FOF7 Scouting Thread

[Ben E Lou]

Scout variance on younger players can be significant. How significant? I just created a test MP league and checked one young player with all 32 teams. Current was virtually the same for all scouts, but the scouting staffs differed greatly on future potential. The variance on future on this player was as high as 23 points. That's meaningful, and of course it's possible that it can be greater than that. This was just one player.

HOUSTON VIEW: 38/78
NYG VIEW: 36/55





On this particular player, once he's on the team, I'm not sure that 23 points makes a huge difference in decision-making. For most squads, he's a likely starter that you're going to want to work into the lineup rather quickly. However, I'm picturing a guy who is really, say, 25/51. Some scouts might see 24/40 while others might see 26/62. That's huge. 24/40 might leave him inactive, while 26/62 starts him in every game of the preseason and signs a mentor to help bring him along. And taking the 24/40 path even further, I could easily see a scenario where he gets little/no playing time, eventually gets cut as a third or fourth year player, and gets picked up and stuck in the starting lineup by someone who scouts him as a 55 future.



FWIW, the two teams farthest on each end of the spectrum for this player both had 3 to 4 staff members with pretty bad scouting. I didn't write down the scout view for every team, but it appeared that the mean and the mode would be somewhere about halfway between the two extremes, and that if graphed, the different scout views of the guy would be basically a bell curve distribution. There were maybe 3 to 5 outliers on either end that had him in the 73-78ish and 55-60ish ranges

[zbuckley]

I'm curious to see how this is going to play out with MP drafts. If combine accuracy was set at 100 it would be far easier to ignore scouting movement. I'm hoping wolf will really dig into the accuracy error and give us an idea how far things can be off. In the little SP i've done i've noticed more random volatility hits (small sample size). Initially I thought static bars were not longer static but I had quite a few booms and busts.

[QuikSand]

Good heavens... is there some chance that we might actually have to look at... gulp... results? Stats?

My land.

[Ben E Lou]

Heh. Remains to be seen. If the potential delta is that high on at least some players rated in the 40-60 range, then absolutely. I'm going to try to take a look at a player or two like that maybe tomorrow morning.

[BowTieSports]

Definitely a nice development if we have to use player performance (stats) and maybe a little trial and error with the depth chart, rather than just being able to plug and play a depth chart based 100% on scout ratings.

[corbes]

Originally Posted by Ben E Lou

Scout variance on younger players can be significant. How significant? I just created a test MP league and checked one young player with all 32 teams. Current was virtually the same for all scouts, but the scouting staffs differed greatly on future potential. The variance on future on this player was as high as 23 points. That's meaningful, and of course it's possible that it can be greater than that. This was just one player.

HOUSTON VIEW: 38/78
NYG VIEW: 36/55

What did the generic scouts have to say about this guy?

[Ben E Lou]

Originally Posted by Ben E Lou

If the potential delta is that high on at least some players rated in the 40-60 range, then absolutely.

I just ran across a rookie in the WOOF dispersal draft who fits this description. My staff is viewing this guy with 57 potential. The export from the league scout has him in the mid 30s in potential. If my guys are right, this guy could be an above average starter, but at least one scouting view has him as probably not worth a roster spot. One thing I've noticed about these players who have wide deltas that passes my "realism sniff test" is that every one of them is low in his current rating and positional experience. In other words, it appears that the most raw players are the ones seen with the most widely varying potential.

[Antmeister]

Originally Posted by Ben E Lou

I just ran across a rookie in the WOOF dispersal draft who fits this description. My staff is viewing this guy with 57 potential. The export from the league scout has him in the mid 30s in potential. If my guys are right, this guy could be an above average starter, but at least one scouting view has him as probably not worth a roster spot. One thing I've noticed about these players who have wide deltas that passes my "realism sniff test" is that every one of them is low in his current rating and positional experience. In other words, it appears that the most raw players are the ones seen with the most widely varying potential.

Ooooh.....which player is this? I gotta know and see how I view him?

