Page 3 of 4 FirstFirst 1234 LastLast
Results 41 to 60 of 78

Thread: FG top 100

  1. #41
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    Quote Originally Posted by Metaphysicist View Post
    I mean, that’s just literally not how they calculated it. Their estimations are based on a database with no 80s.
    I'm a bit less interested in the historical data base of 80s is than what the right assumption should be about the value of the #1 prospect and actually the whole distribution of prospect values. And that should be based on the actual production data of players in the pre free agency years.

  2. #42
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    2,641
    Thanks Thanks Given 
    1,547
    Thanks Thanks Received 
    1,645
    Thanked in
    878 Posts
    Quote Originally Posted by nsacpi View Post
    The grade values have to correspond to the ex post production. Including the tails.
    This is not at all true. The grade includes uncertainty.

  3. The Following User Says Thank You to Metaphysicist For This Useful Post:

    jpx7 (02-12-2020)

  4. #43
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    Quote Originally Posted by Metaphysicist View Post
    This is not at all true. The grade includes uncertainty.
    I think that depends on the purposes of the grades. Lets say you are trading a basket of minor leaguers for a basket of major leaguers. You need expected values of both baskets.. Estimates of uncertainty only come into play to the extent they are significantly different for the two baskets.

    Also related to your point I believe that prospects very far away from the majors get dinged in the current prospect ratings relative to the ones that are older and closer to the majors. In an imperfect way the ratings do seem to take into account uncertainty.

  5. #44
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    I might add that some baskets of major leaguers carry particular risk due to the contracts attached to them. See Arenado for example.

  6. #45
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    2,641
    Thanks Thanks Given 
    1,547
    Thanks Thanks Received 
    1,645
    Thanked in
    878 Posts
    Quote Originally Posted by nsacpi View Post
    I think that depends on the purposes of the grades. Lets say you are trading a basket of minor leaguers for a basket of major leaguers. You need expected values of both baskets.. Estimates of uncertainty only come into play to the extent they are significantly different for the two baskets.
    I don't really know what this means. Uncertainty is baked into the "expected value" of a prospect. That's how FG does it at least, which is what you always seem to use.

    Quote Originally Posted by nsacpi View Post
    Also related to your point I believe that prospects very far away from the majors get dinged in the current prospect ratings relative to the ones that are older and closer to the majors. In an imperfect way the ratings do seem to take into account uncertainty.
    Yes, FG says they will typically lower the FV value of a guy in the lower leagues because their is more uncertainty. But I guess not all players...

  7. #46
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    2,641
    Thanks Thanks Given 
    1,547
    Thanks Thanks Received 
    1,645
    Thanked in
    878 Posts
    Quote Originally Posted by nsacpi View Post
    What is an 80 grade to you.
    Also, to answer this in the more traditional way rather than the FG way, putting an 80 OFP on a prospect would be equivalent to "More likely than not, this player is the next Willie Mays."

  8. #47
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    My point is ideally a GM will want to use a model of valuation of both major leaguers and minor leaguers that takes into account expected value, variance and covariance. I kind of doubt it is fully worked out like that because for most purposes expected value alone is good enough.

  9. #48
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    Quote Originally Posted by Metaphysicist View Post
    Also, to answer this in the more traditional way rather than the FG way, putting an 80 OFP on a prospect would be equivalent to "More likely than not, this player is the next Willie Mays."
    Then I think the standard needs to be redefined so the top guy each year is a Mookie Betts in terms of expected value.

  10. #49
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    Better yet FG should look at the history of its top prospect and define the value based on that history.

  11. #50
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    26,261
    Thanks Thanks Given 
    34
    Thanks Thanks Received 
    10,000
    Thanked in
    6,108 Posts
    Quote Originally Posted by Metaphysicist View Post
    Also, to answer this in the more traditional way rather than the FG way, putting an 80 OFP on a prospect would be equivalent to "More likely than not, this player is the next Willie Mays."
    An 80 grade prospect is not the same thing as an 80 grade player.

    I'm not sure why folks are confusing these concepts with each other.

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

    Snowman (02-13-2020)

  13. #51
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    26,261
    Thanks Thanks Given 
    34
    Thanks Thanks Received 
    10,000
    Thanked in
    6,108 Posts
    Quote Originally Posted by nsacpi View Post
    Better yet FG should look at the history of its top prospect and define the value based on that history.
    You mean exactly like they did here?

    https://blogs.fangraphs.com/an-updat...ect-valuation/

    Prospects are graded based on their expected surplus value. If they project Wander for $120M+ in surplus value, then he is, by definition of the metric, an 80 grade prospect. The methodology is literally spelled out objectively for all to see.

