TBL Rosters 2013-2014; Mistakes, omissions, tweaks etc

[nino33]

Does the game handle weight different with youngsters? I have noticed that prospects weights are usually way off but in the db their weight is listed correctly. Does the game view it as the weight they will grow into instead of their current rate?

Back in 2011 I did some significant testing of Height/Weight development http://www.ehmtheblueline.com/forums/vi ... 110&t=8146


The testing showed younger players usually receive a height/weight at start-up!
My testing results showed virtually all height/weight development occurs only at start-up
Age 12-16 saw a height/weight increase pretty much 100% of the time, age 17-18 the majority saw a height/weight increase, and at age 19-20 about half saw an increase (usually just weight)



The amount of height/weight increase is random

For example, Alex Pietrangelo (age 16 years 1 month at start-up) started with a 189 height and 87 weight. At start-up he was191/99 in the first test,192/95 in the second test and 195/97 in the third test.
Another example...Ryan Nugent-Hopkins (age 13 years 4 months at start-up) started with a 178 height and 67 weight. At start-up he was188/83 in the first test,185/89 in the second test and 185/90 in the third test.

[nino33]

Can someone tell me how I can edit my own rosters on NHL Eastside Hockey Manager 2007.

You can download the pre game editor or the save game editor which are easier to use if its just for you

Pre game: http://www.ehmtheblueline.com/forums/vi ... =88&t=6745

save game: http://www.ehmtheblueline.com/forums/vi ... =88&t=6831

There's also the EHM Updater! http://www.ehmtheblueline.com/forums/vi ... 10&t=11154
If you've used excel before and you're planning to edit more than just a few players I'd recommend the Updater

[boxhamster]

First, sorry for my very bad english.
I find that some players are very unrealistically high rated.
Only a few.
Defenders: Orpik, Staal, Mitchell,Hejda, Smid, Gill

Under rated Defenders:
Ehrhoff, McDonagh,Lindholm

Too high rated Forwards:
Ladd, Selänne, Alfredsson.

Under rated Forwards:
Steen, Kreidler, Hagelin

I think that you should take into consideration for the next update the existing performance of the player in this season.

[Peter_Doherty]

Hagelin is rated fine i think, his attributes are just in the wrong place. He should be quicker with his shooting skills and passing lowered to about 12-13ish. Maybe some other tweaks too but those are the major ones.

[boxhamster]

When I look at the game in the forwards list then Hagelin is behind Players like Brunner, Yakupov, Bouchard. Last and this Season Hagelin is one of the Rangers best Forwards. I hope you know what i mean.

[batdad]

That the other three should be toned down?

[archibalduk]

I posted this in the Goalie Feedback Thread, but also wanted to post this here as it is just as relevant:

That being said...this database may be the best thing that has happened...it seems to be getting some people off their butts to help out. I thank all of the people who have worked so hard on this db and HOPE that those who are writing in here will continue to contribute.

As Batdad says above, it's brilliant to see that the DB has provoked people to really look into the NHL ratings. For years we have been asking for users to help re-rate the NHL and we've never been able to get the level of help we need (we do get a couple of users who very kindly help with one or two teams, but we need enough users/help to cover the entire league). Hence it was left to me to try my hand at re-rating the NHL; otherwise we'd be left again with the same old ratings which are generally both out of date and do not really follow what SI planned when they developed the game.

Although I've followed the Bruins and, to a lesser extent, a number of other NHL teams for many years, there are many, many other users here who could do a better job than I (I'm far more knowledgeable on the British leagues than the NHL). There was also the additional challenge of trying to re-rate approx 1,000 players on my own (as well as spending a lot of time on improving the scoring, PIM and fighting levels). I've had to use my own (limited) knowledge, depth charts, stats and the existing attributes as a basis for the re-rating. I had hoped that things would at least be along the right lines, but the general feedback seems to be that they are simply not good enough. I'm sincerely sorry for that and it's clearly entirely my fault.

On the plus side, the re-rating seems to have gotten everybody talking on a much larger scale than ever before. I think this is really positive for the DB in the long term. Once we have planned out how to go about improving the DB based on all of the feedback we have been getting, I think the NHL ratings will be in the best shape they have ever been - and we'll have achieved exactly what we had hoped for (i.e. a much improved NHL in the game)

DO NOT RATE PLAYERS for the game in a bubble of feeling, emotion or short term success. I mean Lars Eller had a 10 game point streak...so I guess he should be rated like OVenchicken? See what i mean...things can get out of hand and have to be very very careful or you end up with a database of every player being rated with all super star stats.

