Jump to content

Should player attributes scale up/down to a league?


Recommended Posts

I’m playing this new EHM from a different perspective since I played the last retail release several years ago. Back then I played only NHL as that was the only league I had any sort of knowledge about. In the years since we’ve had an EIHL team come in locally. (Braehead Clan) So I’ve been supporting them and following the Elite League since 2010.

I’ve found trying to sign or select players for my line-up as Clan GM rather difficult to suss. Particularly for British players as some of their attributes are so low its makes it so hard to tell if they’d be any use whatsoever on my team. It’s also hard to tell what is deemed “very good” for the EIHL. Does 10 speed actually translate to 20 in the EIHL for example? I’ve been able to sign up Eirk Cole and Keith Aulle for the Clan who have much easier to assess stats, but I doubt I’d be able to get players like that again in future seasons.

Just a few examples, Ben Davies who is a fast skating, very skilled and offers a lot of flair is rated only 7 for deking, flair and accelerating, then just 9 for speed. David Clarke (considered one of the league’s best ever snipers) is only rated 8 for slapshots and 12 for wrist, and even more bizarrely 3 for positioning & 5 for off the puck. Then Jonathan Boxall, a player who may not be the most skilled, is still an absolute work horse type player who gives 100% and is rather physical. Is only rated 7 & 10 for team work and work rate, along with a 5 for hitting and 6 for checking.

I understand these might be fair values in the “global” hockey world scheme of things, and you can’t put those up as higher up teams would just take them for buttons and really mess with things. Should instead player attributes perhaps scale depending on what league a player is in?

Playing in or looking to sign players as an NHL club, you see a player’s “true” attribute values as they are since the NHL is the “best” league in the game with the highest reputation/standard rating. The lower down the tree you go, the more the attributes scale up.

That way Clarke’s shooting attributes in the EIHL can be scaled up to 15+, but if an NHL GM were to look at him, they’d see them as 8 & 12, therefore not worth them signing him.

Hope I got that across ok, would like to see something that makes things a little easier to judge for recruiting & selecting in lower level/junior leagues.

Link to post
Share on other sites

Ben Davies has random 0 declared in attributes: flair, accelerating and deking so what you see in game depends on rating and declared roles for that player. Many players are rated in that way cause user feedback is not much with attributes. Firstly I think we should improving database with missing players and later work on their attributes. Volunteers with help are always welcome. :)

Link to post
Share on other sites

My understanding is Player Attributes are scaled based on their CA and Player Role

CA is between 1-200, and "most Player Attributes, most of the time" fall within a range that connects to CA

So to have an Attribute close to 20 (on the 1-20 scale) you'll usually have a CA in the 160-200 range, and Attribute around 10 would usually indicate a mid range CA.......but it's not that simple! But to explain it initially, the comparison suffices

Other significant considerations are...

Each player has more than 50 Attributes (some Attributes are not visible ingame; CA is "hidden" too).

AND some Player Attributes are not at all based on CA (such as Injury Proneness and Pass Tendency).

AND some Attributes (like the skating Atts) are less likely to use the very bottom of the scale.

AND an individual player may be very good at an individual skill or a few skills, so a player may have a Wristshot that is much better than his CA would indicate

ALSO Player Roles emphasize certain skills over others, so a defensive winger would be more likely to have higher Positioning and Checking than a sniper winger, and those Attributes are more likely to develop faster and to a greater extent for the defensive winger.....looking at offensive Attributes like Off the Puck and Wristshot, the opposite would be true: the sniper would be more likely to develop Off the Puck and Wristshot faster and to a greater extent than a defensive winger

Looking at the same two example players...if at game startup both had 0 for their Attributes and the same CA, the game would give the defensive winger a higher initial value for Positioning and Checking in comparison to the sniper winger (and the sniper winger would be given higher initial values in offensive Attributes like Off the Puck and Wristshot in comparison to the defensive winger)

I think that essentially Attributes scale down to a League, as the players abilities scale down to a League.

When you say "David Clarke (considered one of the league’s best ever snipers) is only rated 8 for slapshots and 12 for wrist" I immediately wonder if he is one of the league's best? You can look at all players Attributes in the League, and see if his 8 and 12 place him at or near the top

My understanding is the typical CA range (from Poor to Top) for the EIHL is 20-90 and the EPIHL is 20-50

If you don't know already and you're interested there's LOTS of information on EHM available at TBL http://www.ehmtheblueline.com/forums/portal.php

Here's the Player Attributes Guide http://www.ehmtheblueline.com/wiki/index.php?title=Attributes:_Players (note Positioning is a defensive skill, and doesn't relate to being a Sniper)

Link to post
Share on other sites

What the OP is asking is if the presentation of attributes should be re-scaled according to what league we play in.

As players skills are essentially a Gaussian distribution curve there are problems if you remove every value above a cut-off point which is what you do when all better players sign in higher rated leagues. If you think of a classic bell-curve and remove the top 30% of the curve you'll see that it becomes a lot harder to differentiate between the remaining top 10-15% of players compared with if you have the complete player curve at your disposal like the top leagues do. The best players will easily stand out if you have the whole curve in your player pool but not if the top players are all signed to higher leagues.

