Player Potential / Development / Progression

[Canuckk]

I think something that's essential for young players in the game (especially during the NHL Entry Draft) is to ensure that we can't see exactly what their current abilities are, numerically-speaking. It should be reliant on what your scouts tell you: I think when scouting a player, they should be able to create a fairly detailed report about them (dependant on how long they scout the player obviously) that lists their strengths and weaknesses in their game by a letter-grade system, as well as estimates their potential and lists character traits. The latter point is especially important as attitude is very rarely properly considered in sports games, and is a key part of a player's development. Not just raw determination, loyalty, etc as in EHM 07, but something more detailed like mental toughness, work ethic, desire to succeed, and "niceness" (ie. if a player is not nice they'll feel less compelled to put up with their coach benching them or playing on a bad team), just to name a few. Things like this should have a huge role in determining how well a player develops - if he has poor work ethic and is not very mentally tough, chances are he won't reach his potential and this should be reflected in the game.

The other thing that I think is important, and this pertains to all players in general, is that the personality attributes should all be hidden, and should instead be simply estimated by scouts, coaches and the media. For example, if a player is a big jerk, he may say something mean to a member of the media, which is then published in an article and, if he is on your team, you could then see that this member of the media thinks your player is a big jerk and has attitude problems. Or if a rift between a player and a coach develops over the player's lack of commitment to the team, the coach could bring it up with you and mention that the player doesn't seem to have a very strong will to succeed. Things like this would make the game feel more realistic and have you staring less at raw numbers and spending more time evaluating players like people do in real life. This idea could also work with on-ice attributes as well, but that may become a little too complicated, at least for an initial release. Though I think it would be very cool to have to make judgment calls based on what scouts, coaches and media say about a player, as well as their statistics and perhaps you watching them in person (if a game engine were to be created). It would provide a far more realistic feeling and give the game more of an unpredictable feel to it as the real NHL does.

Sorry if I posted this in the wrong topic, wasn't sure which one to post it in. Fantastic job in creating this game though dabo, keep up the good work! I've gotta say I'm very excited about the possibility of a wholely new hockey manager game, even if it's a ways away from being released. Best of luck to you!

[dabo]

I agree completely. As I wrote in an earlier post (which I cannot find) I plan on having this letter-grading system:

20-17 -> A
16-13 -> B
12-9 -> C
8-5 -> D
4-1 -> F

Then the attributes (the non-hidden ones) are shown as follows:

A player who is neither scouted nor signed -> Attributes not visible
A player who is scouted but not signed -> Letter-grades visible
A player who is signed (if he is scouted or not doesn't matter) -> Numerical attributes visible

Depending on how good the scout is and how long he has scouted a player determines the likelihood of a correct assessment. Of course a good scout can be wrong too and a bad scout can "guess" right. Let's say a decent scout scouts a player for only a couple of games and gives the player a speed letter-grade of B (16-13), but since he is not the best scout and should have scouted the player for a longer period of time the actual value could have been A (20-17) or C (12-9). So you end up drafting a player who you believe to have a speed attribute of 16-13 but the real value could be as low as 9 or if you are lucky 20. Then in the scouting report we could show what the scout thinks about the player's hidden attributes, this too could depend on how thoroughly the player has been scouted.

Scenario two could be the following:



Instead of letter-grading all Technical-, Mental- and Physical attributes seen above (as suggested first) after a player has been scouted, only the categories in the radar chart could be letter-graded, i.e:

Skating
Physical
Offense
Shooting
Stick Handling
Mental
Defense
Playmaking

Then a radar chart with these letter-grades would be included in the scouting report along with the hidden attributes estimations. Scenario two would bring even more uncertainty to each individual attribute but still give a fair idea about the overall ability of the different parts of a player's game. Then when the player is signed all individual attributes would become visible.

Either of these scenarios together with the feature suggested by one of you to have late bloomers in the game could make the drafting procedure very exciting if you ask me.

