhangen seriously anymore.

  2. The Following User Says Thank You to zbhargrove For This Useful Post:

    jpx7 (02-12-2020)

  3. #22
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    1,184
    Thanks Thanks Given 
    869
    Thanks Thanks Received 
    964
    Thanked in
    452 Posts
    Quote Originally Posted by nsacpi View Post
    They try to make their distribution of grades consistent with historical data on the distribution of production of prospects.
    Which is noble (I guess; it is not obvious to me all that on the high end the distribution of talent should be the same from year to year), but just makes an 80 grade even dumber, since it is completely ahistorical.

  4. #23
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    18,648
    Thanks Thanks Given 
    32
    Thanks Thanks Received 
    6,209
    Thanked in
    3,676 Posts
    Since FG normalized their prospect FVs, they've had somewhere around 45 guys rated as 55+.

  5. #24
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    28,150
    Thanks Thanks Given 
    2,084
    Thanks Thanks Received 
    9,591
    Thanked in
    6,153 Posts
    Quote Originally Posted by Metaphysicist View Post
    Which is noble (I guess; it is not obvious to me all that on the high end the distribution of talent should be the same from year to year), but just makes an 80 grade even dumber, since it is completely ahistorical.
    What is an 80 grade to you.

    Ex post, would an 80 grade be correct for Trout, Harper, Betts, Acuna, deGrom, Yelich, Bellinger?
    I WANT NOTHiNG. I WANT NOTHiNG.

  6. #25
    It's OVER 5,000! msstate7's Avatar
    Join Date
    Jul 2015
    Posts
    16,427
    Thanks Thanks Given 
    180
    Thanks Thanks Received 
    3,291
    Thanked in
    2,123 Posts
    Quote Originally Posted by nsacpi View Post
    What is an 80 grade to you.

    Ex post, would an 80 grade be correct for Trout, Harper, Betts, Acuna, deGrom, Yelich, Bellinger?
    Only for Acuna... rest like 72.5

  7. #26
    Arbitration Eligible
    Join Date
    Aug 2017
    Posts
    3,354
    Thanks Thanks Given 
    68
    Thanks Thanks Received 
    1,145
    Thanked in
    758 Posts
    Quote Originally Posted by nsacpi View Post
    What is an 80 grade to you.

    Ex post, would an 80 grade be correct for Trout, Harper, Betts, Acuna, deGrom, Yelich, Bellinger?
    If you're talking about their median expected outcome at the time they were prospects, I don't think it'd be fair to grade any of those guys as 80 grade. Most of those guys hit the extreme upper end of their possible outcomes, which is what makes them superstars.

  8. #27
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    28,150
    Thanks Thanks Given 
    2,084
    Thanks Thanks Received 
    9,591
    Thanked in
    6,153 Posts
    How many such players are there in the minors per year.

  9. #28
    Where's My Cup of Coffee?
    Join Date
    May 2019
    Posts
    1,238
    Thanks Thanks Given 
    131
    Thanks Thanks Received 
    331
    Thanked in
    227 Posts
    I would somewhat agree this seems overkill, but I absolutely cannot agree with no prospect being worthy of an 80 grade. Simply put, if no one ever, not even a Griffey or an Andruw, is an 80 grade, then you're not grading on a 20-80 scale. If 80 is verboten, it's really only a 20-75,

  10. #29
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    28,150
    Thanks Thanks Given 
    2,084
    Thanks Thanks Received 
    9,591
    Thanked in
    6,153 Posts
    The grade values have to correspond to the ex post production. Including the tails.

  11. #30
    Arbitration Eligible
    Join Date
    Aug 2017
    Posts
    3,354
    Thanks Thanks Given 
    68
    Thanks Thanks Received 
    1,145
    Thanked in
    758 Posts
    Quote Originally Posted by Snowman View Post
    I would somewhat agree this seems overkill, but I absolutely cannot agree with no prospect being worthy of an 80 grade. Simply put, if no one ever, not even a Griffey or an Andruw, is an 80 grade, then you're not grading on a 20-80 scale. If 80 is verboten, it's really only a 20-75,
    I think the 20-80 grade scale is probably more useful for individual tools rather than overall grade, but I don't think its impossible to be an 80 grade prospect, just insanely difficult and probably not realistic. As good of a prospect as Mike Trout was, I don't think its reasonable to presume that his current performance level was his median expected outcome. If it were, could you imagine what his 90th percentile outcome would look like? There would be some otherworldly dimension of baseball performance that not even prime Bonds approached.

    And I think the same can be said for Wander Franco. Him being an 80 grade prospect would mean that, by Fangraph's own standard (https://blogs.fangraphs.com/the-new-...outing-primer/), his median expected outcome is that of a consistent 7+ win player. To me, that is absurd.

