Scouting Inaccuracy in Inaugural draft?

[blah37]

I think the scouting inaccuracies would be pretty realistic in real life, as you don't have any statistics to go by on. Judging players on talent alone in real life might give you similar results. Maybe it would be best to sim 5-10 years in the future then draft, and you'll have stats to work with. Then you'd have a better idea if the guy your director thinks is a 1st rounder is really nothing more than a bench player.

*edit: sorry, just saw that suggestion was already thrown out there.

i don't think the scouting system is broken. i like it the way it is.

[ScottPoseSnag]

Originally Posted by RonCo

Data:

MLB - Only 19 players have been inducted in the MLB HoF that came through the draft, so that sample size is a little small. Some error could exist in it.

Let me first say that RonCo's analysis was fantastic and very intriguing. And while this thread has mostly been about inaugural drafts, scouts matter very heavily in amateur drafts as well. I did some follow up work on where MLB Star players come from.

Here is a quick breakdown of the 2007 MLB All-star teams. I did not include closers or relievers, which you can feel free to argue with me about whether this makes sense to do or not. I noticed two trends (stats to follow). 1). Position players were very likely to have been first rounders, while pitchers were more "scattered" in their selection slot. (Peavy : 15th Round, Webb: 8th round). 2.) 1rst round players dominated.

With players such as Big Papi and Valdamir Guerrero being signed as free agents when they were entereing professional baseball, there were 41 allstars left that were drafted.

1rst Rounders : (20) 48.8%
2nd Rounders : (6) 14.6%
3rd Rounders : (3) 7.3%
4th+ : (12) 29.3%

There were only 4 position players drafted 4th or later, with 14 selected in the first round. There were 8 starting pitchers drafted 4th or later, versus 6 starting pitchers drafted in the first round.

So what does all of this mean? I'm not entirely sure, except to say that a vast majority of 2007 All-stars come from the first three rounds of the draft. (70%), and most of those come from the first round. Its pretty clear that first rounders dominate at the All-Star level, and therefore should be pretty easy to see who exactly these guys are, even if we cant tell which ones will become Prince Fielder and who will become Adrian Gonzalez.

Also, top flight SP seem to come from a more "shotgun" spread of rounds, which could possibly be because they are harder to scout or perhaps they are injured so much more often and seriously.

[RonCo]

Personally, I don't really separate the idea of amateur drafts and innaugural drafts. I think of the innaugural draft more like the elder days of baseball when you knew even less about players than you do in today's amateur draft. There _is_ no MLB for these guys to play in, so the average 10-year-old knows nothing about these guys. Blah, blah, blah.

I think it's understood that the majority of Hall of Very Good players come from the first round, but oddly, the elite come from all over the place.

Regardless, the important thing in the data I've looked at is that (I think) there should be a difference in the population of of the HoF per draft round when scouts are used and when they are not. Otherwise, I don't get it. What I see says that either scouts are too accurate, or they don't make a difference one way or another, or you scouts are working whether you turn them off or not. I would expect this to be a problem for innaurgural drafts, amateur drafts, or any other activity in which scouts are actively used.

[ScottPoseSnag]

Originally Posted by RonCo

Regardless, the important thing in the data I've looked at is that (I think) there should be a difference in the population of of the HoF per draft round when scouts are used and when they are not. Otherwise, I don't get it. What I see says that either scouts are too accurate, or they don't make a difference one way or another, or you scouts are working whether you turn them off or not. I would expect this to be a problem for innaurgural drafts, amateur drafts, or any other activity in which scouts are actively used.

I wonder if your findings support the case I found where a player was rated 3.5 stars by their scouts and yet they drafted him #1 overall in the ammy draft. Perhaps scouts aren't being used by the AI. Which would seem to be confirmed by your tests...wouldn't it?

When I get some time Ill test some more ammy drafts and see who and how the AI are picking.

[RonCo]

Originally Posted by ScottPoseSnag

I wonder if your findings support the case I found where a player was rated 3.5 stars by their scouts and yet they drafted him #1 overall in the ammy draft. Perhaps scouts aren't being used by the AI. Which would seem to be confirmed by your tests...wouldn't it?

I make no judgment about the root dynamic for why I see what I see. There are several possible reasons for it, and only Markus can say for sure what it is.

[ScottPoseSnag]

Just a follow up on another draft.

Picking #1: AI team. The player they selected was the best player in the draft, but they had him rated as their #2 or #3 guy. (There was a guy with 5 stars, and their scout rated the player they chose as only 4.5 stars.) Does this prove anything? Not really, although it suggests the computer is picking based on real ratings, not scouted ratings.

Picking #2: My Team. Took a catcher. My scouts had him as being pretty good, and by "cheating" and checking his ratings, he was definitely a top 5 pick, but probably not the best #2 pick.

Picking #3: AI team. Selects the clear #2 overall prospect in the draft. When I switched to that team as a human manager, they had him (and the #1 guy selected overall) rated as 1 star potential players. So essentially a team just selected #3 in the ammy draft the actual best pick even though their scouting had no clue that this guy was any good. After checking their scouting, they were spending a total of 100k on amateur scouting, well below the default averages. Yet they still managed to "see through" their scouting problems and select a 5 star player they had rated as a 1 star.....as if they had access to the real ratings, and where ignoring their own scouting.

Which, after look at RonCons long term results, suggests that perhaps AI teams have access to the real ratings of players when making ammy draft decisions, and are not using scouted ratings of players. If this is true, does this extend to all decisions the AI makes?

Of course the other possible explanation (that I can think of) is that the scouts and budgets and scouting is changing when I switch teams to human managers......but that doesnt explain roncons results.

[fantom1979]

Originally Posted by ScottPoseSnag

I wonder if your findings support the case I found where a player was rated 3.5 stars by their scouts and yet they drafted him #1 overall in the ammy draft. Perhaps scouts aren't being used by the AI. Which would seem to be confirmed by your tests...wouldn't it?

When I get some time Ill test some more ammy drafts and see who and how the AI are picking.

I was wondering this as well. Maybe the scouts are not too accurate, maybe the computer run teams are seeing through those inaccurate ratings instead.

[deadringer]

Can we get some official word on this please? Is this broken or by design?

[deadringer]

Bump because I see that Markus is back.

[ScottPoseSnag]

Any on word on this? Specifically, if scouting is on, does the AI use scouting? Or does it just "see through" the scouted ratings and use real ratings?