[Ben E Lou]

Originally Posted by Antmeister

Ooooh.....which player is this? I gotta know and see how I view him?

[Ben E Lou]

Dola:

I suppose you will find out when I make my 25th round pick (after all of my starters and punter and kicker are taken.)

[TRO]

Originally Posted by Ben E Lou

Dola:

I suppose you will find out when I make my 25th round pick (after all of my starters and punter and kicker are taken.)

So...he's a long snapper then

[Antmeister]

Originally Posted by Ben E Lou

LOL....was hoping you would slip on one.

[Ben E Lou]

Originally Posted by TRO

So...he's a long snapper then

Fine. Round 26 then.

[Nataris]

I truly dislike not having actual scouts. I don't like that it adds another element to head coaches etc. In my head coach, I'm already looking for player development, young talent, motivation etc.. Yet I find myself going for the guys with more scouting since that is crucially important to evaluating talent.

Scouting is far too important of a stat to be hamstrung and attached to head coaches and coordinators.

I wouldn't mind if they all had a scouting stat, but there should AT LEAST a scouting director that allows you to be better at scouting certain positional talent.

In fact, it might just be the thing that I dislike the most. I want to treat scouting as its own important element. Not tact on to head coaches and coordinators.

IMO, the Assistant Coach is redundant and should have been a scouting director or the like.

Just 1 mans humble opinion though..

[Ben E Lou]

Originally Posted by Ben E Lou

Two behavioral modifications based on the above:

1. Using Extractor for ratings for a MP league is probably a bad idea now. The game's export spits out a generic scout view.
2. MP owners might want to be a lot more cautious in discussing the ratings that they see.

Originally Posted by Ben E Lou

Heh. Remains to be seen. If the potential delta is that high on at least some players rated in the 40-60 range, then absolutely. I'm going to try to take a look at a player or two like that maybe tomorrow morning.

Humpty bumpty. The comments above have moved from "theoretical" to "very relevant" for me.

The GML just completed draft entry and we're doing FA2:1 on Monday. My staff sees a guy who was your standard 25ish/45ish ratings-dropper as 46/58 in the new game, with a 55/79 in one of the key bars for his position. We had the #1 spot in the staff draft and therefore got guys who were solid or better at everything, including scouting. It'll be interesting to see if others offer him, but for me that's a move from "guy I wouldn't bother with signing even as a backup" to "he starts for me if I sign him."

[Ben E Lou]

Bump. Breaking the scouting discussion off from the general "staff" thread, as I think this deserves its own targeted topic.

A little more info on this. I just posted this at the CCFL:

Originally Posted by zullo

damn Ben, you had 5 youngsters all go up 20+ points in future rating and 3 of them were undrafted. You are still the master. What tells are you seeing that we are not that 3 of these guys looked crappy enough to not even be drafted?

I had a nice post tc 20+ bump on my 1st rd pick, but he had a sick combine and low bars and popped like i had hoped. But everyone else i got was pretty mediocre and despite decent combines all went down

MY RESPONSE:

You have to keep in mind that our league scout sucks at scouting, and I have one of the best position-player scouting staffs in the league. My first "+20" on the list is this guy:


My team's scouting staff has him as a much more modest 30/42-->34/44. If I'm reading the tea leaves correctly, it's like this:

CRAPPY SCOUT, BEFORE SEEING HIM IN PRESEASON GAMES: "HE SUCKS!"
CRAPPY SCOUT, AFTER SEEING HIM IN PRESEASON GAMES: "ZOMFG!!! I WAS WRONG!!!! HE'S GONNA BE YOUR STARTER IN A YEAR OR SO!!!!!1"

VERY GOOD SCOUTS, BEFORE SEEING HIM IN PRESEASON GAMES: "He could be a decent backup RB."
VERY GOOD SCOUTS, BEFORE SEEING HIM IN PRESEASON GAMES: "He could be a slightly better decent backup RB than we thought."

And then, my guess, 2ish seasons from now when he's all red and all scouts see him as basically the same...

