The argument about the draft seems to counterdict the original point this thread tried to make. 

The original point of this thread was to argue that projections seen should be more in line with a reasonable expectation of what can be achieved with "proper development".

If that were indeed the case, then would you not expect the results of the HBD draft to be even more top-heavy than it is right now?

You can't have it both ways (more accurate projections AND more potential ML players disbursed throughout the lower rounds of the draft).  That just doesn't work.

I took another look at this because the criteria used to determine who is a "Big Leaguer" in the first analysis may have been easier to meet for the MLB players (basically any service time at the BL level) which resulted in the MLB list containing 133 BL'ers versus only 82 on the HBD list.  This time the MLB* column only includes players who recorded >130 AB or 50 IP which is MLB's current maximum's for ROY consideration and results in 83 total BL'ers, which is nearly the same as the 82 BL'ers on the HBD list and suggests that the new criteria is on the right track.   This list changes the numbers, but I think the overall pattern still holds true that the myth is not a myth.

Rd.	HBD	MLB*
1	35	17
2	22	12
3	13	5
4	4	7
5	3	3
6	1	4
7	1	1
8	1	4
9	0	2
10	1	1
11	0	5
12	0	4
13	1	1
14	0	2
15	0	0
16	0	1
17	0	5
18	0	0
19	0	1
20	0	1
21	0	0
22	0	2
23	0	0
24	0	3
25	0	2

Posted by MikeT23 on 4/6/2011 10:59:00 AM (view original):
I think it's pretty obvious that the talent is taken at the top in both drafts.    For all we know, DITR handles the 14 players taken in rounds 18-25 in HBD. 

I'll just repeat this and add that this is a sim.  Expecting exact results in every world to mirror MLB is an unachieveable goal.   I might also add that expecting the HBD draft, where the players have little numbers next to their names, to mirror MLB drafts is not possible. 

Anyway, if you want to make the draft more "realistic", you have to do several things.

1.  Quit limiting the rounds.  Draft as long as you want.
2.  Do away with projections as we know them.  You'd get current and projections for PHYSICAL ratings only. 
3.  Add some HS/College "stats".

The draft would be a lot more guesswork but you'd find that owners would STILL draft the PHYSICAL beast before the others and the cream would rise to the top again.

Shouldn't DITR players show up in this analysis or have they not been around long enough?

I don't know that I'm taking a side as much as making observations.  And yes, if we want to mirror MLB precisely we'd probably have to simulate the owners' behavior as well, but I can't think it hurts anything to interrogate the differences.  tecwrg's point that the data contradict the original point of the thread is a solid one.  The data imply that either

1) HBD projections are too accurate and/or development is too consistent.

2) HBD owners are more likely to promote players to the BL's based on projections rather than performance (i.e. they assume projections are 100% accurate).

or 3) MLB teams seriously need to increase their scouting budgets.

I think there's probably some truth to #2, especially among less experienced owners (e.g., me, until recently, I hope...), as I see a number of guys sporting 0.900+ minor league OPS or defensive specialists that never even get a shot at the bigs (and may have...ahem...in the past...traded away a few at below market value.)

Another way to look at this data, which may be more revealing, is to list the percentage of big leagues still available at each round.  In HBD only 30% of BL'ers are still available after the 2nd round vs. 70% in MLB.  By round 4 only 10% of HBD BL'ers remain while in MLB 50% are still on the board and MLB doesn't reach the 10% remaining mark until around the 17th round. 

I love the HS/College stats idea.

I think you're missing my point.  

The step up in competitiion in real life exposes a lot of flaws.   The kid who mashed HS fastballs is suddenly overwhelmed by a slider.   While an 78/74 con/pow may never become 94/87 in HBD, he'll still be a 78/74 and hit like one.    The better competition won't matter because his numbers are 78/74 or, if he develops at all, better than 78/74.    The kid who hit .620 with 14 homers in 29 games in HS can't hit .200 in real life because he just can't hit that slider so that's all he sees.

As long as we have accurate current ratings, you don't get Nate McClouth in the 25th round because his current is better than 30/16 con/pow.

I guess the best way to put it is this:   Unless you turn development and the draft into some random spin of the wheel, the people with most time/intelligence will figure it out if given enough info.  So, no matter what you do beyond random randomness, the best players will be drafted at the top in HBD drafts.

And if the draft were to become more random, I will stop spending money on HS/Coll scouting.

I think that's something people don't realize.   If I'm getting busts in the first round so I can get something in the 19th round, I'll let the other guy figure it out.    I'll use that 20m in scouting for a couple of Type A FA.  Some owners already employ that strategy.  I think more would do the same if we didn't know we could get a BL player in the first/second rounds.

Again, I'm not arguing for or against anything, just looking at what is. 

The original comment in this suggestion thread was that either the projections or player development needed to be "fixed" because players seldom achieved their "projections" and that wasn't realistic since in real-life players often exceed their "projections".   The examples provided to support this were players like Maddux (2nd rd) and Pujols (13th rd) who were drafted in later rounds.  Which eventually led to this idea that since BL'ers congregated at the top of the HBD draft then HBD projections were more accurate than MLB projections, but I don't think that's quite right. 

It's curious that more investment in HS/Col scouting would mostly lead to more accurate projections since in MLB I expect more investment mostly means more scouts and more geographical coverage and just maybe some more intensive coverage of the very top guys.  Pujols and Maddux didn't fall to the later rounds because their projections were inaccurate (See Maddux's scouting reports www.stevenellis.com/steven_ellis_the_complete/2009/02/wanna-see-greg.html,) they fell to later rounds because they either weren't scouted by very many teams or there were lingering questions, in Maddux's case his physical size and in Pujols', his actual age and the level of competition they faced in compiling their stats.   Some of this realism is captured in HBD by concealing some portion of the draft pool from each team relative to their scouting budgets.  IF, and I do mean if since fun generally decreases as simulations approach actual realism, you really wanted to see some quality BL'ers fall to later rounds (and I think you have to admit it would be fun every now and then to feel like you'd stolen a quality player) the way to do it would probably be to increase the level of concealment rather than decrease projection accuracy.   

"And if the draft were to become more random, I will stop spending money on HS/Coll scouting."  Some owners would, but, at least in the case of increased concealment, that would leave better players to later rounds for the teams that were still investing in scouting.  I don't think it means that everyone would stop investing in scouting.  

Not ready to call this a good idea, just an idea.   I still think the real problem is that people can't get their head around the idea that the word "projections" as used by HBD is closer in meaning to "ceilings" than "expected outcomes". 

If you want a stud to drop to round 17, concealing him from 31 teams won't make that happen.    The one team that can see him will rank him high.   The ONLY way a stud can drop in the draft is via randomness.  

Yep.

"This dude looks AWESOME! I'm going to roll the dice and bet he's still around at pick 425"..

Not ready to call this a good idea, just an idea.   I still think the real problem is that people can't get their head around the idea that the word "projections" as used by HBD is closer in meaning to "ceilings" than "expected outcomes".

That's fair.  I'm OK with the way it is though.  When I draft someone whose glove is currently 58, and it's "projected" at 90, I know the "projection" is most likely impossible to hit, and I control my expectations.  It's definitely not a ceiling though, because I've seen many guys exceed certain ratings.

That's the thing.  Everyone remembers the 98 that never was but quickly overlooks the 60 that became a 63.

People should temper their expectations when they see a mid to upper 90s number because, quite frankly, that's a top 1% in a world.   Not very many players are going to be top 1% in anything.