  12. #31
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    18,648
    Thanks Thanks Given 
    32
    Thanks Thanks Received 
    6,209
    Thanked in
    3,676 Posts
    A prospect doesn't have to project to being Mike Trout to be an 80 grade prospect. All they have to do is project to be worth something like $100M-$120M of surplus value over their 6-7 years of team control. For a prospect to be worth that much, they probably have to be projected to produce something like ~20 WAR over those 6-7 years (assuming $50M-$60M in earnings over those seasons).

    Guys like Griffey, ARod, and Andrew were probably 80 grade prospects because I'm confident they would have been projected to produce ~20 WAR over their 6-7 years of control. That's what FG is saying about Wander. They are NOT saying he is the next Mike Trout.

    If anything, I would argue there should be 1-2 80 grade prospects almost every year...just like there are 1-2 80 grade everythings every year.

  13. The Following User Says Thank You to Enscheff For This Useful Post:

    Snowman (02-12-2020)

  14. #32
    Where's My Cup of Coffee?
    Join Date
    May 2019
    Posts
    1,238
    Thanks Thanks Given 
    131
    Thanks Thanks Received 
    331
    Thanked in
    227 Posts
    Quote Originally Posted by Enscheff View Post
    A prospect doesn't have to project to being Mike Trout to be an 80 grade prospect. All they have to do is project to be worth something like $100M-$120M of surplus value over their 6-7 years of team control. For a prospect to be worth that much, they probably have to be projected to produce something like ~20 WAR over those 6-7 years (assuming $50M-$60M in earnings over those seasons).

    Guys like Griffey, ARod, and Andrew were probably 80 grade prospects because I'm confident they would have been projected to produce ~20 WAR over their 6-7 years of control. That's what FG is saying about Wander. They are NOT saying he is the next Mike Trout.

    If anything, I would argue there should be 1-2 80 grade prospects almost every year...just like there are 1-2 80 grade everythings every year.
    This sounds far, far more reasonable to me.

  15. #33
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    1,184
    Thanks Thanks Given 
    869
    Thanks Thanks Received 
    964
    Thanked in
    452 Posts
    Quote Originally Posted by nsacpi View Post
    What is an 80 grade to you.

    Ex post, would an 80 grade be correct for Trout, Harper, Betts, Acuna, deGrom, Yelich, Bellinger?
    Their own methodology says, average over their first six years (not one peak season), 7+ WAR, top 5 hitter/top 3 pitcher in baseball. So from your list thats:

    Trout (easily)
    Betts (maybe, depends on how you are counting "six seasons" probably)

    Nobody else is even close. To demonstrate, Bellinger broke out this year, in his third season. 7.8 fWAR is his first year over 7. If (and this is a huge if) he repeats this exact same MVP-caliber season 3 more times he will still not make this threshold (6.46 WAR average).

    So there you go. They are projecting that Wander Franco will probably be better than Mookie Betts. That seems like a thing a smart person would say.

    It's ahistorical because if they are making the "distribution of grades consistent with historical data," as you say, then it makes even less sense to give out a grade that's never been given to any other prospect. The $/WAR/etc. values for the literal best prospects ever (AROD/Andruw) AND all the guys who were rated lower but ended up as PV 80s, are included in the 75 FV calculations and lower. So the 80 FV category is literally based on nothing, but it is being calculated as if it better than all those others. It's gobbledygook nonsense.

  16. #34
    Arbitration Eligible
    Join Date
    Aug 2017
    Posts
    3,354
    Thanks Thanks Given 
    68
    Thanks Thanks Received 
    1,145
    Thanked in
    758 Posts
    Quote Originally Posted by Enscheff View Post
    A prospect doesn't have to project to being Mike Trout to be an 80 grade prospect. All they have to do is project to be worth something like $100M-$120M of surplus value over their 6-7 years of team control. For a prospect to be worth that much, they probably have to be projected to produce something like ~20 WAR over those 6-7 years (assuming $50M-$60M in earnings over those seasons).

    Guys like Griffey, ARod, and Andrew were probably 80 grade prospects because I'm confident they would have been projected to produce ~20 WAR over their 6-7 years of control. That's what FG is saying about Wander. They are NOT saying he is the next Mike Trout.

    If anything, I would argue there should be 1-2 80 grade prospects almost every year...just like there are 1-2 80 grade everythings every year.
    How many players produce ~20 WAR over their first 6-7 years? I know its probably not a ton, but I also think its probably more than >1% of the league, which is what it would have to be in order to be three standard deviations above average in a normal distribution. But I don't know, I could be wrong on that account. Perhaps its even less common than I think it is.

  17. #35
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    28,150
    Thanks Thanks Given 
    2,084
    Thanks Thanks Received 
    9,591
    Thanked in
    6,153 Posts
    Let's say Trout and Betts are the only ones who meet the standard. Trout was in the minors to be eligible for these prospect list for 2 years. Betts for 4. So over a ten year period there was a player that retrospectively have justified that ranking in six of those years.

