EHM The Blue Line

Following the recent release of the 2013/14 TBL Rosters v6.0 for Eastside Hockey Manager 2007, there has been a lot of feedback about how things can be improved. As has been discussed in the TBL Roster Update Forum, we are very keen for users to help us improve the ratings in the DB. Before I explain how everybody can help, here's a bit of background info first:

First of all, I have previously explained the reasoning behind why I updated the NHL (see this post) and I have also explained how I went about working on the update (see this post). I knew the Current Ability ratings were still a work in progress when I imported my research into the DB in December, but I hadn't appreciated how much was still left to be done. I did the CAs back in May/June and, except for various adjustments I made recently, I hadn't touched them since as I moved my concentration on to the attributes and the realism patch (and of course in amongst that I took a break from the whole project for a couple of months whilst I re-wrote the Updater, moved house and finalised the FHM rosters). Looking back through the CAs this week, I can see that in fact I left a lot still to be done (whereas I had been happily assuming since the summer that the CAs were pretty much finalised).

What I'd done as a starting point was to divide the players according to their primary position (Goalies, Defencemen and Forwards). Then I moved a proportion at each position to Top, Good and Average. The reason being was that I found that there were greater proportions of certain positions within some of the four different brackets (i.e. Poor, Average, Good and Top). I can't remember which ones exactly, but by way of illustration, there may have been a disproportionate amount of Forwards at Top/Good and relatively few Defencemen within those brackets. So I pushed up all positions to try and equalise it.

Prior to taking a break from the updating work (and subsequently forgetting that the CAs needed a lot more work), I'd planned on pulling the CAs down. So that at each position and for all players rated with a CA of 100+, the distribution would be more like this (more or less):

[table][tr][th]Bracket[/th][th]CA Range[/th][th]Goalies[/th][th]Defencemen[/th][th]Forwards[/th][/tr]
[tr][td]Top[/td][td]180 - 200[/td][td]1.3%[/td][td]1.3%[/td][td]1.3%[/td][/tr]
[tr][td]Good[/td][td]140 - 179[/td][td]25%[/td][td]25%[/td][td]25%[/td][/tr]
[tr][td]Average[/td][td]110 - 139[/td][td]53.7%[/td][td]53.7%[/td][td]53.7%[/td][/tr]
[tr][td]Poor[/td][td]100 - 109[/td][td]20%[/td][td]20%[/td][td]20%[/td][/tr][/table]

And this would have resulted in a roughly a 120 average CA for NHL players with a 100+ CA. This is the sort of distribution of talent for the NHL that EHM had been designed for (see the first link I posted above for more info).

Where we are at the moment is that the CA ratings are somewhat top heavy. Things need to be brought down as I'd previously planned. Also, there are of course plenty of players who may well need to be boosted too. As I've said before, this is my fault entirely; but on a more positive note we now have plenty of users chipping in with useful suggestions and feedback - and this is exactly what the NHL update needed in the first place (and again, it's a lack of volunteers that led to me taking on the NHL in the first place).


Now that I've explained the background, on to the more pressing issue:

To the best of my knowledge, Manimal hasn't yet finalised his plans on how to deal with all of the feedback received thus far. However, we certainly need the feedback to keep on coming because we have 30 teams of rosters to fine-tune. To this end, there are a couple of ways you can help us:

1) You can post feedback in the Feedback Thread.

2) You can post specific feedback on a group of players using Nino's really useful threads: Goalies / Defencemen / LW / C / RW

3) You can download these spreadsheets. See the details below for more info.


The Spreadsheets

As mentioned above, you can download this zip file containing two spreadsheets. Here's an explanation of each spreadsheet:

NHL v6.0 - OVERVIEW.xls
This spreadsheet is designed to try and make the re-rating a little easier. Although the CAs are not the be all and end all of rating players, they are a useful starting point and it is important we get them right. As mentioned above, there is an approximate ideal distribution of talent across the four Poor/Average/Good/Top brackets.

This spreadsheet shows position by position the CA ratings and CA brackets for every player currently in the NHL with a CA rating of at least 100 (there is a worksheet for each position). At the top of each worksheet it explains approximately how many of the players should fall within each bracket.