I think I prefer scenario two because that would make the draft even more of a guessing game. Which scenario do you prefer? or do you want something completely different?

[Alessandro]

Well, I would add that it woul look a bit dumb if we aren't able to see parameters of let's say Sidney Crosby or Alexander Ovechkin (assuming I'm playing dunno, in teh KHL or in Eliteserien) just because we didn't scout them...
A part this all looks good, and I agree 'bout scenario 2

[dabo]

Regens are the way the game keeps the player pool full and generally talented.

When a player retires he (AFAIK) immediatly is "regen"erated into a player who has random name, and atts. What carries over is the player information page. Height, Weight, Nationality, Position, Offense and Defensive role and Potential ability.

The O/D role will assure that the player plays a similar role as before so that you won't get a guy like Mike Komisarek come back and be the top point producing dman. But the flaw here is that the mentals and physicals don't carry over. A player like Ovechkin or Crosby could regen with 2 agility and 6 speed and over time this drastically diminishes the talent pool. Best way is to have a Physical role to give people a skating role to regen for.

Hmm ok, I don't like it this way and will be looking for another way to create new players.

Well, I would add that it woul look a bit dumb if we aren't able to see parameters of let's say Sidney Crosby or Alexander Ovechkin (assuming I'm playing dunno, in teh KHL or in Eliteserien) just because we didn't scout them...

I see your point but I think it is a unrealistic that you see all the attributes like in NHL:EHM, of course GMs in Sweden has a good idea about Ovechkin's abilities but still I don't want all the attributes to be shown just like that. But maybe we could apply one of the scenarios only to new young players. Not sure really.

[ArtVandelay]

A player who is neither scouted nor signed -> Attributes not visible
A player who is scouted but not signed -> Letter-grades visible
A player who is signed (if he is scouted or not doesn't matter) -> Numerical attributes visible

I think certain players, based on popularity or something to that effect, should have some visible attributes due to media coverage - Crosby is a recent example. Without scouting or signing, I could see him on TV or read newspaper articles about the guy to know he's a big deal.

Depending on how good the scout is and how long he has scouted a player determines the likelihood of a correct assessment. Of course a good scout can be wrong too and a bad scout can "guess" right. Let's say a decent scout scouts a player for only a couple of games and gives the player a speed letter-grade of B (16-13), but since he is not the best scout and should have scouted the player for a longer period of time the actual value could have been A (20-17) or C (12-9). So you end up drafting a player who you believe to have a speed attribute of 16-13 but the real value could be as low as 9 or if you are lucky 20. Then in the scouting report we could show what the scout thinks about the player's hidden attributes, this too could depend on how thoroughly the player has been scouted.

Something like speed being so subjective to a scout bugs me. The worst scout should be able to see that a player is really fast or really slow. To me, what separates a good scout from a bad scout is the ability to identify less obvious attributes like: consistency, leadership, potential, intelligence/decisions, teamwork, "readiness for the NHL" (or whatever league he's scouting for). If I have a scout tell me a guy's speed is B and he turns out to be C (worst case, 16 vs. 9), I should fire the guy and no one else should ever hire him. That's a pretty big potential discrepancy and the scout clearly needs to be in a different line of work or, simply, to be read the frickin' dictionary definition of speed because he clearly doesn't know what it means.

[dabo]

A player who is neither scouted nor signed -> Attributes not visible
A player who is scouted but not signed -> Letter-grades visible
A player who is signed (if he is scouted or not doesn't matter) -> Numerical attributes visible

I think certain players, based on popularity or something to that effect, should have some visible attributes due to media coverage - Crosby is a recent example. Without scouting or signing, I could see him on TV or read newspaper articles about the guy to know he's a big deal.

Sounds like a good idea.