I really want to reiterate this quote from Batdad. Additionally, as a I mentioned last week, I will post an explanation of how I re-rated the league. The aim behind the re-rating was to make the NHL more realistic in terms of ratings. There was a lot of over-rating in the DB in the NHL (and this stems from the original EHM 2007 DB). So I would suggest waiting for my explanation before deciding whether player X should have a 190 CA, etc because I do feel some of those comments are based on the traditional over-rating mindset. It will take me an hour or two to write the explanation, but I hope to have time to do this tomorrow. I think this will help frame how CAs, etc for any particular player should be adjusted.

I think it is much more helpful at this stage to say that Player X should be better/worse than Player Z (which in actual fact is how most people have been commenting thus far ) - i.e. by comparing players with one another rather than arbitrarily setting ratings in isolation. When considering whether a player needs his attributes increasing/decreasing, we need to take into account how they're already performing in the game, how they compare to players of similar real life skill, and their height/weight (as EHM over-emphasises tall and heavy players).

I also want to highlight this very useful thread which gives some basic guidance on how things ought to be done (I've added red, bold typeface to some of the key points):

Riz made extensive changes in EHM2007 to the CA-attribute relationship to allow for more variation in attributes for high- and low- CA players; the earlier versions of the games tended to make it impossible to give NHL star-caliber players very low ratings in a technical attribute - you could enter the number in the database, but the game would see the big CA and inflate it.

Those changes are also the reason why the old approximations that draw a direct line between CA and attribute totals (e.g., attribute average = CA/10, CA = sum of attributes, etc.) aren't a good idea when working with EHM2007 data. You'll wind up with underrated young players and a glut of overrated stars if you do that (unfortunately, some of the 2007 researchers weren't exactly diligent about paying attention to the new guidelines, so the stock game has tendencies in that direction already.)

Nontechnicals, for the most part, are independent of CA and should use the whole 1-20 range, clustering heavily towards the middle. Some of them (e.g., the skating-related ones) do get adjusted a bit by CA, but not to an extent that's worth worrying about when you're doing the ratings.

When I was doing the last revision of the research guide, I spent quite a while talking back-and-forth with Riz about what he'd changed with CA to make sure I understood it properly and could explain it to the other HR's, in the hope that everybody would get the message. That was an unusual amount of time for him to spend on db-related stuff, but it was that important. Unfortunately, while some of the guys paid attention to the big red "read me, it's important" text in their new guide - Matt Bosela's one I remember getting it exactly right - not all did, and there wasn't time to fix all of the problems. But no reason to keep repeating their mistakes when it can be clarified now, right? Much easier to avoid people creating bad data in the first place than it is to find and fix the problems later - believe me, I know, after spending hundreds of hours trying to undo damage done to some of the league db's, with only mixed success.

What I'd suggest doing is working on a case-by-case basis for the players based on what looks "wrong" with their stats - the best starting point is probably getting their shots-per-minute and minutes played into a reasonably accurate range (pass tendency being the best thing to use for the former, and a combination of stamina, natural fitness, CA, and off/def roles for the latter.) Once you've got that, you can see who's doing too little or too much of what, whether it's scoring, hitting, taking penalties, or whatever. From there just adjust the logical attribute(s) that drive the number in question. Then run another season, check the results, and adjust again through as many iterations as you want. Even if you just do it for the NHL regulars and one or two passes through them, you should get significant improvements in the realism of the stats - in the short term, at least.

But the one big impediment I see to getting good long-term results is the attribute inflation that's happened from version to version and then in user-modified db's. After running the game for a while, there winds up being far too many players at the higher reaches of CA and it's just not a realistic environment to model anything in. The only solution I can see to that is a massive re-rate that pushes most CA's and PA's lower, but that's obviously a huge project.

Anyhow, as I promised, I will post hopefully tomorrow a more detailed explanation of how I went about re-rating the NHL. Hopefully this will enable us to then get to work on implementing all of the feedback received thus far and encouraging more users to post further feedback.

I just wanted to post this now so that users don't think I've abandoned the NHL re-rating because of the general disappointment - it's just that I've been away for a few days visiting family.

In the mean time, please keep the feedback coming.

[CJ]

I might be able to contribute in the future. Right now I have too much work on my own.

