Combine/Drafting strategies in 7.2?

[ironsickel]

Hey all,
With 7.2 dropping, anybody had a chance to play with the draft with difficulty below 40?

Thoughts? Reactions? Strategies?

It seems like there are a lot more players now with negative combines that still look draftable. I'll need to play some more, but very curious what other people's experience has been.

[ironsickel]

At first blush, it would appear that the old hard thresh holds are gone.

In my first draft, here's a very nice surprise I found in the middle of the third round:
http://i.imgur.com/6T1R61b.jpg

I normally would not have drafted this guy because the one combine score he did have was below the old thresh hold. Knowing that Jim supposedly "fixed" this...I decided to take a chance because his bars looked intriguing.

As of EX, however, he is currently the second highest potential QB in the draft at 24/69, behind only the #1 overall pick, who my scouts currently see as 20/78.

[Ben E Lou]

At first blush, it would appear that the old hard thresh holds are gone.Well, yeah. Here's the "bottom" of a draft class of 100% 60/60 WRs (with 60 for every bar).

http://www.fof-ihof.com/upload/Ben%20E%20Lou/wrs.png

[ironsickel]

Well, yeah. Here's the "bottom" of a draft class of 100% 60/60 WRs (with 60 for every bar).

http://www.fof-ihof.com/upload/Ben%20E%20Lou/wrs.png

WHAT DOES IT ALL MEAN, BEN?

I mean...everything I've learned in the last few months of playing are basically from your old posts.

My first draft was...ok. That 3rd round QB was a real steal, and my 1st rounder was pretty good, but the rest were all average joes...

[ironsickel]

Is there basically no correlation to combine anymore and we should ignore combines and go for bars?

[Ben E Lou]

Is there basically no correlation to combine anymore and we should ignore combines and go for bars?I don't think so. Let me analyze the rest of that class. I just showed you the very bottom guys. Keep in mind we're talking 1,000 WRs there.

[corbes]

My guess is that there is and ought to be correlation; the change would be that DASH>4.51 is no longer a kiss of death but rather a piece of information to be considered alongside all other pieces of information.

[wustin]

<a href="http://i.imgur.com/k7caZio.png"><img style="max-width:368px; max-height; 465px;" src="http://i.imgur.com/k7caZio.png" alt="Image" /></a>
You wanna see steals? I will attribute every team passing up on him because he's 5'10, character is red flag, and very high volatility. Also the QBs in his draft class was very weak (the first QB was drafted late in the 2nd round).

[ironsickel]

<a href="http://i.imgur.com/k7caZio.png"><img style="max-width:368px; max-height; 465px;" src="http://i.imgur.com/k7caZio.png" alt="Image" /></a>
You wanna see steals? I will attribute every team passing up on him because he's 5'10, character is red flag, and very high volatility. Also the QBs in his draft class was very weak (the first QB was drafted late in the 2nd round).

Don't see his combine data in there. What did it look like?

[wustin]

Don't see his combine data in there. What did it look like?

I don't remember. That late into the draft, I just sort the players by bars with the analyzer and look at whoever potentially good is left available. That QB specifically was low developed/high bars. Since it was late-6th round, virtually no risk.

[ironsickel]

I don't remember. That late into the draft, I just sort the players by bars with the analyzer and look at whoever potentially good is left available. That QB specifically was low developed/high bars. Since it was late-6th round, virtually no risk.

Too bad you didn't import that data into player tracker so it would gave grabbed the combine numbers and bars.

[aston217]

Here's how I would like it to work: meaningful randomness.

In other words, follow intuitions such as high bars/grades/combines = good, with some understanding that things can be all over the map.

It sounds like that's what Jim is going for. Hopefully the uncertainty can be quantified, too -- but needing to watch at all for "7.82 agility, oh, that's terrible because I know how to consult the Oracle's charts" being replaced by "Everyone can see this guy is an 80% lock, this guy is a 50-50 prop, and this guy is a 10% stab in the dark" type of thing.

You know what else wouldn't hurt, having maybe 400-man draft classes instead of 800. Don't really need that many rookies, IMHO. Part of the problem as it were is that the draft is too much of a "Challenge: dig through 80 safeties and know which 50 to discard" deal. New players who haven't the eye for the minutiae yet, end up simply overwhelmed by sheer volume. Want to level it for them, discard the noise and make everyone's life easier [this was the whole point of Stelmack's project, as I understand it].

[Ben E Lou]

You know what else wouldn't hurt, having maybe 400-man draft classes instead of 800. Don't really need that many rookies, IMHO. Part of the problem as it were is that the draft is too much of a "Challenge: dig through 80 safeties and know which 50 to discard" deal. New players who haven't the eye for the minutiae yet, end up simply overwhelmed by sheer volume. Want to level it for them, discard the noise and make everyone's life easier [this was the whole point of Stelmack's project, as I understand it].I've always assumed that the classes are that big so that there's always a supply of minsal roster-filler rookies in case of massive injuries or a bunch of teams mismanaging the cap.

[aston217]

Hm, possibly -- aren't there *plenty* of FAs to cover, though? Ex: Week 1 OSFL sees in FA 90+ QBs, 100+ WRs, 170+ OL.