There are two columns highlighted in red/orange - these are for you to suggest any amendments which you think need to be made. All you have to do is enter your proposed changes under the CA Bracket column. Optionally, you may also propose a CA for that player (but you don't have to).

The challenge is to amend the players' CA Brackets so that they are distributed as explained at the top of each worksheet. You can then email your completed spreadsheets to ehmtheblueline@gmail.com . Please note that you don't have to amend every single player at every single position - just do as much as you can. Hopefully if we can get enough users to help with this then we'll be able to vastly improve the CA ratings.

Btw, I appreciate not every NHL rookie is listed on the spreadsheet. It was going to get too long and complex if I included all of the NHL-contracted players currently with a CA of less than 100. Obviously any comments about players not included within the spreadsheet can be posted in the general Feedback Thread.

Finally, here's an example of how to complete the spreadsheet. You'll see in the screenshot below it is suggested that Price and Rinne move down from the Top bracket to the Good bracket and Ramo move down to the Average bracket. This is just to show how to complete the spreadsheet.




NHL v6.0 - DETAILED.xls
This spreadsheet follows on from Nino's threads (see the links above). This breaks down the data by position and provides details of the CAs, PAs and key attributes. You may find this useful when posting in Nino's threads and/or the Feedback thread. Here's how the spreadsheet looks:

One final point: I just wanted to highlight what Batdad mentioned about team chemistry. It's pretty important to bear this in mind because it has a bearing on how players should be rated (i.e. we want them to perform in game as they do in real life - even if that means that some players, e.g. Kunitz, need to be a little overrated). I have highlighted the key points in bold underline:

batdad wrote:Sure in real life a guy like Kunitz feeds off the stud he plays with--Ie Crosby. But in EHM...there is really little to link this together. So....with a guy like Crosby who is not a big monster (albeit the weight height equalizaton helps this a bit) notoriously underpeforms for his real life ability in EHM....and does NOT make the players around him better in EHM like he does in real life.

This means that if you downgrade a guy like Kunitz too much...he does not do nearly as well as he does in real life. Another example of this is Alex Burrows with the twins in Vancouver. He does not succeed in EHM at all, and ends up being a scrub or 4th line guy on various numbers of teams, in spite of the fact that he scores and does a ton when playing with the twins or Kesler in Vancouver. Brown in LA....all the medium sized guys.

You cannot downgrade them too much or they really have little to no success in the game compared to where they are in real life. Just a fact that the dynamics of the game do not allow for an attribute that could be known as "magic"....making the platers around you look much better than they are. You need a superboost rating in the game that does not exist, or underpeforms if it does exist....

So while Kunitz may only be a 20 goal guy outside of Pitt, fact remains for the time being he is a Pen, does play with Crosby and does light the lamp a ton. The only way to allow this to happen and make the Pens in EHM what they are is to over rate a guy like Kunitz individually. Same for Burrows, same for Brown, same for guys like Martin St Louis.

They just do not have the ability to succeed. Conversely, if you rate a guy like Backes too high he becomes a superstar with ridiculous numbers in EHM. Patrick Marleau from the original DB syndrome.

Size is over rated in the game....ability to make other players around you better....totally non existent. That has to be balanced. THere are few players out there that make the guys around them better...it is really hard to rate.///and then top it off with size being dominant...a guy like Crosby is brutal to rate accurately to perform at the high levels he does, whereas Backes, Thornton, Marleau are way too easy to overrate.

So my personal opinion is ...if a guy is over rated because of who he plays with it is a fact we have to live with. We cannot assume Pit will trade Kunitz in the game, and must leave him the way he is so the Pens peform at a high enough rate. To this point with my experiences in the game....Crosby and Kunitz do not perform at a high enough rate. Rarely does Crosby hit the top 5 in scoring for example.

archibalduk wrote:NHL v6.0 - DETAILED.xlsThis spreadsheet follows on from Nino's threads (see the links above). This breaks down the data by position and provides details of the CAs, PAs and key attributes.