Depending on how good the scout is and how long he has scouted a player determines the likelihood of a correct assessment. Of course a good scout can be wrong too and a bad scout can "guess" right. Let's say a decent scout scouts a player for only a couple of games and gives the player a speed letter-grade of B (16-13), but since he is not the best scout and should have scouted the player for a longer period of time the actual value could have been A (20-17) or C (12-9). So you end up drafting a player who you believe to have a speed attribute of 16-13 but the real value could be as low as 9 or if you are lucky 20. Then in the scouting report we could show what the scout thinks about the player's hidden attributes, this too could depend on how thoroughly the player has been scouted.

Something like speed being so subjective to a scout bugs me. The worst scout should be able to see that a player is really fast or really slow. To me, what separates a good scout from a bad scout is the ability to identify less obvious attributes like: consistency, leadership, potential, intelligence/decisions, teamwork, "readiness for the NHL" (or whatever league he's scouting for). If I have a scout tell me a guy's speed is B and he turns out to be C (worst case, 16 vs. 9), I should fire the guy and no one else should ever hire him. That's a pretty big potential discrepancy and the scout clearly needs to be in a different line of work or, simply, to be read the frickin' dictionary definition of speed because he clearly doesn't know what it means.

I see what you mean but I don't like the idea of having all the attributes visible in front of you from day one. We also have the difference of opinion, perhaps one scout thinks a player's speed is B while some other scout thinks it is A, even if they had scouted the player for years they end up with different conclusions. Then we have god, who is always right, who knows the player is a B (and the actual numerical value shown in the game). So if a scout grades a player differently from the correct grade, it could be a mix between lack of skill and personal opinion. So even if you have two scouts with identical skills, the grades in the reports are not guaranteed to be identical. I agree that it is weird if a scout cannot tell the difference between 16 and 9 but a grade of B instead of C could also mean 13 instead of 12 which isn't too bad. So over time it would be a good idea if you, as a GM, took notice on which scout, in regards to skill + opinion, is the most trustworthy one. This is certainly not a perfect way to do it, I know.

Perhaps scenario 2 would be the way to go then since for example skating is more than just speed.

[ArtVandelay]

Perhaps scenario 2 would be the way to go then since for example skating is more than just speed.

I just re-read that one and I think I misunderstood it earlier. I think I agree that scenario 2 may be the better one.

[bruins72]

The regen system is definitely flawed. It takes a retired player's PA, position, place of birth, along with certain mental attributes, and creates a new player from that. By doing this, it helps keep the talent levels somewhat consistent. You don't end up with periods of too much talent or not enough talent in the "hockey world". Maybe it could be tweaked? Maybe using "player models" is a way to go? Drewst18 mentioned having player types like Sniper, Setup Man, Defensive Defenseman, Puck-Moving Defenseman, Defensive Forward, Tough Guy, and so on. When players are scouted in EHM, the scouts usually use one of these terms (or something similar) to describe what type of player it is. Why not take this and use it when creating a regen? You could take a retiring sniper out of the database and replace him with a new regen that starts off with his skills weighted toward being a sniper (better wrist shot and so on) and he'd have the same PA as the retiring player. You'd probably want to maintain the place of birth or at least region because certain leagues are determined by where a player is from. For example, if a retiring player is from Ontario, you'd want the regen from there because the OHL will be needing a player at some point. If you just let it be random, you could have hardly any players in the OHL but too many in the WHL.

As for the mentoring thing that Laskey and Empach were talking about, it is fairly common. I know often when a young player from another country comes to play with an NHL team, if the team has a young player from his home country on the roster, they'll often take him under their wing. For example, when Patrice Bergeron first came to the Bruins, he was young (18), mostly French speaking and new to the United States. Martin Lapointe played for the Bruins then and came from a similar background. Lapointe had Bergeron move in with his family so he could have a stable family environment around him and acclimate himself to living in Boston. From what I've read, Bergeron says it was a huge help to him sticking with the NHL so soon. It really could be an interesting addition to the game, especially if it affects a player's mental ratings, as these guys suggested.