    Again...80 grade prospect does not mean a future 80 grade player. Perhaps EL is making an error in his projections for Wander, but I would argue there have been too few 80 grade prospects over the course of this FV grading system.

  14. #52
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    Quote Originally Posted by Enscheff View Post
    You mean exactly like they did here?

    https://blogs.fangraphs.com/an-updat...ect-valuation/

    Prospects are graded based on their expected surplus value. If they project Wander for $120M+ in surplus value, then he is, by definition of the metric, an 80 grade prospect. The methodology is literally spelled out objectively for all to see.

    Again...80 grade prospect does not mean a future 80 grade player. Perhaps EL is making an error in his projections for Wander, but I would argue there have been too few 80 grade prospects over the course of this FV grading system.
    More or less. But a couple things to note:

    1) They use WAR over first nine seasons (including minors starting with the season the player was ranked). Why not use the pre-free agency WAR numbers.

    2) I think also they need to divide up their samples to try to detect improvements that get made over time in identifying and developing prospects. This can also possibly pick up other changes in the game (testing for PED, increase in velocity which works in favor of value being concentrated in the early years of hitters) that affect the value of those pre-free agent seasons.
    "I am a victim, I will tell you. I am a victim."

    "I am your retribution."

  15. #53
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    2,641
    Thanks Thanks Given 
    1,547
    Thanks Thanks Received 
    1,645
    Thanked in
    878 Posts
    Quote Originally Posted by Enscheff View Post
    An 80 grade prospect is not the same thing as an 80 grade player.

    I'm not sure why folks are confusing these concepts with each other.
    That has nothing to do with what you quoted.

  16. #54
    Not Actually Brian Hunter Metaphysicist's Avatar
    Join Date
    Jul 2013
    Posts
    2,641
    Thanks Thanks Given 
    1,547
    Thanks Thanks Received 
    1,645
    Thanked in
    878 Posts
    Quote Originally Posted by Enscheff View Post
    You mean exactly like they did here?

    https://blogs.fangraphs.com/an-updat...ect-valuation/

    Prospects are graded based on their expected surplus value. If they project Wander for $120M+ in surplus value, then he is, by definition of the metric, an 80 grade prospect. The methodology is literally spelled out objectively for all to see.

    Again...80 grade prospect does not mean a future 80 grade player. Perhaps EL is making an error in his projections for Wander, but I would argue there have been too few 80 grade prospects over the course of this FV grading system.
    Another old man rant incoming...

    You know, I think I'm coming to understand why you have so many sniping haters on this board. It's been awhile since I've been super-active here, so I thought it was strange, but seemingly everything you post contains at least a few of the same traits:

    1) Condescending, as if what you are saying should be obvious.
    2) Not really bothering to read what you are responding to.
    2) Vaguely tethered to something FanGraphs once wrote.
    3) Usually wrong on the merits or details or both.

    I can understand why this would annoy people. It certainly annoys me. And I am myself a longtime condescending stathead, so you should really be my people.

    For example here, you have repeated a couple points: (a) 80 FV = $120M surplus value, and (b) 80 grade prospects ≠ projecting future 80 grade players. And you have asserted this as if is either the objective reality or the actual framework employed by Fangraphs. Everything you've said I guess seems facially sensible, but it is all aggressively wrong:

    - In terms of traditional OFP scores, point (b) is just 100% wrong. A player is graded on both his present tools/skills and his future tools/skills and then the final score of OFP is exactly what it says, a projection of "Overall Future Potential." If you give a guy an OFP of 80 that is literally you saying "in the future I expect this guy will be an 80 player." That's why 80 OFPs don't ever get handed out (to prospects at least; again, you could easily justify putting that on Mike Trout at 22), but for even the best prospects that kind of a crazy thing to say. Now FG is free to do their own thing, but that the normal scouting parlance that they are referencing (embracing?), and presumably trying to improve upon. If you refuse to recognize this basic fact, you are just not living in reality and should not be expressing any opinion on the topic. Further Reading: The Scouting Scale

    - In terms of fangraphs FV, both (a) and (b) are also 100% wrong. Let's look at (a) 80 FV = $120 surplus value. How do I know that what you are saying is wrong? Because as you (condescendingly) note: "The methodology is literally spelled out objectively for all to see." Here are the actual numbers from the article you linked to:





    So the first, and very obvious point is that these numbers don't say anything about 80 FV. 80 FV players have never existed. There are no expected WAR calculations for them. There are no surplus value calculations for them. Everything you've said about what 80 FV is just explicitly made up and not in this article. Second, $120 surplus value is actually almost exactly the 70 FV number plus inflation since 2018. So if anything, the article directly contradicts what you've said. Each move up the rung, 50 -> 60 -> 70 -> 80, is a standard deviation. You will note the value basically doubles each jump:

    50: 21M (3.5 Surplus WAR)
    60: 55M (6 Surplus WAR)
    70: 112M (13 surplus WAR)
    80: ?