[ForNever]

As reading this thread through and with the Finnish leagues in the state they are in I just couldn't avoid finally registering. I'm the person whose list was posted on page 2 by my friend lapaz (because I didn't have an account then). I would've added them straight in the rosters but I now couldn't because of the updater glitch. Also I did have a part of the updates with 5.x rosters like lapaz stated, mainly updating my favourite team SaiPa (last season we had several updaters for just the Finnish Elite League so that's why only that team) + creating a bunch of juniors before Lazion created pretty much all of them which was pretty awesome. Though I didn't really add anyone straight to the rosters, just used sheets and sent them to either Koikkari or Lazion (they may not be with the same nicknames here as they are in EHM Finland, nor am I). Both of those guys have quit updating the rosters if I'm not wholly incorrect so that's why Finnish leagues are kind of undone and a part of the transactions are wrong/not updated. I may have talked something about the rosters with Manimal through Windows Live back then but I can't remember certainly anymore.

Anyways I would like to help updating the Finnish Elite League as it is in a bad state while still being a big European League. In my personal opinion I have knowledge enough to update anything but if people (like jhcjobpb above) are helping I of course do not want to overpower anyone. Actually I've kept updating Saimaan Pallo Lappeenranta from Finland for myself on a regular basis so already 1/14 done. No but seriously, at least CA and PA updates are really needed for the league and could be easily done given the time. As any help is needed I would want to lend a hand on that.

About the ''weakened'' NHL, I do not know how would that impact updating other leagues. Mainly every league should be downgraded but that's one hell of a job to do and I do not even know if that's possible. But for example in the Finnish Elite League some of the best goalies like Jussi Rynnäs and Leland Irving now have attributes that make them some of the worst goalies in the league (on goalie attributes, their CAs seem about right, like the situation is also with the NHL). So, I do not really know should they be bettered or the opposite. I do not know (yet at least) why are their goalie attributes between 7 and 9... Best Finnish Elite League goalies have those between 12 and 15 anyways (because they are not updated), which, with the new NHL back-up goalie attributes, seems a bit too much. The problem is like I stated: either every league has to be weakened like this or otherwise we can just change players like Rynnäs back to yellow attributes.

I do not really know how their CAs affect their attributes as the game goes on. I guess the attributes should grow rapidly but I guess I might be wrong. While waiting on Archi's explanation I have to thank for the big work done once again. I'm quite sad about the editor glitch but I understand the situation and think I can wait.

PS. I use this nickname hardly anywhere so if you're from EHM Finland and want to know who I am there, just PM me, though my favourite team should already give that away.

[nino33]

I had hoped that things would at least be along the right lines, but the general feedback seems to be that they are simply not good enough. I'm sincerely sorry for that and it's clearly entirely my fault.

IMO you have absolutely no need to apologize Archi!
Even if there are many dozens of players with significant ratings issues, that's still a very small fraction of the work you did!

Your EHM Updater has done so much to make editing EHM07 easier!

I am for sure glad there is growing interest in improving the TBL Rosters roster updates, and hope the number of volunteers grows...but I just don't think Archi has anything to apologize for (and wanted to say so publicly)



P.S. I'm still not convinced that the database has been "downgraded/lowered" much

Average NHL CA before was 131, average NHL CA after is 132
There were 30 NHLers with a CA 180-200 before, and now there are 23 NHLers with a CA 180-200
There was 66 NHLers with a CA 160-179 before, and now there are 67 NHLers with a CA 160-179
There was 170 NHLers with a CA 140-159 before, and now there are 185 NHLers with a CA 140-159

[archibalduk]

I've written up an explanation, but need to add some tables to it and check some stats from the original DB. It'll likely be tomorrow that I'll post it online.

[Chopper]

Long time lurker here. Just wanted to log on and say the work that gets put into these rosters year in year out is really fantastic and appreciated. I've been playing this game for years now and love downloading the new rosters when they come out.
I'd love to help - I'm not much of a spreadsheet guy and don't have much of an idea of how the game works under the hood. I've never used the editor or anything like that. I watch and follow the league pretty closely though (Calgary guy). So I can offer some more qualitative observations. I don't know if that will be helpful for you guys.
Anyway, just wanted to say thanks for keeping this game relevant and fun!

[Peter_Doherty]

Long time lurker here. Just wanted to log on and say the work that gets put into these rosters year in year out is really fantastic and appreciated. I've been playing this game for years now and love downloading the new rosters when they come out.
I'd love to help - I'm not much of a spreadsheet guy and don't have much of an idea of how the game works under the hood. I've never used the editor or anything like that. I watch and follow the league pretty closely though (Calgary guy). So I can offer some more qualitative observations. I don't know if that will be helpful for you guys.
Anyway, just wanted to say thanks for keeping this game relevant and fun!