[batdad]

Crosby and Lemieux did the same. However apparently Patrick Kane did not need it. Nor did Richards and Jeff Carter. Kane appears to have corrupted his mentor John Madden.

[A9L3E]

I think the whole regen system should be replaced. Maybe you could just set values how likely the new player would be playmaker/sniper/power forward... It could depend of how much team's junior academy is investing to each player type. And I think it could be calculated how likely a good player will appear in specific country depending how good junior academies are in that country, how much people are in that country and how strong stature hockey has in that country. Stature could depend on past years success in international competitions and other factors.

[dabo]

My plan is to create new players based on a template (power forward, sniper etc.), for attributes that is (some hidden ones are not included in this), completely independent of who retires or not. bruins72 is right though, still need a way to make sure enough new players are created for each region/league.

I want the talent level of new players to change from year to year.

[empach]

Scenario 2 sounds best.

If you are using the reputation rating EHM had why not use that to determine whether veteran players have attributes automatically visible. Thanks to the media it is very easy for anyone to find out what Crosby or Ovechkin's main skills are, so some or all of their attributes would be visible automatically. Stars would have a certain amount of attributes auto visible, superbs would have less and so on.

Using this Reputation would have to be global reputation. So if someone is a star in the KHL but not well known in North America he would take a hit to his reputation rating. Which I think is how EHM does it right now anyway.

EDIT: To extend this a bit, the top draft picks of every year - Hall, Seguin - are covered pretty well in the media so maybe a few of their attributes would be visible. Just the very top guys. If you asked a moderate hockey fan about Hall they could probably tell you a few things, but they might not even know who Jeff Skinner even is.

If I recall correctly in EHM, sometimes the top draft eligible players have a slightly higher rep than everyone else their age.

I want the talent level of new players to change from year to year.

That sounds good.
The trick to any new player system is to make sure we get a wide variety of players. We don't want a continuous stream of template clones.

[dabo]

I agree, the reputation attribute would probably do it.

[bruins72]

I really like that idea, empach!

[dabo]

Hopefully someone knows the CA/PA system in NHL:EHM enough to answer some questions.

1. What is the general idea of CA and PA?

2. What determines when the player has reached his peak and won't improve any further?

3. How is the CA related to the values of the attributes (technical, mental and physical)?

4. What prevents players who are for example practicing skating intensively to get all 20 ratings on skating attributes (acc and speed etc)?

5. What determines how much each attribute will improve over a players career? Certain attributes are perhaps not improving that much (depending on the type of player I assume).

Obviously you wanna make it so that if a player practices a certain area of his game it is likely to improve, but how much will be different for each player.

I wanna have some kind of attributes cap set when the player is created and then have it adjustable depending on how his career plays out. Let's say a player has a major knee injury at a young age which would prevent him from reaching the projected ability, then the cap of some attributes (skating) would me decreased.

I am trying to get some ideas started on how to do the whole player development system. Any input is welcomed. This is, to me, one of the biggest challenges.

[Animal]

I don't know, CA/PA really confused me when I tried to edit data in EHM07. I think it'd be best, if CA was combined with other attributes, so it'd be kind of "overall" attribute. It'd also make it easier to increase/decrease players' skills.

[archibalduk]

I can partially answer some of those questions for you, just to get the ball rolling. I'm sure there are other users here who can provide more details than me.


1. What is the general idea of CA and PA?

CA = Current Ability
PA = Potential Ability

CA is the player's current ability rating. In other words, his overall ability is rated out of 200. It's an overall rating of how good he currently is.

PA is how good the player could be (if he is still developing/improving) or how good he was at this peak (if he has reached the end of his career and his abilities are starting to decrease). This is the maximum limit that the CA rating can reach. E.g. if a player has a PA of 170 then his CA will never go above 170.


2. What determines when the player has reached his peak and won't improve any further?

There are two factors that I can think of:

1) If his CA has reached his PA. A player's CA cannot go higher than his PA.

2) The player's age. Players reach a point in their career where their skills/abilities peak. Once the player is past their peak, some of their skills will start to decline.

The peak age varies according to position (e.g. goaltenders peak at a much older age than wingers) - I seem to remember a useful guide posted somewhere on TBL. I wonder where it is?...


3. How is the CA related to the values of the attributes (technical, mental and physical)?

The CA influences technical attribute ratings (it doesn't influence non-technical attributes). The CA is used to create an upper and lower ceiling in which the average of the technical attributes cannot exceed.

EHM calculates the average of the technical attributes (i.e. it adds all of the technical attributes together and finds the average) and adjusts the attributes up/down if the average is outside of the CA's ceiling.

So if the player has a CA of 150 then the technical attribute average ought to be roughly 11.5. If it is much higher or lower than this then EHM will adjust the attributes up/down to make the average 11.5.

As a general rule, a technical attribute should be within +/- 6 of the average. Obviously there will be exceptions to this but this is the general rule.

You might want to look at this post by JeffR which refers to a guide posted by Selne. Selne's guide mostly refers to EHM 05 but it will give you some idea of how EHM works with CA/PA. EHM 07's attribute average ceilings are different to EHM 05 in that 05 simply uses CA divided by 100 whereas 07 does not do this. The comments provided by JeffR in the above thread are extremely useful.

Clearly as a player develops, his CA will increase and this will allow a larger average of technical attributes.


4. What prevents players who are for example practicing skating intensively to get all 20 ratings on skating attributes (acc and speed etc)?

I can't fully answer this question (training has always been a source of confusion for me), but:

As mentioned above, the CA/PA will prevent players reaching 20 in all attributes.

I believe that training only gives a very small boost to a player's attribute. In the space of a season you might see a player only go up by 1 or 2 attribute points if training intensively. Also, if you stop training, their attributes tend to slowly drop back down. It's almost as if training gives a player a temporary boost. Or at least this would appear to be the case with older players - not sure about those who are younger and still developing...

As a player trains intensively, he will tire quite quickly. I guess there must come a point when he tires so much that he will no longer progress - he will need a break.


5. What determines how much each attribute will improve over a players career? Certain attributes are perhaps not improving that much (depending on the type of player I assume).

I believe non-technical attributes don't change over the course of a player's career. Could somebody correct me if I'm wrong?

I think it's just the technical ones that develop and change over time. Again, as I said above, CA and PA will determine how a player's attribute can develop over time.

It's also going to be things like having a decent average rating, training, not getting serious injuries, etc.

I think there are three key things here:

1) How game experience develops players
2) How training develops players
3) How a player's attributes will eventually decline - and how does this interact with playing/training.

[dabo]

Thanks a lot.

I believe non-technical attributes don't change over the course of a player's career. Could somebody correct me if I'm wrong?

Can this really be? I am pretty sure I have seen physical attributes increase (and decrease when they get older (skating)). At least I think they should in order to make it realistic. Also mental attributes (some at least) should increase with experience.

[axwel3221]

Thanks a lot.

I believe non-technical attributes don't change over the course of a player's career. Could somebody correct me if I'm wrong?

Can this really be? I am pretty sure I have seen physical attributes increase (and decrease when they get older (skating)). At least I think they should in order to make it realistic. Also mental attributes (some at least) should increase with experience.

I bet Archi meant to say mentals are the only ones that don't change, no?

[empach]

Also mental attributes (some at least) should increase with experience.

My understanding is that they are supposed to. I'm not if it actually works though.

[archibalduk]

Thanks a lot.

I believe non-technical attributes don't change over the course of a player's career. Could somebody correct me if I'm wrong?

Can this really be? I am pretty sure I have seen physical attributes increase (and decrease when they get older (skating)). At least I think they should in order to make it realistic. Also mental attributes (some at least) should increase with experience.