CRAPPY SCOUT: "He's a nice backup at 46/46, but not someone you want as your every-down back."
VERY GOOD SCOUTS: "He's a nice backup at 46/46, but not someone you want as your every-down back."

My own scouts didn't have any players improving +20. The best was my punter. My scouts are showing 61/61-->76/76 on him. League scout? 70/70-->71/71. And how about one guess at what my scouting staff sucks at. Yup, you guessed it. My Ass't Co-atch is rated 24 in scouting. The next best change by my scouts is a +11 for my 3rd-round WR. You know, that guy near the top that the league scout says is...


Same deal, my scouting staff is much better than the generic, so they saw him as significantly better than 13/13 to begin with.

I'm sharing these particular ratings because there's zero chance that these guys are getting cut or traded. I'm not sure what I have on my hands in these players and I want to see it play out. By the time they hit FA, all of our scouts will basically agree on who they are. But from looking at just these two cases, it becomes extremely obvious why JG is strongly recommending not sharing what your scouts see.

[Ben E Lou]

Oh...probably worth highlighting for those who don't pay attention to the Help File. This was added in the latest version:

Originally Posted by The 7.0b version of the Help File

Since personal scouting data can vary quite a bit by team, this data should not be shared with other owners in your league. This data should be kept private. Having it in .csv form may be useful when you need to make decisions away from the game.

[Ben E Lou]

Oh, and when you export that data, you get this message:

[Ben E Lou]

OK. Now THIS is worth noting, also from the CCFL. Check out this undrafted FA, first, according to the league scout:

Atlanta Falcons

FB

Kendall Irizarry

1

35

35

43

57

+8

+22

0(0)

So the league scout is showing a big bump. However, my scouts had him at 36/54 pre-camp, 42/63 now.

And when I look at the money he was offered to sign and compare that to scouting ratings, I get this breakdown.

GBY offered 1.5M and average of HC and OC scout ratings is 80.
ATL offered 2.5M and average of HC and OC scout ratings is 76.
CLE offered 1.85M and average of HC and OC scout ratings is 61.

but then...

ARI--1.10M, scout ratings 43
CHI--1.15M, scout ratings 38.5
PIT--1.11M, scout ratings 46

(1.10M is the rookie minimum in the CCFL.)

From that, I can't help but guess that the ATL, CLE, and GBY staffs saw this guy as a potential solid starter, while the other three (and perhaps the rest of the league...31 owners submitted exports for FA2, so it's not like it was owner inactivity) had him with much lower potential, like the league scout.

(Oh, and all of the offers above were for one year.)

[QuikSand]

I really hope this pans out. If your investment in scouting really matters in this game, that could be a pretty interesting new wrinkle.

Of course, the method of acquiring said investment in scouting will, likely, then prove to be all the more frustrating. But for now, I'll focus on the upside.

[Yoda]

I still think that letting the best teams have the best coaches unbalances the game, since scouting means so much more now.

[Ben E Lou]

Originally Posted by Yoda

I still think that letting the best teams have the best coaches unbalances the game, since scouting means so much more now.

The best teams aren't getting the best coaches. With the way staff retention works, that's just not possible.

[AlexB]

I always thought the scout report was based on an average somehow of the whole staff.

But am I right in reading from the above that the scouted ratings are from the HC all the time, plus only the OC for offense, the DC for defence and the AC for special teams?

[Ben E Lou]

Originally Posted by AlexB

I always thought the scout report was based on an average somehow of the whole staff.

But am I right in reading from the above that the scouted ratings are from the HC all the time, plus only the OC for offense, the DC for defence and the AC for special teams?

Originally Posted by Help File

Your coaches and coordinators, with the exception of the strength coordinator, all develop players and scout players. The offensive coordinator plays a bigger role with offensive players. Your defensive coordinator plays a bigger role with defensive players and your assistant coach plays a bigger role with special teams players. When developing talent, having a head coach and/or an assistant coach that matches the position group of the player can often give the player a development boost.