I'm the same as you, never really used editors and stuff. I wrote my input about the team i felt i had alot of knowledge about in a pm and sent to manimal. I think it was appreciated, so thats one way to help a bit atleast

[Manimal]

I'm the same as you, never really used editors and stuff. I wrote my input about the team i felt i had alot of knowledge about in a pm and sent to manimal. I think it was appreciated, so thats one way to help a bit atleast

Yeah, what you did was great!
If anyone have anything they want edited post here or PM me. We are open for negotiation and all hekp is appreciated!

Anyways I would like to help updating the Finnish Elite League as it is in a bad state while still being a big European League. In my personal opinion I have knowledge enough to update anything but if people (like jhcjobpb above) are helping I of course do not want to overpower anyone. Actually I've kept updating Saimaan Pallo Lappeenranta from Finland for myself on a regular basis so already 1/14 done. No but seriously, at least CA and PA updates are really needed for the league and could be easily done given the time. As any help is needed I would want to lend a hand on that.

About the ''weakened'' NHL, I do not know how would that impact updating other leagues. Mainly every league should be downgraded but that's one hell of a job to do and I do not even know if that's possible. But for example in the Finnish Elite League some of the best goalies like Jussi Rynnäs and Leland Irving now have attributes that make them some of the worst goalies in the league (on goalie attributes, their CAs seem about right, like the situation is also with the NHL). So, I do not really know should they be bettered or the opposite. I do not know (yet at least) why are their goalie attributes between 7 and 9... Best Finnish Elite League goalies have those between 12 and 15 anyways (because they are not updated), which, with the new NHL back-up goalie attributes, seems a bit too much. The problem is like I stated: either every league has to be weakened like this or otherwise we can just change players like Rynnäs back to yellow attributes.

I do not really know how their CAs affect their attributes as the game goes on. I guess the attributes should grow rapidly but I guess I might be wrong. While waiting on Archi's explanation I have to thank for the big work done once again. I'm quite sad about the editor glitch but I understand the situation and think I can wait.

PS. I use this nickname hardly anywhere so if you're from EHM Finland and want to know who I am there, just PM me, though my favourite team should already give that away.

I would very much appreciate any help on the Finnish leagues!
Drop me a PM and we'll talk some more

[pantsukki]

Could something be done about the NHL teams' finances? It seems that most (maybe even all) teams run heavily into debt during the season, and constantly require investments. Is it possible to edit the amount of TV money which is currently only 2m $?

[Manimal]

Could something be done about the NHL teams' finances? It seems that most (maybe even all) teams run heavily into debt during the season, and constantly require investments. Is it possible to edit the amount of TV money which is currently only 2m $?

The TV money is not editable but I'll have a look at the team's budgets and cash at start up to see what needs to be done

[archibalduk]

As promised, here's some detail on how I updated the NHL for the TBL Rosters v6.0. I have added headings to try and break things up a little. This all follows on from my earlier post here: http://www.ehmtheblueline.com/forums/vi ... 30#p163730

Note that due to the time it took to write this and to draw up the tables, I haven't yet had time to proof-read it yet. I'll proof-read it when I have a spare moment (and if I think of anything additional, I'll update this post).


SCOPE AND BASIS OF THE UPDATE
I started the research back in around April or May 2013. Most of the work was done prior to July 2013 and I finished it off after the summer, in between moving house and a couple of FHM 2013 deadlines I had.

I used the following as a basis for the update:

• The existing TBL Rosters v5.3d ratings
• 2011/12 and 2012/13 regular season stats
• Depth charts from various online sources

The update covered 962 NHL players and free agents who played in the NHL in 2012/13 or who didn't play that season but played in 2011/12. This covered virtually all players on an NHL roster in the v5.3d DB, other than those who had not played an NHL game. This consisted of:

• 84 Goalies
• 301 Defencemen
• 343 Wingers
• 234 Centres

In terms of testing the update, we were pretty much testing this right up to the release date. Most of my testing concentrated on getting the scoring, PIM and fighting levels more realistic. This involving using the information from bobmcgoo's Realism Patch thread and then experimenting with various attributes to see how whether some form of realism patch could be implemented into the DB without the extensive changes involved in bobmcgoo's patch. The work on improving the realism of the game took 2 or 3 months to complete.

As I mentioned in my previous post, trying to update 950+ players on my own was pretty tough. When testing my update, I could only review things as more of a general overview kind of way. So it was checking that the correct people were at the top of the stats tables for each team and that teams were performing reasonably well. Unfortunately there are a number of teams who were still not performing very realistically at the time of release (e.g. Colorado, Toronto and Carolina). But we agreed that the v6.0 DB would be released by Christmas regardless of how things were because it had been 8 months since the last DB and everybody was understandably anxious for the first 2013/14 update of the season. The v6.0 should really be regarded as a work in progress; there is still much to be done (and as I've said before, we had fewer researchers for this version and several of us who have helped haven't had as much free time as on previous versions. The NHL update was also tested by another researcher (I don't know who) and their feedback was that it seemed good.

Manimal also went through my updated CAs and changed some of them (Brooks Orpik was also supposed to have been downgraded, but his amended CA was inadvertantly missed from the list imported into the DB).

With regards to NHL rookies, I didn't update any of these. As mentioned above, my update was based on those who had played last season. Manimal updated the CAs for a number of rookies, but didn't have time to work on their attributes for v6.0.


CURRENT ABILITY
One of the key aims of the update was to get the CAs more in line with how they ought to be in EHM. As mentioned in my previous post, the NHL ratings are higher than what EHM had been designed for and NHL players were generally too perfect.

The NHL CAs should be rated like this in EHM 2007:

[table][tr][th]Descriptor[/th][th]CA[/th][/tr]
[tr][td]Top[/td][td]190[/td][/tr]
[tr][td]Good[/td][td]160[/td][/tr]
[tr][td]Average[/td][td]120[/td][/tr]
[tr][td]Poor[/td][td]100[/td][/tr][/table]

Although the CAs should range from 100 - 190, the bulk of the players should cluster around 120 because this is what the CA should be for the average NHL player. Additionally, if you calculate the average CA of all of the players in the NHL, it should be 120. Therefore it follows that there should be fewer players at towards the top end of CAs and more clustered around the Average.

Based on the above, I calculated the CA range of each of the four categories as per the table below. Obviously there could be some minor variations on how you could calculate these ranges. These ranges are just for comparing the CA ranges between DB updates and not for setting the CAs themselves.

[table][tr][th]Descriptor[/th][th]CA Range[/th][/tr]
[tr][td]Top[/td][td]180 - 200[/td][/tr]
[tr][td]Good[/td][td]140 - 179[/td][/tr]
[tr][td]Average[/td][td]110 - 139[/td][/tr]
[tr][td]Poor[/td][td]100 - 109[/td][/tr][/table]

The table below shows how the DB v6.0 compares with the previous TBL DB and the original v3.0.4 DB (i.e. the final DB released by SI). The data sample is based on players contracted to an NHL team and with a CA of at least 100. I have made the assumption that players with a CA less than 100 are not of NHL quality and therefore should not be considered. Such players are typically playing in the juniors or minors (e.g. the AHL/ECHL and the Canadian Major Juniors). My update has concentrated on those who have played in the NHL and are therefore of NHL quality; and should thus have CAs in the 100 - 190 range.

Note that the percentage figures below are rounded to one decimal place. Thus the figures might not quite add up to 100%.

[table][tr][th]Descriptor[/th][th]CA Range[/th][th]Original v3.0.4[/th][th]TBL DB v5.3d[/th][th]TBL DB v6.0[/th][/tr]
[tr][td]Top[/td][td]180 - 200[/td][td]1.6%[/td][td]1.3%[/td][td]2.5%[/td][/tr]
[tr][td]Good[/td][td]140 - 179[/td][td]25.5%[/td][td]34.2%[/td][td]27.0%[/td][/tr]
[tr][td]Average[/td][td]110 - 139[/td][td]51.5%[/td][td]52.3%[/td][td]64.2%[/td][/tr]
[tr][td]Poor[/td][td]100 - 109[/td][td]21.4%[/td][td]12.3%[/td][td]6.3%[/td][/tr]
[tr][td]AVERAGE CA[/td][td]-[/td][td]127[/td][td]132[/td][td]130[/td][/tr][/table]

Contrary to what some people have said, the NHL hasn't really been downgraded at all. In some respects it has in fact been upgraded albeit the average CA is now 1.5% lower than the v5.3d DB - but it is still not as low as the original EHM rosters (127) and is not as low as the ideal average (120).

The reason for leaving things this way is because I had originally planned to release the NHL update as a separate update for "beta" testing and feedback. However, because of various things that took up my time (e.g. moving house and working on the realism patch), I never had the opportunity to release it in advance of the TBL DB update.

The plan was that, on receiving feedback, we could decide which players need dropping down from Top to Good, Good to Average, etc. By dropping down these players, we'd have a spread of talent which more or less reflects the the original 3.0.4 spread whilst also getting much closer to the 120 average. We have already received quite a lot of feedback and hopefully this will help us adjust the CAs so we're more in line with how it ought to be - but we of course need more feedback because we have 30 teams to cover. Perhaps a starting point for finalising the CAs would be to identify those players who are overrated and drop their CAs down.

One final note: It's worth keeping in mind that I used steps of 5 when rating the CAs. So whereas a player may be have been given a 107 or a 117 CA under previous updates, I would have given him a 110 or 120. It just made things a little easier to manage whilst also giving some freedom to tweak things later on following feedback. At the top of the ratings, I capped CAs at 187 rather than 190, but these players could always be increased to 190 if needs be.


POTENTIAL ABILITY
The only work on PAs I did was to cap older players' PAs according to their CAs. In other words, to prevent older players improving any further.

I didn't make any other adjustments to PAs seeing as the CAs still needed more work.


RE-RATING/RE-DISTRIBUTING THE ATTRIBUTES (PART 1)
As mentioned when I announced the release of the v6.0 DB, I made two types of re-rating to the NHL: Re-rating and re-distributing. Here's what each means:


1) Re-Rating
Completely re-rating a particular attribute, based on the existing rating as well as the 2011/12 and 2012/13 stats. For example, the Wristshot attributes were based on the existing Wrist attributes along with goals per ATOI and shot quality (I used either Corsi or Fenwick - I can't remember which).

Other examples include Aggression which was based mostly on PIM per ATOI, Fighting which was based mostly on fight win% and total fights and Dirtiness which was based partly on number/length of suspensions and number of major penalties (amongst other things).

Another example is Faceoffs which was based mostly on faceoff win% and total faceoffs. The range of ratings I used when re-rating this attribute was higher for centres than wingers and higher for wingers than defencemen.


2) Re-Distributing
This is where I maintaing the order of the existing ratings but redistributed the ratings in order to properly reflect how these ratings should be distributed in EHM (and in fact we do something very similar for FHM). This is actually how I did most of the work for the NHL update. I.e. most of the attributes I adjusted we simply re-distributed rather than re-rated.

I don't think I've explained this well and so I will use an example: One attribute I re-distributed was Pace (aka Speed). In order to re-distribute, I ranked all of the players by their existing Pace attribute.

Obviously a lot of players have the same existing Pace rating and so I had to sub-sort these players in order to properly rank the players. For Pace, I used a number of factors to sub-sort equally rated players; such as age, CA and how good their present team is. Each of these factors had a different weight to the sub-sorting (age had the biggest weighting whereas how good their team is had a very minute weighting).

Although the sub-sorting would affect whether the player is ranked higher or lower, it would only in reality have an effect of raising/lowering the final attribute by around about 1 point and so it didn't need to be a precise science IMO.

Having sub-sorted the players so that they were ranked from first to last, I then adjusted players down slightly if they were old. This was done on the assumption that the NHL players in DB haven't been aged much over the years. For example, in the v5.3d DB Marian Gaborik still has the 18 Pace rating he had from the v3.0.4 EHM DB. There's no denying he's still a fast skater, but I'd argue he's not as quick aged 31 as he was in 2006/07 at age 25. Thus for a number of physical attributes, I reduced older players down the ranks to reflect ageing.

Once I had the players ranked, I then assigned their new ratings. The range of ratings I used depended on whether the attribute in question is a Technical or a Non-Technical attribute (in the case of Pace, it is a Non-Technical attribute). I will explain the difference in a moment (see below), but what I'm trying to explain here is that when re-distributing attributes, the actual order of where players are (in terms of who is best and who is worst) is the same or very similar to the order in the v6.0 DB. With the exception of those players I have aged, there shouldn't be all that much difference between the v5.3d DB and the v6.0 DB in terms of who is best and worst in re-distributed attributes (which formed the vast majority of the updating work I did).

There are however two caveats I want to explain (well, one caveat and one more of a note):

1. I did the re-distributing on a position by position basis. Thus I re-distributed the goalies, defencemen and forwards separately. I did this on the assumption/argument that there should be an equal spread of talent between positions in relation to attributes. This prevents us having the situation where all of the forwards are faster than the defencemen, etc.

2. Strictly-speaking I didn't quite use a ranking system, but more of a points system which took into account the various weighted sub-sorting I did. But it's easier to explain it as a ranking system for present purposes.


RE-RATING/RE-DISTRIBUTING THE ATTRIBUTES (PART 2)
Now that I've explained the difference between re-rating and re-distributing, I just wanted to make some final comments:

When re-rating the players, I followed a very similar methodology to re-distributing. It's just that when re-rating, I didn't use the existing attributes as the main starting point for ranking the players. But as with the re-distributing, I re-rated the attributes on a position by position basis (i.e. the goalies, defencemen and forwards were re-rated separately) in order to maintain an appropriate spread of talent.

Another reason for doing things on a position by position basis is because obviously certain attributes should be rated more highly for certain positions. For example, Faceoff attributes should be typically higher for forwards than defencemen. The Faceoff attribute was also rated more highly for centres than wingers.


THE TWO TYPES OF ATTRIBUTE: TECHNICAL AND NON-TECHNICAL
Now that I've explained how the attributes were re-rated/redistributed, I'll explain about the two types of attribute.

How the players are rated for a particular depends on whether the attribute is Technical or Non-Technical. See the two headings below for more detail.

Something I found when working on the NHL goalies is that their non-goaltending ratings were highly overrated and didn't follow the correct guidelines. I have re-worked many of these attributes so that they correctly follow the distribution designed for EHM. I didn't however alter goalies' Pokechecking to the correct level because it seemed to undo the effects of the realism patch.


1) Non-Technical Attributes
Regardless of the CA of the player, the full 1-20 range should be used. Thee attributes are NOT related to CA.

For any league in the game, there should be a good spread/distribution of Non-Technical attributes. Most players should be clustered around the middle (i.e. the most common ratings should be 10s and 11s) and few, if any, players should be at the extremes (i.e. 1 and 20). In theory, if you take the average rating of any Non-Technical attribute for all players in any given league, the average should be approximately 10.5.

There are some exceptions to this rule:

1. Some leagues may have certain characteristics which merit a slightly different distribution. E.g. the LNAH would typically have higher ratings for Aggression.

2. For many Non-Technical attributes, goalies should have a different distribution. E.g. goalies should have an average-to-low Acceleration rating and an average-to-high Agility rating.

This is how the Non-Technical attributes are supposed to be set in EHM and this is how I did it for the NHL update. As mentioned earlier, most of the work on the Non-Technical attributes involved re-distributing rather than re-rating.

Below is a list of all Non-Technical attributes in EHM (those with an asterix indicate those that do not quite follow the standard 1 - 20 range). I have put in bold underline those which I updated. Note that for Consistency and Temperament these were adjusted en masse as part of the realism patch work.

[table][tr][td] Acceleration [/td] [td]Important Matches[/td] [td] Stamina [/td][/tr]
[tr][td]Adaptability[/td] [td]Injury Proneness*[/td] [td] Strength [/td][/tr]
[tr][td] Aggression [/td] [td]Leadership/Influence[/td] [td]Teamwork[/td][/tr]
[tr][td] Agility [/td] [td]Loyalty[/td] [td] Temperament [/td][/tr]
[tr][td]Ambition[/td] [td] Natural Fitness [/td] [td]Versatility[/td][/tr]
[tr][td] Bravery [/td] [td] Pace/Speed [/td] [td][/td][/tr]
[tr][td] Consistency* [/td] [td] Pass Tendency* [/td] [td][/td][/tr]
[tr][td]Determination[/td] [td]Pressure[/td] [td][/td][/tr]
[tr][td] Dirtiness [/td] [td]Professionalism[/td] [td][/td][/tr]
[tr][td] Flair [/td] [td]Sportsmanship[/td] [td][/td][/tr][/table]

2) Technical Attributes
These attributes ARE related to CA.

However, contrary what was mentioned once or twice earlier in this thread, NHL players shouldn't have all Technical attributes of at least 10. That's not how Technical attributes work in EHM. If you take that approach, NHL players will appear too perfect (and this was in fact an issue with EHM 2004 and 2005 which resulted in a change in the way the CA and Technical attribute systems work in EHM 2007).