I bet Archi meant to say mentals are the only ones that don't change, no?

Actually I meant all non-technical attributes, but I'm wrong - well, kind of!

Technical attributes are those that are linked to current ability and increase/decrease as the player develops or regresses (i.e. when they're getting older and slowing down). Non-technical attributes do not change very much during the player's career and they don't get better or worse as a player develops and/or ages - they can change but it's very minor changes. I suppose mental attributes would be the same as non-technical attributes - they can change but not by very much.

Based on Selne's guide which he put together with some other users (this one) and on what you can see in game, the following attributes are technical attributes:

Code: Select all

Agitation
Anticipation
Balance
Blocker
Checking
Creativity
Decisions
Deflections
Deking
Faceoffs
Fighting
Glove
Hitting
Off the puck
One on one
Passing
Pokecheck
Positioning
Rebounds
Recovery
Slapshot
Stickhandling
Wristshot
Work rate

I might have missed some, though. Anybody notice any mistakes?

[dabo]

On the page you linked to you can find the following:

All of the technical and and physical attributes can be developed through practise and game experience.
Of the mental attributes only the following can be developed through practise and game experience: anticipation, creativity, teamwork;

But as you wrote only some are used in the CA calculations.

[archibalduk]

On the page you linked to you can find the following:

All of the technical and and physical attributes can be developed through practise and game experience.
Of the mental attributes only the following can be developed through practise and game experience: anticipation, creativity, teamwork;

But as you wrote only some are used in the CA calculations.

That guide is a mix up of EHM 2005 and 2007. As JeffR mentions in a link I posted earlier, there is a fair bit of inaccuracy in that guide - but it is a very useful rough guide.

What I said in the post above is accurate with regards to EHM 2007

[vilifyingforce]

Does anyone know if it is a common thing irl?

Didn't Sidney Crosby live with Lemieux? Team bring in veterans all the time to help the kids know how to act like a pro. I bet Phaneuf has taken Schenn under his wing.

[vilifyingforce]

Also mental attributes (some at least) should increase with experience.

My understanding is that they are supposed to. I'm not if it actually works though.

Some change, but, at very young ages before they ever hit draft radars.

Hopefully someone knows the CA/PA system in NHL:EHM enough to answer some questions.

1. What is the general idea of CA and PA?

Talent of a player currently and their potential to improve.

2. What determines when the player has reached his peak and won't improve any further?

Determination for the base. Training will further improve atts but only so far based off training and ability of coaches.

3. How is the CA related to the values of the attributes (technical, mental and physical)?

CA is only for Technical excluding faceoffs (which can't improve oddly) and hitting.

4. What prevents players who are for example practicing skating intensively to get all 20 ratings on skating attributes (acc and speed etc)?

Not sure, I've only ever been able to improve physical atts a couple of points.

5. What determines how much each attribute will improve over a players career? Certain attributes are perhaps not improving that much (depending on the type of player I assume).

For techs, most improve roughly equal depending on training.


Obviously you wanna make it so that if a player practices a certain area of his game it is likely to improve, but how much will be different for each player.

I wanna have some kind of attributes cap set when the player is created and then have it adjustable depending on how his career plays out. Let's say a player has a major knee injury at a young age which would prevent him from reaching the projected ability, then the cap of some attributes (skating) would me decreased.

I am trying to get some ideas started on how to do the whole player development system. Any input is welcomed. This is, to me, one of the biggest challenges.

I think making injuries give negative atts would work. You could also base att limits around roles/mental makeup. You could have a hidden att for progression making some guys be late bloomers (or have them role a number each season so a player could better have ups/downs/plateaus of remaining pa, this would also make it so some guys never reach their potential.) I think consistent ice time should help a prospect and ability at a level. You can force feed a guy big minutes, but, if he's simply over his head it should hurt his development. Not sure how to handle that one.