So the DC and AC might play a small role in scouting this offensive player, but based on the help file entry and the small amount of looking into this that I've done, it appears that the HC+Coordinator combo is what's most relevant for position players, and HC+AC for special teams. I tend to suspect that the HC is the lesser factor in the equation than the Coordinator. So my best guess at this point is that it's...

OFFENSIVE PLAYERS
1. Offensive Coordinator: primary
2. Head Coach: less than OC, but significant
3. Assistant Coach: trivial/nothing
4. Defensive Coordinator: trivial/nothing

DEFENSIVE PLAYERS
1. Defensive Coordinator: primary
2. Head Coach: less than DC, but significant
3. Assistant Coach: trivial/nothing
4. Offfensive Coordinator: trivial/nothing

SPECIAL TEAMS
1. Assistant Coach: primary
2. Head Coach: less than AC, but significant
3. Offfensive Coordinator: trivial/nothing
4. Defensive Coordinator: trivial/nothing

[AlexB]

Gotcha. Thanks for that, I had picked up on the development of players, but for some reason I can't fathom hadn't connected it through to scouted ratings as well. Will change the way I pick up staff for sure.

[QuikSand]

Originally Posted by Ben E Lou

The best teams aren't getting the best coaches. With the way staff retention works, that's just not possible.

Well... people who pay attention, figure stuff out, and tend to the details have an advantage in every portion of this game.

In theory, the staff advantage won't go to the "best teams" due to financial reasons. If that theory ever really manifests, then I think I'd agree with your assessment.

But alas, here we are.

[Ben E Lou]

Originally Posted by QuikSand

Well... people who pay attention, figure stuff out, and tend to the details have an advantage in every portion of this game.

In theory, the staff advantage won't go to the "best teams" due to financial reasons. If that theory ever really manifests, then I think I'd agree with your assessment.

But alas, here we are.

Oh, sure. I'm pretty sure I made a comment about this in the other thread, but it's worth repeating: no matter how slow or fast a MP league moves, there is only one opportunity to retain staff members. The biggest advantage in all of this will go to the people who rarely/never miss that stage, no matter how good their teams are. There are a decent number of people who are consistently mediocre who'll get/keep good staff simply because they don't miss many stages. And MalcPow will continue to out-draft all of us and then send his coaches our way.

[MalcPow]

*fist pump*

Hey, wait a second...

[AlexB]

Following on from the OC/DC/AC scouting effect, is there any evidence that the league scout accuracy is affected by his Primary Focus position?

[JohnRckr]

Originally Posted by Ben E Lou

Oh, and when you export that data, you get this message:

Is it possible for a non-developer type to get player name information into the personal scouting data csv? With only player_id the file doesn't provide useful information until a merge of player Name happens. Any help is appreciated.

[albionmoonlight]

Originally Posted by QuikSand

Good heavens... is there some chance that we might actually have to look at... gulp... results? Stats?

My land.

In IHOF right now, I am torn between my starting QB, who my scout says is decent if not great. And my mid-round rookie QB who has spot started this year based on injury and has put up better numbers, but who my scout says looks like a mid-round rookie QB should look.

I am assuming that small sample size has more to do with the rookie's numbers than anything. But there's a small part of me that wonders if I should be paying more attention to how these guys actually play.

Basically, if you take the rookie part out of it, this is a Doug Flutie/Rob Johnson situation here. And I admit to being somewhat swayed in my thinking by the general mantra of "when possible, do the opposite of what Wade Phillips did."

[gstelmack]

Originally Posted by JohnRckr

Is it possible for a non-developer type to get player name information into the personal scouting data csv? With only player_id the file doesn't provide useful information until a merge of player Name happens. Any help is appreciated.

This is all generally intended for import into a database. You could possibly use something like Access here. Or load both CSVs into Excel and create a third sheet that handles the merge.

What is the specific use-case? If there is a need for a utility that I would find useful, I might do something.

[Ben E Lou]

Before I could write any code whatsoever, I just used Excel and VLOOKUP to view Interrogator data when I wanted to do chemistry searches.

[JohnRckr]

Originally Posted by gstelmack

This is all generally intended for import into a database. You could possibly use something like Access here. Or load both CSVs into Excel and create a third sheet that handles the merge.