  18. #36
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    1,184
    Thanks Thanks Given 
    869
    Thanks Thanks Received 
    964
    Thanked in
    452 Posts
    Quote Originally Posted by Snowman View Post
    I would somewhat agree this seems overkill, but I absolutely cannot agree with no prospect being worthy of an 80 grade. Simply put, if no one ever, not even a Griffey or an Andruw, is an 80 grade, then you're not grading on a 20-80 scale. If 80 is verboten, it's really only a 20-75,
    80 FV isn't verboten; you just have to remember that it isn't only a minors only scale. It's an MLB scale. Mike Trout was at his FV was probably already there by his first season. Griffey was an 80 FV once he hit the majors. Same maybe for Andruw. Definitely for AROD.

    But prospects are prospects. Not every guy who look like 5-tool, 7-skill, can't miss, consensus stud becomes an MVP. Delmon Young is an extreme case, but even guys like Cliff Floyd, JD Drew, Justin Upton seemed like the end all be all as prospects, but just ended up as "very good." That risk is baked into the grade.

    But the long and the short of it, if you are seriously projecting "Future Value," and your method gives a wunderkind in A ball the value projection grade as Mike Trout had when he put up 10 WAR as a rookie, then your method is obviously dumb.

  19. #37
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    1,184
    Thanks Thanks Given 
    869
    Thanks Thanks Received 
    964
    Thanked in
    452 Posts
    Quote Originally Posted by Snowman View Post
    This sounds far, far more reasonable to me.
    Well, I don't know what he's talking about. Their own methodology says 80 is a guy who averages 7+ over his first six seasons (42+ WAR - 5 for "earnings"). Not a "guy who produces 20 WAR."

  20. #38
    Arbitration Eligible
    Join Date
    Aug 2017
    Posts
    3,354
    Thanks Thanks Given 
    68
    Thanks Thanks Received 
    1,145
    Thanked in
    758 Posts
    Quote Originally Posted by Metaphysicist View Post
    Their own methodology says, average over their first six years (not one peak season), 7+ WAR, top 5 hitter/top 3 pitcher in baseball. So from your list thats:

    Trout (easily)
    Betts (maybe, depends on how you are counting "six seasons" probably)

    Nobody else is even close. To demonstrate, Bellinger broke out this year, in his third season. 7.8 fWAR is his first year over 7. If (and this is a huge if) he repeats this exact same MVP-caliber season 3 more times he will still not make this threshold (6.46 WAR average).

    So there you go. They are projecting that Wander Franco will probably be better than Mookie Betts. That seems like a thing a smart person would say.

    It's ahistorical because if they are making the "distribution of grades consistent with historical data," as you say, then it makes even less sense to give out a grade that's never been given to any other prospect. The $/WAR/etc. values for the literal best prospects ever (AROD/Andruw) AND all the guys who were rated lower but ended up as PV 80s, are included in the 75 FV calculations and lower. So the 80 FV category is literally based on nothing, but it is being calculated as if it better than all those others. It's gobbledygook nonsense.
    Honestly, if Fangraph's own standard were the ~20 WAR over their first 6-7 seasons, I'd probably have less of a problem with them giving Wander an 80 grade.

    I don't really think that 7 WAR per season or 20 over 6-7 is entirely accurate. I'd like to see someone analyze a sample of highly rated prospects and compare their results to the general understanding of the 20-80 scouting scale and standard deviations. How many 60 grade prospects were around the 68th percentile, how many 70s were around the 95th, how many 80s were around the 99th, etc.

    I'd just be interested to see the results just so we'd have a better idea of what exactly we should expect when we say someone is a 70 grade prospect, or whatever. Its also pretty confusing to know when people are referring to a player's MEO or their ceiling, which can make a massive difference.

  21. #39
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    28,150
    Thanks Thanks Given 
    2,084
    Thanks Thanks Received 
    9,591
    Thanked in
    6,153 Posts
    Mike Trout is an extreme outlier. But I think it is reasonable to work on the basis of a distribution where the best prospect puts up Betts-like production before he reaches free agency.

  22. #40
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    1,184
    Thanks Thanks Given 
    869
    Thanks Thanks Received 
    964
    Thanked in
    452 Posts
    Quote Originally Posted by nsacpi View Post
    Let's say Trout and Betts are the only ones who meet the standard. Trout was in the minors to be eligible for these prospect list for 2 years. Betts for 4. So over a ten year period there was a player that retrospectively have justified that ranking in six of those years.
    I mean, that’s just literally not how they calculated it. Their estimations are based on a database with no 80s.

    You’ll also notice that there is a FV outcome distribution graph for each prospect on the list. This is our attempt to graphically represent how likely each FV outcome is for each prospect. Using the work of Craig Edwards, I found the base rates for each FV tier of prospect (separately for hitters and pitchers), and the likelihood of each FV of outcome.
    The article he links to has 70 as the max grade. There are no 80s.
    Last edited by Metaphysicist; 02-12-2020 at 08:19 PM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •