A Player Development Tell

[MalcPow]

I wanted to share with the community a fairly recent discovery of mine that points to a strong correlation between player development and the player's initial fan popularity. I'll outline some of my more specific findings as well, but I'll note that it is a significant discovery in light of the fact that the Interrogator program in the StelmackSuite software actually reveals this information before it might otherwise be available in game. Recognizing that advantage, it's best that information about this "tell" be publicized in a way that levels the playing field for multi-player league participants.

Specifically, it appears that a player that has a fan popularity rating >40 as a member of the draft class is a player of at least backup quality. It also appears that a player with a fan pop > 80 is a good starter, and pop >90 seems to be a strong indicator of an elite player. Many of these players have other signs (such as good bars and combines) that give some hint of their talent, but some do not. Particularly in the later rounds, this information can be used to get much more certainty and value from a large pool of very similar players.

I'll open up for discussion here, and try to clarify in response to any questions people have. I do not have vast quantities of data to back all of this up, but I have seen enough that I find it to be compelling enough to reveal. I will say that low popularity does not seem to mean a player sucks, only that popularity >40 seems to be a surefire indicator that you have a solid player, regardless of any other data points.

For reference, the average draft class has approximately 50 players with fan pop >40.

[cuervo72]

Originally Posted by MalcPow

I wanted to share with the community a fairly recent discovery of mine that points to a strong correlation between player development and the player's initial fan popularity. I'll outline some of my more specific findings as well, but I'll note that it is a significant discovery in light of the fact that the Interrogator program in the StelmackSuite software actually reveals this information before it might otherwise be available in game. Recognizing that advantage, it's best that information about this "tell" be publicized in a way that levels the playing field for multi-player league participants.

I wasn't aware that there were specific values associated with this "tell", but I have heard it kicked around that high fan popularity could generally indicate that a player was good (as few have really really high fan pop).

I was also aware that Interrogator could extract this information, as I stumbled upon it in scripting the FOFL site. Specifically, the team chemistry/personality screen:

http://www.thefofl.com/teampages/personality.php?t=6

I found that if I plugged in the team id for draftees, the personality ratings showed up, even if they didn't in-game. I pretty quickly figured that this wasn't the best thing to have out there, and put an if/then statement in to prevent that team id from revealing anything. This may not be an issue just for fan popularity, but also for intelligence if you subscribe to RKG's theory that Solecismic score is derived from both intelligence and the hidden avoid INT rating (where high Sol + dumb QB = WIN).


The other thing to be noted here is that when these players enter the FA pool undrafted, their personality traits will appear on the Personality View roster page. I don't think they appear on the player pages yet, though I could be wrong. Either way, it seems to me that such players are immediately targeted by some of the shrewder owners out there, assuming they make it through the draft.

[Ben E Lou]

On the one hand, I'm glad that MalcPow decided to go public with this. On the other, I have some concern that it might lead to "witch hunts" in MP leagues: "Oh no! So-and-so drafted 5 guys with higher pop, so he must be looking at it ahead of time!" This point is probably worth emphasizing:

Originally Posted by MalcPow

Many of these players have other signs (such as good bars and combines) that give some hint of their talent, but some do not.

The vast majority of these guys can be identified as good players by the many other tells that are available from the front end of FOF. For example, 9 of the 50 players who fit the identified profile in the most recent WOOF draft went to the two teams owned by QuikSand and Celeval. This should not exactly surprise anyone.

Originally Posted by cuervo72

Either way, it seems to me that such players are immediately targeted by some of the shrewder owners out there, assuming they make it through the draft.

Well sure. That's information available right there in FOF. You should have seen the crazy bidding wars that MalcPow and I had over a small handful of guys in the most recent ECFL season during FA2.

[Eaglesfan27]

This is disappointing, but I really appreciate MalcPow going public with this.

[RedKingGold]

The bigger question is, does it work in the opposite direction (i.e. guy with studly bars but low popularity is a fake?)

[Ben E Lou]

No.

[MalcPow]

Originally Posted by RedKingGold

The bigger question is, does it work in the opposite direction (i.e. guy with studly bars but low popularity is a fake?)

It's not really clear. I wish I knew. I certainly view a guy like that with more skepticism than I did previously. It's important to note, popularity can change a lot over the course of a career, and I am speaking specifically to a player's initial popularity. Some guys lose popularity very rapidly, even if they're starting games. Other guys gain it rapidly, there's a fairly standard +3 to popularity every time a player wins POG in fact, so QBs shoot up quickly.

[VPI97]

From a design point of view, it makes some sense....if a player is popular/well known coming into the draft, then it's probably because he was good in college. If a draftee was good enough in college to be popular, then there's a better than average chance he'll be a solid pro.

What doesn't make sense is if that's the design decision for initial popularity values, then why didn't Jim expose that number in the draft report.

From a MP point of view, I would expect all leagues to have a rule that prohibits using data that's not exposed by the game. Obviously, the pre-draft popularity falls under that category. Enforcing the rule can be sticky, but I know of at least one league that dismissed someone for an almost eerily similar situation where proof was based primarily on speculation and supposition.


Shurg...either way it's just another gutshot to the rotting corpse of FOF MP.

[Ben E Lou]

MP, I make that comment based on a couple of different data points:

1. I've seen massive creeping from guys with low popularity. For example, ILB Jared Caniff in WOOF is a +35 creeper with only 15 pop.

2. I've seen players who are created using FOF's players.csv file who I *know* are good players but have lower pop.

[Eaglesfan27]

Originally Posted by VPI97

Shurg...either way it's just another gutshot to the rotting corpse of FOF MP.

Agreed. The more I think about this, the more disappointing it becomes to me.

[albionmoonlight]

I trust the people with whom I am in a MP FOF league. But, geeze. A commonly available utility allows people to see something that is important but not in the game. And, if they use this information with any sense, it will be hard to catch them because other, legal, tells will make their choices seem legit.

I mean, I trust the people in my MP FOF league. But that doesn't mean that I'm ready to give them my ATM card and PIN number either, you know?

[MalcPow]

Well an interesting wrinkle to things is the AI's refusal to trade players that are "fan favorites" in SP games. Even if those fan favorites appear to be fairly average players, you can't wrest them from AI teams no matter what. Having built some pretty 'popular' teams trying to test this, I can attest that the impact on ticket sales is minimal at best from what I've seen. On some level, popularity may serve as a kind of cover story for the AI protecting top players.

[gstelmack]

Originally Posted by albionmoonlight

A commonly available utility allows people to see something that is important but not in the game.

And I remain embarrased at this. For most players, this data is available. And once the draft is complete, it's available for everyone (thus, without Interrogator you could use this along with the now-exposed future to make more intelligent decisions about who to go after in Late FA). But it's hidden for the draft class, and thus should be hidden in Interrogator (much like I did for staff data in 2k4, and still do for HoF votes in 2k7). I've tried hard to only expose data that is also exposed by the game, and I missed this.

[cuervo72]

Well in fairness Greg, things like this were missed in the game before, weren't they? At one point couldn't you get accurate rating information from the Roster Attributes screen before the players had actually gone through camp?

[Ben E Lou]

I don't know. I think it's important to note that a good majority of these guys are getting drafted in the first few rounds just from the front-end information that's out there. It's probably more significant during FA2 when the bidding free-for-all takes place. I can't help but wonder if the solution is to publish a "Fifty Fan Favorites" list pre-draft, and level the playing field a bit. It's not like they're all studs. A "1" player (1 - Replacement-Level Starter (decent backup) - 23.9%) can fit into this group, so it's not like that's a guy I want to be drafting in the 1st or 2nd round. And like I said above, there are plenty of examples of quite solid players who come out of the draft with low pop.

[albionmoonlight]

Originally Posted by gstelmack

And I remain embarrased at this. For most players, this data is available. And once the draft is complete, it's available for everyone (thus, without Interrogator you could use this along with the now-exposed future to make more intelligent decisions about who to go after in Late FA). But it's hidden for the draft class, and thus should be hidden in Interrogator (much like I did for staff data in 2k4, and still do for HoF votes in 2k7). I've tried hard to only expose data that is also exposed by the game, and I missed this.

You have a great tool. Don't apologize. It's really like you have it right and the game has it wrong. (Which, I know, isn't the point, but still . . .)

[Dutch]

Fan Popularity just became the "TCY/Intelligence" bug.

Thanks for posting.

[Subby]

Off topic comment to Greg....

Why not expose HoF votes? That would be pretty sweet to see and I am not sure that gives any advantage.

[gstelmack]

Originally Posted by Subby

Off topic comment to Greg....

Why not expose HoF votes? That would be pretty sweet to see and I am not sure that gives any advantage.

'cause I try not to expose info the game doesn't expose . Once the guy enters the hall of fame, it's exposed.

[Kodos]

Ugh.

[Subby]

Originally Posted by VPI97

From a MP point of view, I would expect all leagues to have a rule that prohibits using data that's not exposed by the game. Obviously, the pre-draft popularity falls under that category.

Didn't have any specific language about this in the FOFL Constitution but added it earlier today...

Enforcement is going to be tough considering most of these guys get drafted anyway and have a lot of other things in common. An owner that drafts five popularity guys per season isn't necessarily breaking the rules.

[Ben E Lou]

Originally Posted by Subby

Enforcement is going to be tough considering most of these guys get drafted anyway and have a lot of other things in common. An owner that drafts five popularity guys per season isn't necessarily breaking the rules.

I'm really chewing on this for WOOF and about to put it up for league discussion. There's a big part of me that thinks an immersive "50 Fan Favorites From The Draft Class" list being published is a better solution than people being suspicious of one another when they see high pop players getting drafted, but I don't know what the league thinks.

[MalcPow]

I'll be posting a list of > 40 pop players when the draft class is released in my leagues from here on out (assuming commissioner's are cool with that). I don't know if that addresses all concerns, but it should open things up for people from a fairness perspective.

[primelord]

I think people are over reacting a bit to this. This isn't an end all tell or anything. You can't just make your way down the popularity list and get all studs. As has been said, in the top leagues the vast majority of these guys are drafted already as it is.

It's important that everyone know this kind of stuff so everyone is working from the same page, but this did not suddenly unlock the draft. It is not equivelant to a key to print out the real ratings for all the players.

[MalcPow]

Originally Posted by primelord

I think people are over reacting a bit to this. This isn't an end all tell or anything. You can't just make your way down the popularity list and get all studs. As has been said, in the top leagues the vast majority of these guys are drafted already as it is.

It's important that everyone know this kind of stuff so everyone is working from the same page, but this did not suddenly unlock the draft. It is not equivelant to a key to print out the real ratings for all the players.

This is pretty true. The advantage here is not so much in guaranteeing you get all studs, it's in value selections. You're not going to have a roster stocked with guys in the 90s by using this information, but you will accumulate solid players where you may have been drafting duds previously. This is not a silver bullet to success, but there is an advantage here.

[larrymcg421]

If anything, I think this might make the draft more interesting. I bet you'll see some strong combine guys fall later int he draft than they normally would because they have lower popularity. I'll gladly scoop these guys up while everyone else fights over the "safe" picks.

My late 1st round RT has 12 popularity and he is at 59/59 in his 3rd year.

[VPI97]

Originally Posted by MalcPow

I'll be posting a list of > 40 pop players when the draft class is released in my leagues from here on out (assuming commissioner's are cool with that). I don't know if that addresses all concerns, but it should open things up for people from a fairness perspective.

I wouldn't be cool with that. Breaking a rule and giving the results to everyone may be better than breaking a rule by yourself...but it's still breaking a rule.

[larrymcg421]

Originally Posted by VPI97

I wouldn't be cool with that. Breaking a rule and giving the results to everyone may be better than breaking a rule by yourself...but it's still breaking a rule.

I don't see how it's breaking a rule if the commissioner is cool with it.

Also, I think it's obvious that some people will use this cheat. It'd be naive to suggest otherwise. Since that's the case, the information should just be revealed to everyone so there is no unfair advantage.

[MalcPow]

Originally Posted by VPI97

I wouldn't be cool with that. Breaking a rule and giving the results to everyone may be better than breaking a rule by yourself...but it's still breaking a rule.

Noted, as you're the commissioner of a league I'm in. I tend to lean toward Larry's thinking, but I'll obviously respect your wishes.

[VPI97]

Originally Posted by larrymcg421

I don't see how it's breaking a rule if the commissioner is cool with it.

Commish doesn't make the rules in any leagues I'm in...he just enforces them. If the owners change the rule, then the commish changes his stance.

It would just seem hypocritical to ask someone to stop enforcing a rule after x amount of time...especially if violations of this rule have been dealt with differently in the past.

[cartman]

Those of you ready to declare FOF MP dead, because of this I'll just throw out that I have never used Interrogator, and I made it to the Conference championship game in WOOF last season.

[headtrauma]

Originally Posted by cartman

Those of you ready to declare FOF MP dead...

Yeah, it does seem to me that there's an inordinate amount of hand-wringing over this. I think we're all underestimating the role of randomness in this game (or just using this as an excuse to decry the oft-cited shortcomings of the game yet again).

[MalcPow]

Originally Posted by VPI97

Commish doesn't make the rules in any leagues I'm in...he just enforces them. If the owners change the rule, then the commish changes his stance.

It would just seem hypocritical to ask someone to stop enforcing a rule after x amount of time...especially if violations of this rule have been dealt with differently in the past.

Not to really linger with this, but I wouldn't interpret this as against the rules in the constitution of the GEFL, or any of the leagues I'm familiar with, unless it was deemed to be so as part of a broad "commissioner's discretion" mandate with regard to general conduct.

And I agree with those that are calling a lot of the responses here overblown. Again, this is not some instawin superstrat. It's a slight edge that can be exploited a tad too much with the (unintended) use of one of Greg's awesome utilities. I'm wary of having given this angle too much of a spotlight at this point. I think it is a datapoint worth noting, but this shouldn't revolutionize the way anyone plays (or doesn't play) the game. *shurg*

[RedKingGold]

Originally Posted by primelord

I think people are over reacting a bit to this. This isn't an end all tell or anything. You can't just make your way down the popularity list and get all studs. As has been said, in the top leagues the vast majority of these guys are drafted already as it is.

It's important that everyone know this kind of stuff so everyone is working from the same page, but this did not suddenly unlock the draft. It is not equivelant to a key to print out the real ratings for all the players.

+1

Is it a tell? Perhaps yes. Is it good for the community to know? Yes. Are we going to see super teams built off of Fan Favorite players? I doubt it.

[Ben E Lou]

Somewhat related to the point about exposing/not exposing is the fact that it's actually very odd where the personality stuff is and isn't exposed in the game. For example, in the most recent WOOF draft (completed yesterday), I didn't interview TE Dixon Grant, who ended up being my 6th round pick. He's now on my roster in FA2. Because he wasn't intereviewed, I can't see any of his personality information. However, I know from other in-game information that he's 80 leadership, 10 intelligence, 92 loyalty, 88 play for winner, 8 personality, and 13 popularity. I wonder if one or the other is an oversight.

[sovereignstar]

Redo. I've been drafting nerds.

[sovereignstar]

lol.. which team is mine and which team is MalcPow's?

[Antmeister]

Well I brought this up in WOOF, so I will post it here as well. How do you think this will affect the draft if a list is posted. From my understanding, it appears that people with a popularity of 80 or above, they are just about guaranteed a starter. So my concern is that when used in combination with combines and/or bars, this will drastically reduce 1st round busts since more people are only gonna look for that combination in the first round. I am not saying that anything under 80 is a bust, but who is gonna take that risk when they have a guarantee.

In one respect it will make the draft much more exciting because there will be those that take those risk on low popularity players. However, I doubt you are gonna see many people doing that in the 1st round and 1st round busts aren't gonna happen much anymore.

[primelord]

Originally Posted by Antmeister

Well I brought this up in WOOF, so I will post it here as well. How do you think this will affect the draft if a list is posted. From my understanding, it appears that people with a popularity of 80 or above, they are just about guaranteed a starter. So my concern is that when used in combination with combines and/or bars, this will drastically reduce 1st round busts since more people are only gonna look for that combination in the first round. I am not saying that anything under 80 is a bust, but who is gonna take that risk when they have a guarantee.

In one respect it will make the draft much more exciting because there will be those that take those risk on low popularity players. However, I doubt you are gonna see many people doing that in the 1st round and 1st round busts aren't gonna happen much anymore.

Ant,

I don't think it is that big of a deal. Without having really looked too deeply into this, I am willing to bet that most of those 80+ guys are guys that had combine numbers and what not good enough to go in the spots they will go in anyway.

It's also important to remember that a guy with a lower fan pop can still be a beast. So in other words seeing a guy with great combine scores and an 85 fan pop may make you feel very confident in your pick, but my guess is the vast majority of the time his combines alone would have made you feel very confident about that pick.

[Ben E Lou]

Originally Posted by Antmeister

Well I brought this up in WOOF, so I will post it here as well. How do you think this will affect the draft if a list is posted. From my understanding, it appears that people with a popularity of 80 or above, they are just about guaranteed a starter. So my concern is that when used in combination with combines and/or bars, this will drastically reduce 1st round busts since more people are only gonna look for that combination in the first round. I am not saying that anything under 80 is a bust, but who is gonna take that risk when they have a guarantee.

Well, the sheer numbers here won't allow that to happen. In the WOOF class that we just drafted, there are only 8 players with pop >80. I did 6 Interrogator runs on SP and MP leagues this evening, and there are a total of 35 >80 guys, so we're probably just talking 4-8 guys a season.

In one respect it will make the draft much more exciting because there will be those that take those risk on low popularity players. However, I doubt you are gonna see many people doing that in the 1st round and 1st round busts aren't gonna happen much anymore.

Even a 100 pop rookie can get hit by the Volatility Stick Of Death.

[Antmeister]

Originally Posted by primelord

Ant,

I don't think it is that big of a deal. Without having really looked too deeply into this, I am willing to bet that most of those 80+ guys are guys that had combine numbers and what not good enough to go in the spots they will go in anyway.

It's also important to remember that a guy with a lower fan pop can still be a beast. So in other words seeing a guy with great combine scores and an 85 fan pop may make you feel very confident in your pick, but my guess is the vast majority of the time his combines alone would have made you feel very confident about that pick.

Oh...don't get me wrong, I know there are under indicators. My concern is the term "guarantee". I am having this discussion on the other board, but does that also mean that regardless of volatility, the players with >80 popularity and good/great combines and/or bar not destined to be a bust.....ever. In other words does, volatility only become a factor if the player has <40 popularity.

The reason why I ask this is because of the verbage used in the first post. I am not sure how far out in the future he looked, past the draft, to determine that the players with >90 popularity will be elite.

[Celeval]

FWIW, it looks like there were 7 players with a popularity under 40 taken in the first round of the WOOF draft; and 2 with 90+.

[Celeval]

Originally Posted by Antmeister

Oh...don't get me wrong, I know there are under indicators. My concern is the term "guarantee". I am having this discussion on the other board, but does that also mean that regardless of volatility, the players with >80 popularity and good/great combines and/or bar not destined to be a bust.....ever. In other words does, volatility only become a factor if the player has <40 popularity.

The reason why I ask this is because of the verbage used in the first post. I am not sure how far out in the future he looked, past the draft, to determine that the players with >90 popularity will be elite.

I don't think that's the case at all. The most likely scenario is that initial popularity is generated in some fashion based on the existing talent of the player - i.e. higher talent = higher popularity, on average. To have future booms and busts - which have been confirmed many times as being purely random - dependent on the popularity of the player is just silly.

[Ben E Lou]

I doubt he went out into the future. You can come to this conclusion simply by creating a player files filled with rookies of the same number on the 0-9 scale.

[QuikSand]

I think the vision of having a public list of high-popularity draftees completely upending the draft is a bit misguided. If we're talking about fewer than 100 players in a typical draft... and the sizable bulk of them are already guys who are standouts in some other way (likely great combines)... then what's the practical effect here? Maybe a dozen or so guys in the entire draft suddenly get tagged as looking a bit better than they would have otherwise? And that's including a fair number who move from "undraftable" to "late round project" not just *sure star* or anything. I'm not seeing the practical effect of that (if these numbers are accurate) as totally overhauling the way we evaluate drafts... it's not like the predictive power of this suddenly overwhelms everything else we understand about the draft.

Obviously, each MP league ought to consider what is best to do about the subject...but I really don't see a public revelation of this data rendering the draft pointless or empty. It would just add one more layer of information to the mess we already have to navigate through.

[Antmeister]

Ahhh....okay. For some reason I interpreted the post to mean that the players with >80 popularity was discovered with players who were superstars in the future.

I didn't know you meant the rating prior to training camp. Because, in my way of thinking, I don't know how you can determine that someone is a good starter or elite without having played any seasons.

[QuikSand]

In one respect it will make the draft much more exciting because there will be those that take those risk on low popularity players. However, I doubt you are gonna see many people doing that in the 1st round and 1st round busts aren't gonna happen much anymore.

I think this is the sort of thing I mean above. If you're thinking that there will be a pool of 50 or so guys sitting out there with 95 popularity, and that we'd just see all those guys getting taken with picks #1-50 or so... that's pretty misguided. Look at what Celeval dug up above - from the whole first round in the last WOOF draft, only *two* guys with superhigh popularity, and I can vouch for the fact that there were no fewer than three sure-fire can't-miss franchise-building type of players in there, who naturally went #1 to #3. Tell me exactly who were the guys with 90 popularity, even if both were other players, and I'd *still* rank the same three studs #1-2-3. It's not like I'd junk all the obvious signs telling me that these three studs are studs just because they have initial popularity ratings of, say, 75, 50, and 25 and some other guy has a 95.

[Ben E Lou]

Originally Posted by QuikSand

I think this is the sort of thing I mean above. If you're thinking that there will be a pool of 50 or so guys sitting out there with 95 popularity, and that we'd just see all those guys getting taken with picks #1-50 or so... that's pretty misguided.

To further expound on this, I've pulled together six draft classes and checked. Here are some numbers:

>90: 6 players (avg. 1.0 per class)
>40: 245 players (avg. 40.8 per class)

[Front Office Midget]

Maybe it's because I just finally got into online multiplayer over the winter and have enjoyed being at the top of the eNFL for the last two seasons, and I've never used the Interrogator (don't really know what it does), but ummm... why/how is FOF multiplayer dying?

[Commo_Soldier]

I am not in an online league or anything so maybe I am missing something. But I am not sure how this would be much of an issue since it is viewable in free software that anyone in the league could have access too. If people are willing to take more time in doing research than others then they should be rewarded just like I am sure teams in RL are rewarded. This may be more of an issue if not everyone is able to access the software, but if accessible to everyone in the league than I do not see where a problem would really lie. Other than some owners not wanting to go through as much work, but than as stated as in RL if you do less work than someone else you have to be prepared to possibly pay for doing less.