What is the specific use-case? If there is a need for a utility that I would find useful, I might do something.

My thought was to find a way to use my staff scouting data (player_personal) in an Excel spreadsheet with last name, first name, DOB, position, height, weight, and current team etc. Same premise of 'Recommend Player' in-game but, in Excel I could compound my sort criteria. For example, if I wanted to sort within cornerbacks highest rated at BnR and M2M.

[gstelmack]

You should be able to do that inside Excel.

[w24olfpack]

Originally Posted by Ben E Lou

Before I could write any code whatsoever, I just used Excel and VLOOKUP to view Interrogator data when I wanted to do chemistry searches.

Same. I have a big lookup table with the BDAY (MMDD format) with their sign, their conflict and their two affinities.

0101	Capricorn	Leo	Aquarius	Libra
0102	Capricorn	Leo	Aquarius	Libra
0103	Capricorn	Leo	Aquarius	Libra
0104	Capricorn	Leo	Aquarius	Libra
0105	Capricorn	Leo	Aquarius	Libra
0106	Capricorn	Leo	Aquarius	Libra
0107	Capricorn	Leo	Aquarius	Libra
0108	Capricorn	Leo	Aquarius	Libra
0109	Capricorn	Leo	Aquarius	Libra
0110	Capricorn	Leo	Aquarius	Libra
0111	Capricorn	Leo	Aquarius	Libra
0112	Capricorn	Leo	Aquarius	Libra
0113	Capricorn	Leo	Aquarius	Libra
0114	Capricorn	Leo	Aquarius	Libra
0115	Capricorn	Leo	Aquarius	Libra
0116	Capricorn	Leo	Aquarius	Libra
0117	Capricorn	Leo	Aquarius	Libra
0118	Capricorn	Leo	Aquarius	Libra
0119	Capricorn	Leo	Aquarius	Libra
0120	Capricorn	Leo	Aquarius	Libra
0121	Aquarius	Aries	Libra	Capricorn

[w24olfpack]

Originally Posted by Ben E Lou

Before I could write any code whatsoever, I just used Excel and VLOOKUP to view Interrogator data when I wanted to do chemistry searches.

Same. I have a big lookup table with the BDAY (MMDD format) with their sign, their conflict and their two affinities.

Then do a vlookup comparing the leader to the player in an if statement.

0101	Capricorn	Leo	Aquarius	Libra
0102	Capricorn	Leo	Aquarius	Libra
0103	Capricorn	Leo	Aquarius	Libra
0104	Capricorn	Leo	Aquarius	Libra
0105	Capricorn	Leo	Aquarius	Libra
0106	Capricorn	Leo	Aquarius	Libra
0107	Capricorn	Leo	Aquarius	Libra
0108	Capricorn	Leo	Aquarius	Libra
0109	Capricorn	Leo	Aquarius	Libra
0110	Capricorn	Leo	Aquarius	Libra
0111	Capricorn	Leo	Aquarius	Libra
0112	Capricorn	Leo	Aquarius	Libra
0113	Capricorn	Leo	Aquarius	Libra
0114	Capricorn	Leo	Aquarius	Libra
0115	Capricorn	Leo	Aquarius	Libra
0116	Capricorn	Leo	Aquarius	Libra
0117	Capricorn	Leo	Aquarius	Libra
0118	Capricorn	Leo	Aquarius	Libra
0119	Capricorn	Leo	Aquarius	Libra
0120	Capricorn	Leo	Aquarius	Libra
0121	Aquarius	Aries	Libra	Capricorn

[Ben E Lou]

Heh. Just posted this at WOOF regarding the changes we say during this preseason:

A note on change tracking before I post it: there's not enough data yet to draw a definitive conclusion, but I'm most definitely seeing a pattern here: the worse the scouting rating of the league scout, the bigger the changes he reports. I'll talk about the two guys I have (one on each list) as great examples:

Safety T.J. Jefferson is showing 23/51-->26/29 according to the league scout: a big, fat minus 22. My scouting staff? 21/44-->25/40.
RB Garrett Irwin, according to the crap league scout, is showing 26/27-->26/50. My staff tells me that he's 26/31-->27/36.