    A similar jump would mean an 80 FV player would have ~$224m surplus value (or about 25 surplus WAR). Extremely different than what you've stated.

    The second point is that, elsewhere fangraphs has actually said explicitly what 80 FV means, and it is absolutely not "$120 surplus value." Here's an actual explanation linked to in the actual top 100 explanation: (link - The New FanGraphs Scouting Primer). Here is what Fangraphs actually says 80 FV means.





    Thus, by FanGraphs own methodology, an 80 FV means projecting to average 7 WAR per year over the first six seasons. That means at least 42 WAR over six years. That is FAR beyond $120M surplus value. An 80 FV will have already exceeded 13 surplus WAR (equivalent to $120M) after only their second year. (See Aaron Judge, Mike Trout)

    Again, that's FanGraphs own statement about what they said that are doing with FV. That's why an 80 FV on Franco is ridiculous, judging by their own principles. Escheff you are free to construct your own framework around whatever you think makes sense. Make a system where their are 2 80s every year. Knock yourself out, that sounds interesting. But it just incredibly annoying for you pop in add just flagrantly wrong information about the topic being discussed. And it is double annoying for you to be condescending about it.
    Last edited by Metaphysicist; 02-13-2020 at 03:00 PM.

  17. The Following 3 Users Say Thank You to Metaphysicist For This Useful Post:

    Carp (02-13-2020), chop2chip (02-13-2020), jpx7 (02-13-2020)

  18. #55
    Steve Harvey'd
    Join Date
    Sep 2015
    Posts
    18,946
    Thanks Thanks Given 
    1,856
    Thanks Thanks Received 
    5,329
    Thanked in
    3,353 Posts
    What the hell is this novel that took the whole page. Did Clv get hold of someone’s account. Jesus can someone let me know if it is worth reading.
    Coppy

  19. The Following 2 Users Say Thank You to bravesfanMatt For This Useful Post:

    jpx7 (02-13-2020), Metaphysicist (02-14-2020)

  20. #56
    Expects Yuge Games nsacpi's Avatar
    Join Date
    Jul 2013
    Posts
    47,430
    Thanks Thanks Given 
    2,704
    Thanks Thanks Received 
    11,384
    Thanked in
    7,533 Posts
    I think FG has made things unnecessarily confusing by using the grading system that scouts use.

    Here is a similar system of scoring the expected surplus value of top 100 prospects by their ranking that doesn't have that problem.

    https://www.thepointofpittsburgh.com...dated-edition/
    "I am a victim, I will tell you. I am a victim."

    "I am your retribution."

  21. #57
    Making Atlanta Great Again!
    #MAGA!

    Promises MADE, Promises KEPT!
    The Chosen One's Avatar
    Join Date
    Jul 2013
    Location
    School of Hard Cox
    Posts
    25,337
    Thanks Thanks Given 
    8,593
    Thanks Thanks Received 
    9,752
    Thanked in
    5,746 Posts
    Quote Originally Posted by Metaphysicist View Post
    Another old man rant incoming...

    You know, I think I'm coming to understand why you have so many sniping haters on this board. It's been awhile since I've been super-active here, so I thought it was strange, but seemingly everything you post contains at least a few of the same traits:

    1) Condescending, as if what you are saying should be obvious.
    2) Not really bothering to read what you are responding to.
    2) Vaguely tethered to something FanGraphs once wrote.
    3) Usually wrong on the merits or details or both.

    I can understand why this would annoy people. It certainly annoys me. And I am myself a longtime condescending stathead, so you should really be my people.

    For example here, you have repeated a couple points: (a) 80 FV = $120M surplus value, and (b) 80 grade prospects ≠ projecting future 80 grade players. And you have asserted this as if is either the objective reality or the actual framework employed by Fangraphs. Everything you've said I guess seems facially sensible, but it is all aggressively wrong:

    - In terms of traditional OFP scores, point (b) is just 100% wrong. A player is graded on both his present tools/skills and his future tools/skills and then the final score of OFP is exactly what it says, a projection of "Overall Future Potential." If you give a guy an OFP of 80 that is literally you saying "in the future I expect this guy will be an 80 player." That's why 80 OFPs don't ever get handed out (to prospects at least; again, you could easily justify putting that on Mike Trout at 22), but for even the best prospects that kind of a crazy thing to say. Now FG is free to do their own thing, but that the normal scouting parlance that they are referencing (embracing?), and presumably trying to improve upon. If you refuse to recognize this basic fact, you are just not living in reality and should not be expressing any opinion on the topic. Further Reading: The Scouting Scale

    - In terms of fangraphs FV, both (a) and (b) are also 100% wrong. Let's look at (a) 80 FV = $120 surplus value. How do I know that what you are saying is wrong? Because as you (condescendingly) note: "The methodology is literally spelled out objectively for all to see." Here are the actual numbers from the article you linked to:





    So the first, and very obvious point is that these numbers don't say anything about 80 FV. 80 FV players have never existed. There are no expected WAR calculations for them. There are no surplus value calculations for them. Everything you've said about what 80 FV is just explicitly made up and not in this article. Second, $120 surplus value is actually almost exactly the 70 FV number plus inflation since 2018. So if anything, the article directly contradicts what you've said. Each move up the rung, 50 -> 60 -> 70 -> 80, is a standard deviation. You will note the value basically doubles each jump:

    50: 21M (3.5 Surplus WAR)
    60: 55M (6 Surplus WAR)
    70: 112M (13 surplus WAR)
    80: ?

    A similar jump would mean an 80 FV player would have ~$224m surplus value (or about 25 surplus WAR). Extremely different than what you've stated.

    The second point is that, elsewhere fangraphs has actually said explicitly what 80 FV means, and it is absolutely not "$120 surplus value." Here's an actual explanation linked to in the actual top 100 explanation: (link - The New FanGraphs Scouting Primer). Here is what Fangraphs actually says 80 FV means.





    Thus, by FanGraphs own methodology, an 80 FV means projecting to average 7 WAR per year over the first six seasons. That means at least 42 WAR over six years. That is FAR beyond $120M surplus value. An 80 FV will have already exceeded 13 surplus WAR (equivalent to $120M) after only their second year. (See Aaron Judge, Mike Trout)

    Again, that's FanGraphs own statement about what they said that are doing with FV. That's why an 80 FV on Franco is ridiculous, judging by their own principles. Escheff you are free to construct your own framework around whatever you think makes sense. Make a system where their are 2 80s every year. Knock yourself out, that sounds interesting. But it just incredibly annoying for you pop in add just flagrantly wrong information about the topic being discussed. And it is double annoying for you to be condescending about it.
    Forever Fredi


  22. The Following User Says Thank You to The Chosen One For This Useful Post:

    chop2chip (02-13-2020)

  23. #58
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    26,261
    Thanks Thanks Given 
    34
    Thanks Thanks Received 
    10,000
    Thanked in
    6,108 Posts
    He’s trying to prove me “wrong”.

    I say an FV 80 prospect is not the same thing as an 80 grade hitter.

    He then posted a pic showing how FG defines an 80 grade hitter, despite the fact I literally just explained why those aren’t the same thing.

    The FV scale is set on the 20-80 grade scale because folks in the scouting sphere are comfortable with it. Mapping projected surplus value to that scale is the basis of the FV scale.

    Again, for the 15th time....

    A FV 80 PROSPECT IS NOT THE SAME THING AS AN 80 GRADE HITTER.

    I linked the article explaining the FV scale, and it should be clear to the reader why it’s different than an 80 grade hitter.

  24. #59
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    26,261
    Thanks Thanks Given 
    34
    Thanks Thanks Received 
    10,000
    Thanked in
    6,108 Posts
    Quote Originally Posted by Metaphysicist View Post

    Thus, by FanGraphs own methodology, an 80 FV means projecting to average 7 WAR per year over the first six seasons.
    Wrong. Completely and utterly wrong.

    A FV 80 PROSPECT IS NOT AN 80 GRADE HITTER

    Trout was not an 80 grade prospect, but turned into an 80 grade player. Part of that was lack of info on a a cold weather player, and part was due to him not being a prospect for very long.

    Wander may be an 80 grade prospect who turns into a 60 grade player due to variance in player development.

    Until this misunderstanding is cleared up there is no point in discussing the topic further.
    Last edited by Enscheff; 02-13-2020 at 03:54 PM.

  25. #60
    It's OVER 5,000!
    Join Date
    Jul 2013
    Posts
    26,261
    Thanks Thanks Given 
    34
    Thanks Thanks Received 
    10,000
    Thanked in
    6,108 Posts
    Quote Originally Posted by nsacpi View Post
    I think FG has made things unnecessarily confusing by using the grading system that scouts use.

    Here is a similar system of scoring the expected surplus value of top 100 prospects by their ranking that doesn't have that problem.

    https://www.thepointofpittsburgh.com...dated-edition/
    Pretty much, but it seems some are willingly letting themselves be confused by a concept that isn’t very confusing.

    And we are well past the point where they will concede. All that’s left now further convoluted attempts at proving themselves “right” with longer and longer posts.

Posting Permissions

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