Fuzzy Projections

by Pajammies 12:19 PM 0 comments

Or, Why You'll (Likely) Never Draft a Prospect That Exceeds Your Projections. 

I'd like to balance out the new, exciting, optimistic blog re-design with a negative, pessimistic inaugural blog post.

Approximately 1 year ago, WhatIfSports introduced the largest Hardball Dynasty update in quite some time. Included in this update was an overhaul to the scouting system:

In the new system, the accuracy of the high school, college and international scouting departments is now different and less accurate than the advanced scouting department. And because current ratings are no longer visible for prospects, scouts can now under project a prospect's ratings. This means the shift in projected ratings once a player enters your advanced scouting department can be potentially larger in the past.

This also means the Amateur Draft will be less formulaic.  Budgets still matter very much, but there is more variability.  Can you occasionally have a scout be way off on a player with a $15 budget? Sure, but it'll be less likely than if you had a $10M or $5M budget. But it also means your scout may be wrong in a favorable way as well.

Prior to this update, the draft was relatively formulaic -- as long as you had a decent enough scouting budget and didn't get stuck with a terrible draft class, a top 10 pick would usually turn into an All-Star or better. This generally eliminated from HBD the concept of a "draft bust", but it also eliminated much hope of getting a top 10 talent at the end of Round 1 or later. The idea that you could draft a player that exceeds your scouting expectations was generally well-received.

In the year since the update, I have anecdotally heard many accounts of owners being disappointed by their scouting, and very few being pleasantly surprised. Examples of TRUE diamonds in the rough appear few and far between.

But as WiS stated above, scouts can now under-project a prospect's rating. I've personally experienced a prospects current rating in a category end up better than my scout's projected rating, so it certainly is possible. I'm sure if an owner compared every player scouted in the draft to the player's actual peak development rating, we'd find a good percentage that ended up better than the projections predicted. 

But will you ever draft such a player?

I don't think so.

For a given prospect, 31 other teams are potentially scouting that prospect. And statistically speaking, it's likely a good percentage of those teams either projects the prospect accurately, or over-projects the prospect. Assuming teams rank players based on their projections, that player is going to be ranked higher for other teams than for myself, and will therefore in all likelihood get drafted before he reaches me. 

An over-simplified example:

If we have a group of 10 players, one rated 80 and the other nine rated 70. I see the 80-rated guy as a 65, but I see the other nine accurately at 70. If I drafted him I'd be super stoked -- he's better than any of the others I'd seen!

However, if I draft before any other owners, I'm screwed -- I'll take one of the nine guys rated 70ish.

In order for me to take the 80-rated player, the following two conditions must be met:

Condition 1: I pretty much have to draft after everyone else to have hope of getting this guy, since my board will take any of the 70 guys before the "65" guy. Ok this is easy if you have a lower draft pick.

Condition 2: I also have to hope zero of the nine-plus owners drafting in front of me see that 80 guy as better than any of the 70 guys -- some combo of under-rating the 80 guy, over-rating the 70 guys, and perhaps not seeing the 80 guy. For all nine-plus owners.

So, in order to draft a player your scouts significantly under-rate, all teams drafting in front of you have to draft all your "incorrectly-higher-projected" players before your spot, AND either not see the "under-rated-good-guy" or also consistently under-project in similar fashion.

All that said, because projections are "fuzzier", there appears to be a higher chance a guy you accurately project as good could fall further than in the past, because so many other teams are over-projecting and under-projecting other players. But I still don't think we'll be hearing many stories of owners drafting top-ten talent at the end of Round 1 or lower.

Unless we can figure out WTF happened with Kolten Cook.