In order for this to work the [presented] attribute scale can no longer be linear but has to have a higher granularity at the top end than at the bottom and take the form of an inverse logarithmic curve. Furthermore the curve has to follow a function that is weighted differently for every league depending on the "quality" of that league.

It's an interesting idea and would make it easier for everyone playing in leagues other than the top three or four. How hard it is to get it right and make it workable is a totally different question...

Link to post
Share on other sites

I don't know if I'd want scaled attributes (although I see your point), but one thing that I would like that may help along with your needs: Something like a league average. What is the average speed of an EIHL, AHL or NHL defenseman at that time? I seem some of the other side, where if everyone is 13-17 outside of the stars, are they above average of the talent in the league?

Link to post
Share on other sites

What the OP is asking is if the presentation of attributes should be re-scaled according to what league we play in.

Exactly this. Since the game runs on an older version of the FM engine, League standard and reputation remain static so you could use that as a base value for attribute scaling.

Since the NHL is the "best" league, there isn't any change so you see everything as it is in the DB if you're the GM of an NHL team. KHL is generally considered the next best league, so you could say maybe 1.10 for attributes in that league. (Scale up 10%.) SHL 1.15, Liga 1.20, AHL 1.25 etc and so on.

If not that, it would be extremely helpful as I've already said to have some sort of information to go on that tells me what is a good attribute value for the league I'm managing in. Especially for junior leagues as well, as virtually all the junior players British players I see have virtually every attribute in red, so it's very hard to tell within the game mechanics who could be the stand outs, they all look of the same standard right now.

Link to post
Share on other sites

I’m playing this new EHM from a different perspective since I played the last retail release several years ago. Back then I played only NHL as that was the only league I had any sort of knowledge about. In the years since we’ve had an EIHL team come in locally. (Braehead Clan) So I’ve been supporting them and following the Elite League since 2010.

I’ve found trying to sign or select players for my line-up as Clan GM rather difficult to suss. Particularly for British players as some of their attributes are so low its makes it so hard to tell if they’d be any use whatsoever on my team. It’s also hard to tell what is deemed “very good” for the EIHL. Does 10 speed actually translate to 20 in the EIHL for example? I’ve been able to sign up Eirk Cole and Keith Aulle for the Clan who have much easier to assess stats, but I doubt I’d be able to get players like that again in future seasons.

Just a few examples, Ben Davies who is a fast skating, very skilled and offers a lot of flair is rated only 7 for deking, flair and accelerating, then just 9 for speed. David Clarke (considered one of the league’s best ever snipers) is only rated 8 for slapshots and 12 for wrist, and even more bizarrely 3 for positioning & 5 for off the puck. Then Jonathan Boxall, a player who may not be the most skilled, is still an absolute work horse type player who gives 100% and is rather physical. Is only rated 7 & 10 for team work and work rate, along with a 5 for hitting and 6 for checking.

I understand these might be fair values in the “global” hockey world scheme of things, and you can’t put those up as higher up teams would just take them for buttons and really mess with things. Should instead player attributes perhaps scale depending on what league a player is in?

Playing in or looking to sign players as an NHL club, you see a player’s “true” attribute values as they are since the NHL is the “best” league in the game with the highest reputation/standard rating. The lower down the tree you go, the more the attributes scale up.

That way Clarke’s shooting attributes in the EIHL can be scaled up to 15+, but if an NHL GM were to look at him, they’d see them as 8 & 12, therefore not worth them signing him.

Hope I got that across ok, would like to see something that makes things a little easier to judge for recruiting & selecting in lower level/junior leagues.

Hello!

Don't be so sure about E Cole being the last NHL guy you'll see... even in real life the EIHL usually attracts one or 2 ex NHLers each season, same will happen on EHM. The EIHL is always fun to play. The way I look at it (in a very basic way) is basically get the best possible imports I can first. So if ex NHL man wants to come, get him before he goes to Russia or Germany! Obviously try not to bankrupt my team at the same time.

With the Brits I try to load up on the 'main' ones that 1. I know are actually good, say Ben O'Connor and 2. Keep a check on Team GB, sign them if possible to fill the 2/3rd lines. If I have to go hunting for Brits I don't really know I just look for the absolute basic stats to start with, so I want a D man. I want him to have some size, be patient, able to tackle, check. I want a Centre then as long as he can skate and win a faceoff it will do, then asses them over the season and see if they are worth while

Not rocket science but It helps me. As you say how do we know who is good and who isn't.

Link to post
Share on other sites

Since the NHL is the "best" league, there isn't any change so you see everything as it is in the DB if you're the GM of an NHL team. KHL is generally considered the next best league, so you could say maybe 1.10 for attributes in that league. (Scale up 10%.) SHL 1.15, Liga 1.20, AHL 1.25 etc and so on.

That will unfortunately not help you. All such a change will do is to make all your "9-10" players turn into "10-11" players and you'll still not be able to see any difference between them. It will only shift the numbers upwards not make them more useful.

You need to magnify the differences between the players for it to work hence the "inverse logarithmic function" I suggested. Such an approach would show greater differences between players the better they get and by calibrating the function properly (for the specific league) you'd get a meaningful difference between the players (that are relevant to your league).

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...