New scout every year?

[QuikSand]

Okay, first of all -- this is an idea that I'm not sold on myself. It will likely prove to be laborious, tedious, fraught with uncertainties, and all for very little "edge" if at all. Oh, and not at all "immersive." So, if that warning keeps you from making your eyes burn, great. Otherwise read on.


Okay, here's the basic premise of the idea. I'm assuming that whatever it is scouts do in this game (and I'm not really 100% clear on that, to be honest) they presumably do a better job in the position groups where they are highly rated. Pretty much stands to reason.

So... why not cycle out your scout every year with a new guy? Particularly if your team is in a league where you can't land the tip-top scout who's pretty good at everything, for whatever reason.

Here's the theoretical idea:

-You generally have at least soem idea what position(s) are your highest priority for drafting each year... so you could go out and get a "LB specialist" for that year, or a guy who is strong in two areas like RB and DL. Any maybe get such a guy without much competition, and/or more cheaply than the top guys would cost.

-Each year, by seeing how your guys are rated before and after a scout change, you would get some insight into what's going on with that player's scout error. If your starting WR goes from 60/60 with a "very good" scout to a 56/56 with a "fair" scout at that position... maybe that's more insight into him than you'd have by just getting the same guy's opinion year after year. You would, I think, always have the same scout in place to do a before/after perspective on the major pivot points (start of FA and training camp).


Anyway... requires some record-keeping, I suppose. But it might be better than just suffering through with whatever lame-ass guy you get stuck with once the big money teams have their pick of the litter.

[cuervo72]

Hmm. I'm not sure you'd even need that many, depending on how scout error works. I'm still of the mind that on most vet guys, there is a bit of a "dampening" effect...i.e. most of them have false masks downward, and the worse the scout the larger the mask. The best scouts show the true rating. My thought had been that this was formulaic - EX scout shows 100% of ability, VG might show 96%, G 93%, etc. Except for guys who have no masking at all, and are seen the same by everyone. So the value in having different scouts might be that you get to identify the guys who *don't* have error, rather than the guys who do (who you could just extrapolate).

I might have to look at the last few IHOF files. Some savvy owner swiped my scout, so I believe in the file before today's I'd be viewing my team with the "No Scout" - essentially a scout that is Poor in each area. I'll have to confirm this and relay any player "changes".

[gstelmack]

Many assumptions here. For one, there aren't that many scouts who are particularly good with young talent, which blows out the specialist concept. Plus scouts can err both ways, so your 60 -> 56 example doesn't necessarily say much. Also, didn't we determine a while ago that the difference between scouts views of players wasn't necessarily all that much, typically just a point or two?

[cuervo72]

Ok, when I first figured this, I guessed the following:

EX = 100% of true rating
VG = 97% of true rating
G = 94% of true rating
A = 91% of true rating
F = 88% of true rating
P = 85% of true rating (also "no scout")

Example:

Here's an example of a guy who "works" under my theory.

TE Ricky Wooden
Scout 1: 47/47 (G)
No Scout: 44/44
Scout 2: 49/49 (VG)

Ok, not a huge change here. But the "good" scout had him at 47, the default had him at 44, and the "VG" scout had him at 49. My theory would work something like...."Poor scout has him at 44, which is 85% of his real rating - somewhere around 51.8, give or take. 97% of this is 50.2, close to what Scout 2 said (maybe Scout 2 is a weak VG). 94% of this is 48.7, close to what Scout 1 said (maybe he's a weak G, even).

I have a WR who followed a similar pattern as the TE - 44 under the old scout (G), 42 under the No Scout (P), 47 under the new scout (VG). % might be different, but same general idea.

Another example is P Tony Amey. In 2k4, Amey was 100/100, which I believe is his true ability (makes him a good test dummy too). With my old scout, who was Poor with P/K, he was 89/89 (about where he came over in the conversion, which led me to look at this whole deal in the first place). With "No Scout", he was 88/88. So maybe my old scout was a little better than rock bottom. With this new guy, he is a 92. My new guy is Average for P/K, so this makes sense.

All of these changes are fairly small though, especially for guys who have lesser ratings (if it is % of ability, those with less ability can show less change). My K showed a similar, but smaller trend (61 -> 60 -> 62)


Where this gets turned upside-down is young players. Changes here seem to be flipped. For instance, QB Mel Wickland:

Scout 1 (G in YT): 51
No Scout: 45
Scout 2 (VG in YT): 52

Wickland had a bad first camp, and I have no doubt he will continue to decline. My intuition here is that the No Scout actually has it right - he is applying NO false positive mask, where the other "better" scouts actually are. Young players are weird like that.

[zullojer]

My strategy is more along the lines of not relying fully on my scout. I take his opinion in, but i make my own decisions. Instead of trying to improve my scout every year, or getting in a bidding war with someone over the top scouts. i look for a guy that's strong in young talent and o-line. I get better every off season at judging talent on my own. I look at my scout's opinion and factor that in with everything else (bars, pop, leadership, solcismic, height, weight, and scout opinion) etc. etc.

[Celeval]

A couple of additions to this:

- Unless things have changed for the less-complex in 2k7, scouts have the same 0-100 scale as any other attribute, it's just not shown (and abstracted to G/VG/Ex/etc.).

- Going /way/ back to TCY and working from memory, even 100-rated scouts weren't perfectly accurate for all players. I don't think they saw through masking, for one.

[cuervo72]

Originally Posted by Celeval

A couple of additions to this:

- Unless things have changed for the less-complex in 2k7, scouts have the same 0-100 scale as any other attribute, it's just not shown (and abstracted to G/VG/Ex/etc.).

- Going /way/ back to TCY and working from memory, even 100-rated scouts weren't perfectly accurate for all players. I don't think they saw through masking, for one.

Right...which is why it's not as easily grabbed in the data like it was before (to the dismay of my lovely php staff pages). Makes sense too considering you see changes from year to year (maybe even more often? might be interesting to check).

[cuervo72]

Ah, what the heck. Here are the tracker results for my entire team, going from old scout to no scout to new scout. My finding is that any player who had a positive change from step 1 to step 2 can be identified as being either a creeper or a *former* creeper who is now a vet. Those guys are oddballs, and seem to stay that way. The others are either regular vets or young guys who are not creepers.

Beaumont (TX) Player Report
(2018 Staff 01 - 2018 Staff 03)

Name	Pos	Team	2018 Staff 01 Current	2018 Staff 01 Future	2018 Staff 02 Current	2018 Staff 02 Future	2018 Staff 03 Current	2018 Staff 03 Future	Overall Current Change	Overall Future Change
Bensen, Cornell	QB	Beaumont	34	48	31 (-3)	48 (0)	32 (1)	48 (0)	-2	0
Marquardt, Rick	QB	Beaumont	39	45	37 (-2)	45 (0)	38 (1)	45 (0)	-1	0
Percy, Jackie	QB	Beaumont	81	81	81 (0)	81 (0)	81 (0)	81 (0)	0	0
Prior, Hunter	QB	Beaumont	23	33	24 (1)	35 (2)	23 (-1)	33 (-2)	0	0
Wickland, Mel	QB	Beaumont	20	51	20 (0)	45 (-6)	20 (0)	52 (7)	0	1
Bernius, Gino	RB	Beaumont	31	41	32 (1)	42 (1)	31 (-1)	41 (-1)	0	0
Buff, Riddick	RB	Beaumont	59	59	59 (0)	59 (0)	59 (0)	59 (0)	0	0
Cassity, Stephen	RB	Beaumont	62	62	63 (1)	63 (1)	62 (-1)	62 (-1)	0	0
Pope, Andre	RB	Beaumont	34	34	34 (0)	34 (0)	34 (0)	34 (0)	0	0
Brock, Emmitt	FB	Beaumont	67	67	68 (1)	68 (1)	67 (-1)	67 (-1)	0	0
Cooley, Calvin	TE	Beaumont	39	39	35 (-4)	35 (-4)	39 (4)	39 (4)	0	0
Murphy, Raymond	TE	Beaumont	15	23	16 (1)	25 (2)	15 (-1)	23 (-2)	0	0
Wooden, Ricky	TE	Beaumont	47	47	44 (-3)	44 (-3)	49 (5)	49 (5)	2	2
Bates, Billy	FL	Beaumont	37	37	33 (-4)	33 (-4)	37 (4)	37 (4)	0	0
Fromme, Gino	FL	Beaumont	27	39	26 (-1)	36 (-3)	27 (1)	40 (4)	0	1
Galloway, Monty	FL	Beaumont	19	27	20 (1)	28 (1)	18 (-2)	26 (-2)	-1	-1
Heyward, Tyrone	FL	Beaumont	63	63	63 (0)	63 (0)	63 (0)	63 (0)	0	0
Bernard, Michael	SE	Beaumont	44	44	42 (-2)	42 (-2)	47 (5)	47 (5)	3	3
Hancock, Albert	SE	Beaumont	35	35	33 (-2)	33 (-2)	36 (3)	36 (3)	1	1
Patterson, Ty	SE	Beaumont	46	46	43 (-3)	43 (-3)	48 (5)	48 (5)	2	2
Andersen, Britt	C	Beaumont	56	56	57 (1)	57 (1)	55 (-2)	55 (-2)	-1	-1
Castillo, Heath	C	Beaumont	46	46	46 (0)	46 (0)	46 (0)	46 (0)	0	0
Barton, Randy	LG	Beaumont	42	52	42 (0)	52 (0)	42 (0)	52 (0)	0	0
Probin, Donnell	LG	Beaumont	22	32	22 (0)	33 (1)	22 (0)	32 (-1)	0	0
Wagner, Ty	RG	Beaumont	60	60	54 (-6)	54 (-6)	60 (6)	60 (6)	0	0
Woodson, Claude	RG	Beaumont	53	53	53 (0)	53 (0)	53 (0)	53 (0)	0	0
Fleming, Riddick	LT	Beaumont	53	53	53 (0)	53 (0)	53 (0)	53 (0)	0	0
Norton, Willie	LT	Beaumont	48	48	49 (1)	49 (1)	47 (-2)	47 (-2)	-1	-1
Franklin, J.R.	RT	Beaumont	47	47	45 (-2)	45 (-2)	48 (3)	48 (3)	1	1
Borders, J.R.	K	Beaumont	61	61	60 (-1)	60 (-1)	62 (2)	62 (2)	1	1
Amey, Tony	P	Beaumont	89	89	88 (-1)	88 (-1)	92 (4)	92 (4)	3	3
Dotson, Rufus	LDE	Beaumont	45	45	47 (2)	47 (2)	46 (-1)	46 (-1)	1	1
Moriarty, Glen	LDE	Beaumont	36	38	36 (0)	38 (0)	36 (0)	38 (0)	0	0
Neujahr, Roosevelt	LDE	Beaumont	52	52	49 (-3)	49 (-3)	51 (2)	51 (2)	-1	-1
Kirtsey, Daryl	LDT	Beaumont	50	50	48 (-2)	48 (-2)	50 (2)	50 (2)	0	0
Barber, Danny	RDT	Beaumont	50	50	47 (-3)	47 (-3)	47 (0)	47 (0)	-3	-3
Sutton, Willie	RDT	Beaumont	49	49	52 (3)	52 (3)	50 (-2)	50 (-2)	1	1
Bronson, Freddie	RDE	Beaumont	48	48	50 (2)	50 (2)	49 (-1)	49 (-1)	1	1
Hardy, Reggie	RDE	Beaumont	38	38	37 (-1)	37 (-1)	38 (1)	38 (1)	0	0
Abreu, Bobby	SLB	Beaumont	49	49	45 (-4)	45 (-4)	49 (4)	49 (4)	0	0
Ward, Phil	SLB	Beaumont	24	39	21 (-3)	39 (0)	24 (3)	39 (0)	0	0
Adams, Harry	MLB	Beaumont	55	55	57 (2)	57 (2)	56 (-1)	56 (-1)	1	1
Albertson, Todd	MLB	Beaumont	43	56	42 (-1)	56 (0)	43 (1)	56 (0)	0	0
Erbaugh, Duane	MLB	Beaumont	53	53	53 (0)	53 (0)	53 (0)	53 (0)	0	0
Dugan, Deon	WLB	Beaumont	39	39	36 (-3)	36 (-3)	37 (1)	37 (1)	-2	-2
Feugill, Terrance	WLB	Beaumont	34	34	33 (-1)	33 (-1)	34 (1)	34 (1)	0	0
Greene, Randall	WLB	Beaumont	38	38	34 (-4)	34 (-4)	36 (2)	36 (2)	-2	-2
Farley, Ricky	LCB	Beaumont	40	40	41 (1)	41 (1)	41 (0)	41 (0)	1	1
Scarlett, Randall	LCB	Beaumont	44	53	40 (-4)	53 (0)	41 (1)	53 (0)	-3	0
Morris, Troy	RCB	Beaumont	40	40	36 (-4)	36 (-4)	37 (1)	37 (1)	-3	-3
Sieber, Nolan	RCB	Beaumont	42	42	43 (1)	43 (1)	42 (-1)	42 (-1)	0	0
Fells, Luther	SS	Beaumont	41	41	36 (-5)	36 (-5)	40 (4)	40 (4)	-1	-1
Gooden, Everett	SS	Beaumont	57	57	58 (1)	58 (1)	58 (0)	58 (0)	1	1
Greenwood, Kris	SS	Beaumont	36	36	39 (3)	39 (3)	37 (-2)	37 (-2)	1	1
Craig, Alonzo	FS	Beaumont	39	39	39 (0)	39 (0)	39 (0)	39 (0)	0	0
Hawkins, Larry	FS	Beaumont	43	43	42 (-1)	42 (-1)	43 (1)	43 (1)	0	0

[QuikSand]

Geau Beaumont!

[Ben E Lou]

Cuervo: "I will overlay pre-interview and post-interview screen shots, track noscout/oldscout/newscout, cut test like crazy.

BUT I REFUSE TO USE A BATCH FILE, DADGUMMIT!!! "

[MIJB#19]

Ah yes, I thought about doing the expert scout routine if I'll ever end up sitting outside a top half finances situation. In a way I've already added it to my routine, but I don't even remember ever having a scout with a single 'poor'. Anyway, usually I try to project which position groups I will be more heavily scouting in the next off-season or two (especially if I expect to draft players) and base my scout needs on that. Additionally, if I'm loaded at another position group, I don't mind downgrading in that area at all.

I think I've done the same with coordinators, focussing on guys with excellent or at least 'VG' in a young position groups and willing settle for 'average' or maybe even 'fair' for the largely developed groups.

Down the road, it can even help your cash flow a bit, by going with a cheap bid to a two or three position groups expert, theoretically leaving you some extra money in seasons down the road to spend on a new scout or coordinator.

[TheMeat]

Just a quick thought: I agree that scouting error seems to be minor but I have witnessed some players being out 10 pts or so, which i would say is major. I notice this because I run a MP league with just me and one close friend of mine and on occasion he has me switch control to his team on my comp and make changes to his team to save from having to export/import. We do this because we usually just chat on msn and run through chunks of the season in a couple hours so we try to be quick. Anyways, i guess we could easily start up a new MP league, and do some testing on certain players from the perspective of each team, with varying scouts. I suspect that this would just affirm the assumption that good scouts are just closer to the actual rating as opposed to whatever masking is in effect.

This is pretty well a non-factor in my league as the two of us have very good scouts due to lack of competition and I believe mine is at least VG in all categories. But several players are a point or two different if viewed from the perspective of his scout, and in earlier years when we had scouts with some bad ratings there was much more of a difference on certain players.

[Firefly]

I haven't read the whole thread, but I suppose no one has considered the fact that scouts basically suck... all of them.

[OldSchool]

Originally Posted by cuervo72

Ok, when I first figured this, I guessed the following:

EX = 100% of true rating
VG = 97% of true rating
G = 94% of true rating
A = 91% of true rating
F = 88% of true rating
P = 85% of true rating (also "no scout")

Just ran into a player that fits this pattern while assessing a possible trade in a multiplayer league. A good scout was showing a WR as a 54, and a fair scout had him at 49. If the true rating was 57 then possible the second scout was a week fair, or maybe the true rating is closer to 56 and the first scout is a high-end good.