I was difficult to decide what to include when I created the threads...I didn't want to overwhelm people with "to much" and so I did my best to limit the Attributes I included...at this point I'd like to add a few thoughts...


For goaltenders Positioning is also very important.

Anticipation is very important for "top players" (top 6 F/top 4 D)

For the “top offensive players” I’d recommend looking at Stickhandling, Deflections and Deking too.

For “defensive players” Bravery is also relevant (i.e. shot blocking)


Work Rate and Determination are relevant too (if left as "0" the game to often generates values that are to low), as is Stamina and Natural Fitness...

I definitely think...ranking by position instead of doing individual teams is the way to go with the db. It will help in making people think outside their team. I totally get that people feel they are experts on their own home team...by default I know more about the Canucks than I do any other NHL team, just because that is all they talk about here. But...to think outside the box of your own team and the top players on it, is the way to make people really think...and to learn more about the game and players in the game.

I wish...really wish I had the time to rank what I believe the order of players in each position should be and group them into categories. I would just get shot though at home for doing it as it would take up too much time.

Finally....it will have to be constantly reminded that big players will need to be toned down, players playing with the elite studs need to be upgraded a bit beyond what they normally are, and that players who are small but succesful in real life will need to be tuned up beyond what they really are in order to make them perform WITHIN EHM.

I would suggest that we try and avoid discussing players by teams for now, and move to discussing how they fit in within their positions across the league. I will be pointing that out in the other thread as well....

I know guys do not feel comfortable rating outside their own home team, but I think they need to start looking at expanding their knowledge base in order to contribute effectively and make the DB as good as it could be.

again....not easy to do. People favour their home team and yes...the one they have the most knowledge of.

But..we need it to be realistic and that is possible only if we try and move them out of rating their own teams.

Nifty, I'll spend some time on this. I like doing thing kind of thing.

Thanks for all this effort.

Mr.Pickles wrote:Nifty, I'll spend some time on this. I like doing thing kind of thing.

Thanks for all this effort.

And thanks for all your help and feedback thus far.

I noticed a couple of issues and queries coming up in the Defencemen and Left Wingers threads which I thought I'd address in here. I think otherwise we're at risk of being side-tracked.


The Crosby/Kunitz, etc Chemistry/"Magic" Point
I don't think it's worth getting too carried away with this. As with all players in the game, we want to rate them as accurately as possible within the limits of the game. We either rate them according to their performances with their "magic" line-mates (i.e. over-rating their raw talent) or we rate them according to how they are in isolation (i.e. under-rating how well they are playing with their magic line-mates). Ultimately we're looking to realistically portray the NHL within the limitations of EHM. Thus if Kunitz is getting 50+ pts per season in real life with the Pens then that's probably what we want to see in game (or alternatively some sort of compromise where he's achieving more-or-less near that mark). Of course if he subsequently moves to another team and suddenly drops down then we can revise his ratings. That's not a problem.

Equally, we will do our best to ensure the likes of Crosby and Thornton are accurately rated in game within the limitations of the game and according to how players are supposed to be rated.

It's not worth spending too much time on this issue. Manimal has been running the roster updates for a number of years now and he'll know how to best deal with this.


Rating by Team or Rating by the Entire League
We absolutely have to rate all players in the league as a whole. The only extent to which we can break it down is by position (this is to ensure an equal spread of talent between each position - otherwise the temptation is to put all of the superstar forwards at the top of the ratings and forget about the top defencemen, etc).

Of course we can look at players on a team-by-team basis in order to consider who are the best/worst players in each team. But we cannot actually rate a player's CA until we consider them within the context of the entire league. A player might be an average defenceman for his particular team, but that doesn't necessarily mean he is an average defenceman when compared to all defencemen in the NHL - he might actually be better placed within the Poor or Good bracket.

I have tried to make things as easy as possible by uploading the "NHL v6.0 - OVERVIEW.xls" spreadsheet. As I said in the first post in this thread, users can do as much or as little as they want. Just use the spreadsheet to highlight which players are under-rated and over-rated. We can then collate all of the spreadsheets returned by users and start to finalise the ratings.


Reconciling Defensive and Offensive Players
I can't find the post right now, but I saw someone mentioning a short while ago the difficulty about rating players grouped by position because some players are strong offensively and others are better defensively. Unfortunately this is the drawback of the game's Current Ability system (contrast this with FHM which has separate defensive and offensive CA ratings which makes things so much easier). We just need to do the best we can within the limitations of the game. Perhaps if you find, for example, a defensive defenceman and an offensive defenceman hard to compare, try considering the overall value to their respective teams and perhaps even what lines they play on. If in theory both players were on the same roster, which player would offer more value to the team? I appreciate the players' value would vary by the needs of an individual team, but at least thinking along these lines might help a bit.

But as I've said before (and as I explain further below), the CA rating is not the decisive factor in a player's performance in EHM. So it doesn't matter too much whether Player X has a CA of 120 or 130. It's the attributes which make a player good or bad (as well as the Defensive and Offensive Role ratings which affect how good a player is in each role).

So it's not worth agonising over too much. Just get the players within the correct bracket (i.e. Poor/Average/Good/Top) and then roughly in the correct order.


Current Ability is NOT the Key Rating for a Player
The CA is just the starting point. As I've mentioned before, EHM uses it to ensure that a player's Technical Attributes are within the correct range. It is also used for player development (i.e. so a player does not progress beyond his PA). It's the attributes that affect whether a player is good/bad; not necessarily the CA. And the CA does not affect any of the Non-Technical attributes. A player with 120 CA could well be better than a player with 150+ CA if his Non-Technical attributes are much better than the other's.

It's not worth getting too caught up in whether Player X should have a CA of 120 or 130, etc. Like I say, just get the players within the correct bracket (i.e. Poor/Average/Good/Top) and then roughly in the correct order.

Ideally we'd follow the process mentioned in this post. We'd have users commenting about all of the different attributes players currently have and advising whether they should be increased or decreased. However, experience tells us that in reality we get very few volunteers to actually do this for such a big league. Hence I've started this thread and have used CA re-rating as a starting point. Also, because the CAs do still need a lot of work (for the reasons detailed in the first post of this thread). Once the CAs have been completed, it would be good to move on to the PAs, then we could ask for feedback on specific attributes (such as those listed by Nino in his threads) and then of course there's the prospects.

With regards to attributes, I'm of the view it's better users just say a player is under/over-rated for a particular attribute rather than necessarily giving a specific new rating. Then we leave it to somebody like Manimal or another experienced researcher to re-rate those attributes as necessary (this ensures consistency of ratings).

archibalduk wrote:Equally, we will do our best to ensure the likes of Crosby and Thornton are accurately rated in game within the limitations of the game and according to how players are supposed to be rated.

My two cents...I'd choose the above over artificially inflating a players ratings

archibalduk wrote:Reconciling Defensive and Offensive Players

I think there's some decent examples now...I just took a quick look at the "key Atts averages" in the NHL Defensemen Ratings thread and the listed D (with Top/Good CAs) show both low offensive/high defense as well as high offense/low defense

archibalduk wrote:the CA rating is not the decisive factor in a player's performance in EHM. So it doesn't matter too much whether Player X has a CA of 120 or 130. It's the attributes which make a player good or bad

Agreed (this is why I included the Attributes I did)

Hi, first time commenting so sorry if this is in the wrong place. Certain contracts with the roster seems to be incorrect, for example Corey Crawford, Evgeni Malkin, Alex Steen, Henrik Lundqvist and a few more all went on to FA at the end of the first year. Now whilst this was exciting it's not exactly the most realistic of games. Is this due to be sorted in the next update?

danmara87 wrote:Hi, first time commenting so sorry if this is in the wrong place. Certain contracts with the roster seems to be incorrect, for example Corey Crawford, Evgeni Malkin, Alex Steen, Henrik Lundqvist and a few more all went on to FA at the end of the first year. Now whilst this was exciting it's not exactly the most realistic of games. Is this due to be sorted in the next update?

Oh and keep up the generally great work. Love all the roster updates

danmara87 wrote:Hi, first time commenting so sorry if this is in the wrong place. Certain contracts with the roster seems to be incorrect, for example Corey Crawford, Evgeni Malkin, Alex Steen, Henrik Lundqvist and a few more all went on to FA at the end of the first year. Now whilst this was exciting it's not exactly the most realistic of games. Is this due to be sorted in the next update?

No, it is not.
The 6.0 roster update is set with 2013-14 contracts and extensions are not possible to add

Is it known whether Favourite Staff has any in-game effect at all?

just want to say that the san antonio rampage the panthers affiliate team is in echl, not ahl, its a pain cos it makes it very hard to move players around

How long have you been here? have a look around and you will find you answer as to why they are there. How much do you know about the game? How many NHL teams are there? How many AHL Teams are there in EHM? Please change that for us if you can.

Maybe this is a dumb question but in the spreadsheets under PA what does a number like -9 mean under say Stamkos? Does it mean that their attributes will either go up or down by 9? I want to make sure I have a clear answer before rating forwards.

Thanks!

darn it I was in that thread earlier but forgot where it was. -9 is like one of the best player in the league. Like a 190-200 PA. Then you have -8, -7 etc... and IIRC -10, -11, -12, -13 and -14 are random for young players that we can't really know their full PA.

I'll try to find the thread.

Edit:

-1 Potential will be between 1-20 (they will likely retire at a young age and keep getting recycled back into the regen pool)
-2 Potential will be between 10-40
-3 Potential will be between 30-60
-4 Potential will be between 50-80
-5 Potential will be between 70-100
-6 Potential will be between 90-120
-7 Potential will be between 110-140 (quite rare – a good prospect with a decent chance of being a regular NHL player)
-8 Potential will be between 130-160 (rare – a top prospect with a good chance of being a talented NHL player)
-9 Potential will be between 150-180 (very rare – a top 10 prospect who is considered a can’t miss NHL talent)
-10 Potential will be between 170-200 (extremely rare – the top player available in a draft, but not every year)
-11 Potential will be between 20-80 (intended for very young, hard-to-predict players)
-12 Potential will be between 40-100 (intended for very young, hard-to-predict players)
-13 Potential will be between 60-130 (intended for very young, hard-to-predict players)
-14 Potential will be between 90-160 (intended for very young, hard-to-predict players)
-15 Potential will be between 110-190 (intended for very young, hard-to-predict players

Hope it helps! I know somewhere Archi posted his way of doing it and it was good but the core is there.

That is great. Thank you!

You're welcome.

philou21 wrote: I know somewhere Archi posted his way of doing it

There's the Research Guidelines Archi posted last year http://www.ehmtheblueline.com/forums/vi ... 40#p145540
There's also the Player Attributes Guide http://www.ehmtheblueline.com/forums/vi ... it=realist

I've moved the posts about scoring levels to the Scoring Feedback Thread: http://www.ehmtheblueline.com/forums/vi ... 45&t=11768

Now I know this is in the wrong place and there is probably a topic on this but I do not have the time nor do I have the patience to go and look for it.

Can somebody explain to me how the function of not using real player names work? Does the game generate completely new people with random attributes or are the names of the real NHL stars just changed and all the attributes are the same. Any help is appreciated and thank you.

If you want an answer, find the time and patience. Or try it and find out. Yes i am a meanie.

Welcome aboard and hope you have a look around and learn from the site ... hopefully without asking questions in wrong threads and derailing them like I do with frivolous comments.

To answer you...guys from show with different names for most part.

Why not Duchene, Landeskog or many more extension contract?

Cannot do the extensions in the game as it messes up their cap hit for this season. That info is all over the site, have a look around.

batdad wrote:Cannot do the extensions in the game as it messes up their cap hit for this season.

Would that really be so bad? Because currently many of the players who've had recent extensions seem to become FA's in the game, which will lead to totally wrong rosters fast. Personally I'd rather have believable rosters at the expense of a few incorrect cap hits. Of course one might need to make compromises with the cap hits to ensure that teams wouldn't go over the limit.