And keep in mind that both of those guys might change directions next year.

So take what you're about to see with that big disclaimer in mind.

For those interested in seeing the big ol' incorrect changes, here ya go: 2014 Preseason Weeks 1-3 Complete

[Ben E Lou]

Ultimately, I think the big takeaway here is this: "if you don't have at least a decent scouting staff, do not trust what they're telling you about younger players."

[Antmeister]

Yep....came to this conclusion a while back in my single player campaign (Front Office Football Central - View Single Post - Thoughts on Ratings History). Basically it seems regardless of how bad the scout is, he doesn't have too much variance from a good to great scout for the veteran players (4+ years).

[Antmeister]

With that said, if you see a huge bump in current, though, that may be a good sign (as long as there is a bit of movement upward for future). It is the future that is wobbly for a young player. Heck even for a good to great scouting it can go up and down until it settles later in the career.

[corbes]

My working theory of scouting, posted here for the purpose of inviting review and criticism:

1. Each player gets a development curve which guides their development in the absence of other random variables (volatility, injuries, etc.)

2. Every player’s scouting becomes more accurate as the player advances further along the development curve.

3. Scouting involves both accuracy and precision.

4. A good scout has better precision than a bad scout.

5. So even if a bad scout’s ratings might generally get more accurate over time, beware the low precision of the evaluations.

6. My *suspicion* is that good scouts also get an accuracy bonus (in other words, a good scout may see with an accuracy that is one step ahead of where the player actually is on the development curve, or something like that). I don’t know exactly how this works in code, but this observation is meant to address case examples in the nature of Kendall Irizarry (example posted above in this thread).

[corbes]

A corollary to #2 would be that during the draft, the player's development % tells you something about the level of accuracy you can expect in the bars you are seeing (but it doesn't tell you anything about the precision of those bars).

[Sef0r]

There are not too many "Draft - what did you get" threads going on anymore because it it very hard to tell what you got. Gone are the days of the telling bar/combine combinations, at least from what I've seen.

If I had to review my drafts now it boils down to this statement...

"I am pretty sure I got some good players in all 7 rounds, but I'm also pretty sure they suck big balls".

[Pacersfan46]

Okay .... I didn't know FOF7 existed until this weekend. Imagine my excitement when I was scanning around and decided to see what was going on here after a year or two.

All right, I'm still in the experimenting stage of trying the system to see what works and how to scout but can anyone explain to me why this guy outside of 2 years has been a run of the mill QB? These are his ratings in year 13, at his peak I believe he was a 79 or 80 overall. I'm Tampa Bay and he was my first big draft pick. I'm confused as to why he's never played outstanding in his career. After his ridiculous rookie year I was expecting greatness.

When he was on my team I rotated through a couple offensive style types and the result was always the same. I thought maybe I was doing something wrong, so I traded him to see what the CPU would do with him, and it was much of the same. Anyway, any idea's?

I'm posting this here because if there's an issue that can be seen on these screens, then it becomes a scouting issue to look for.

[Sef0r]

Eli Manning?

[Pacersfan46]

Actually, going through my games current QB's I've noticed a trend. QB's with the "Long Passes" style don't perform up to their ratings as well as roll out or short passes QB's.

In fact, the guys currently playing in my game (can't see retired guys style) even roll out style QB's in the 40's overall have similar careers as the 2 highly rated long passers.

May be abnormal, and it's an extremely small sample size but something I think to keep an eye on.

[mrtourette]

Could it be that players aren't utilising QBs with a long passing style properly, and the results of doing so are greater than incorectly utilising a short/rollout passer? Or that it's easier to correctly utilise a short/rollout passer?

[Yoda]

Unless something has changed- I believe that the QBs style of play just affected the formations he'd learn, or had a better chance to learn.

[Mike7273]

How much has QB Style mattered in the past?