If there's a danger in lowering that, then maybe (down the road; I know it's a lot of work) tweak some of these roster requirements. A team with 8 healthy OL currently has to sign a FA (or position switch - ugh) if their two centers are out. An NFL team can apparently carry 6 OL into a game, and the nominal T/G/C counts are irrelevant.

Also, doesn't AI assist already ignore cap/demands where necessary? If the spare numbers can somehow not have to be rookies, then perhaps the game could stop vanishing all those youngish players at every season turn, reduce rookie pools, and maintain FA numbers anyway.

In FA too, though, I feel the combination of volume and noise overwhelms people. Now, maybe dealing is the Name of the Game, but I do think it gives rise to imbalance and calls to level the field (an interesting topic on its own). If Jim can find a way to reduce the numbers, it'll be a more accessible and pleasant experience for many, IMO -- so I'm kinda rooting for it!

[Ben E Lou]

Hm, possibly -- aren't there *plenty* of FAs to cover, though? Ex: Week 1 OSFL sees in FA 90+ QBs, 100+ WRs, 170+ OL.There are those who play with injuries set as high as 400, and in SP there are 31 AI-run teams that have to manage rosters in that setting. I tend to do SP with injuries either off or very low, so it's not my thing, but I can certainly see it being an issue. I just opened a year 2 career (year 1 had injuries off) and ran season 2 with injuries at 400. There were only 9 Cs, for example, by the end of FA1. After a few seasons at that setting and with a tighter cap increase than I was using, that number could easily push toward 0. (And no, that doesn't appear to be a huge anomaly. There are 19 DEs, 23 OTs, and 11 ILBs.)

And I'm not sure that AI teams get the minsal exception in SP. I think that's a MP-only thing to keep leagues from crashing because humans are capable of doing a much worse job than the AI.

[ironsickel]

As someone who's new to the game in the last couple months, the amount of players available in the draft was not an issue for me.

The biggest learning curve in the draft, for me, was figuring out why my #1 picks would only be fair or average while the computer kept getting G/VG players. Which is the entire point of the draft. I don't think reducing the number of draftees would actually make anything easier.

But then again, I don't WANT anything easier.

[ironsickel]

Allow me to follow up on my last post.

I want the draft to be very challenging, but not just be a crap shoot. It should reward you, either by player skill, or by in game decisions (such as your coaching/scouting staff hire decisions), but still have enough element of chance that there are few "sure things" (just like in the real draft).

To that end, in the few seasons I've done since 7.2 dropped (and i'm using the 39 combine accuracy), I'm very happy with how the draft seems to be working.

In my experience, it the best strategy, beyond the first round or so, where sure things are more likely, is to draft based on bars. Well, bars are highly dependent on scouts. And having good scouts means you've either prioritized monetization to afford the best coaches, or have made deliberate sacrifices in other coaching disciplines to hire the guys with the best scouting talent.

I'm okay with that.

I've also noticed that, even given that, there are still no sure things.
I'm, finding that I'm having a LOT more late round G/VG players, but at the same time, I'm having a lot more early round AVG players. And that feels (at least in my limited experience) much more like how the real draft plays out.

What was happening before for me, in SP, was all the G/VG players were typically gone by round 3 and what few good players left were very rare or VSOL players.

And in the real draft, franchises that value scouting manage to find late round gems all the time.

Not sure if my experience mirrors anyone elses though.

[Yoda]

Use the force.

[Ben E Lou]

Worth noting...We strongly recommend not going below 40 in the single-player game, as the AI for computer-controlled teams is not good at this level.

[ironsickel]

Worth noting...

Lame.

And it's true...they're not. But it hasn't diminished the experience for me.

I've found there's just a lot more late round gems for everybody.

There doesn't seem to be a competitive advantage because I'm not any better at figuring out the good from the bad than they are.

Btw, Ben, where did you see this quote posted originally?

[aston217]

Worth noting...

Damn.

Do you suppose that's because of the "problem fix", or the fact that the accuracy number is so far from 50? If the latter, I hope we can get to choose both 50 and "problem fix".

[Ben E Lou]

Damn.

Do you suppose that's because of the "problem fix", or the fact that the accuracy number is so far from 50? If the latter, I hope we can get to choose both 50 and "problem fix".I'm not following you. I thought you didn't play SP.

[Ben E Lou]

Btw, Ben, where did you see this quote posted originally?http://www.solecismic.com/support/index.php

[aston217]

I'm not following you. I thought you didn't play SP.

Do computer controlled teams behave differently in MP?

*Wait* -- nevermind. Conscriptor handles that in MP. Although that's something else to consider, I suppose -- I'm unfamiliar with what the existing logic for autopicks is.

[Ben E Lou]

I'm unfamiliar with what the existing logic for autopicks is.Yes, it's 100% Conscriptor-based. FOF's logic has nothing to do with it. (I always get a chuckle when some nitwit whines about how FOF screwed him on the missed pick. Ummm...no. I didn't manually enter 100+ picks to see who FOF would draft for you.) It seems fairly obvious that the Conscriptor logic is based on perceived positional need and adjusted rating. I'm pretty sure it *always* drafts the highest adjusted rating guy at a given position, so that's a strong indicator that it's a $positional_need*$adjusted_rating setup.

[aston217]

Ah, okay, thanks. Yeah, total brainfart for a moment there. That seems simple enough to just leave as is.