For any given player, the average of his Technical attributes relates to his CA. However, there should be some variation of +/-6 attribute points for each player. Following on from the CA table I posted earlier on, here is how the Technical attributes should be rated as per the four key CA levels:

[table][tr][th]Descriptor[/th][th]Minimum[/th][th]Average[/th][th]Maximum[/th][/tr]
[tr][td]Top[/td][td]190[/td][td]7[/td][td]13.5[/td][td]19[/td][/tr]
[tr][td]Good[/td][td]160[/td][td]6[/td][td]12[/td][td]18[/td][/tr]
[tr][td]Average[/td][td]120[/td][td]4[/td][td]10[/td][td]16[/td][/tr]
[tr][td]Poor[/td][td]100[/td][td]3[/td][td]9[/td][td]15[/td][/tr][/table]

For example, the average NHL player should have a 120 CA. This means that his Technical attributes should average 10, but each individual Technical attribute can range from 4 - 16.

This is the approximate guidelines EHM will use to adjust a player's Technical attributes if you try to under-rate or over-rate a player. If you create a player using the Editor and give him a CA of 120 and all Technical attributes set to 20, you'll see that when you start a new game in EHM his Technical attributes will be greatly reduced and will be line with the rules above.

NHL players aren't perfect and they should therefore have a range of Technical attributes. They shouldn't all be rated 10+ otherwise they'll be over-rated and too "perfect". When re-rating and re-distributing Technical attributes, I have tried to adhere to these rules by using a range that is more appropriate according to the player's position and according to the range of attributes that should be expected according the NHL range of CAs. Typically I have used a 5 - 19 range with the most common ratings being in the 9 - 12 range (which follows the sort average Technical attributes we should expect for most NHL players - as per the table above).

The range used for goalies depended on the type of Technical attribute being rated. For goalie-specific attributes (e.g. Glove and Blocker), a 7 - 20 range was used (with 7 - 12 being the most common ratings). Slightly lowered ranges were used for non-goalie-specific Technical attributes (e.g. Passing).

Below is a list of all Technical attributes in EHM. I have put in bold underline those which I updated. You'll see I haven't re-rated/re-distributed every Technical attribute category. Unfortunately I ran out of time.

[table][tr][td]Agitation[/td] [td] Fighting [/td] [td] Reflexes [/td][/tr]
[tr][td] Anticipation [/td] [td] Glove [/td] [td]Slapshot[/td][/tr]
[tr][td] Balance/Skating [/td] [td]Hitting[/td] [td] Stickhandling [/td][/tr]
[tr][td] Blocker [/td] [td]Movement[/td] [td]Work Rate[/td][/tr]
[tr][td] Checking [/td] [td] One-on-Ones [/td] [td] Wristshot [/td][/tr]
[tr][td] Creativity/Vision [/td] [td] Passing [/td] [td][/td][/tr]
[tr][td]Decisions[/td] [td] Pokecheck [/td] [td][/td][/tr]
[tr][td]Deflections[/td] [td] Positioning [/td] [td][/td][/tr]
[tr][td]Deking[/td] [td] Rebounds [/td] [td][/td][/tr]
[tr][td] Faceoffs [/td] [td] Recovery [/td] [td][/td][/tr][/table]

[Peter_Doherty]

Great job overall archibalduk and a nice read Seems like this patch woke the whole community up to try and help the best they can aswell, so the re-rating should become really realistic soon.

Edit; please check your pm and answer me when you have a minute over

[jesterx7769]

One of the things I have noticed is that aggression seems almost too low but I admit I am not sure how that plays out when fully tested as I usually use aggression to help my teams toughness and I know it was to try to lower PIM

[CJ]

I could help with updating this database on a later stage. Right now I'm too busy with all other stuff.

[Mr.Pickles]

I've started posting in the threads that nino has been making but needed to post here to say thanks for all the work that everyone has put into these rosters.

I'm a long time EHM player and have lurked here for years. It's amazing to see that the game playing experience actually continues to get better years after development ended because of all the work you guys put in. Thanks.

I'd be happy to help with any re-rating work that needs to be done. I follow the Jets but have a good overall knowledge of the league and prospects. I can take a few teams or a division or whatever to start working on.

[Manimal]

An editable(pre-game editor) version of the rosters will be released within a week.
Some things will be edited between 6.0 and that version but not very much.

[philou21]

The DB looks good! I only checked the Avalanche and the Habs so far though but I think the skating abilities like acceleration, agility and speed and kinda low for alot of guys.

In MTL Galchenyuk is really underated he wouldn't even fit on a fourth line, even Ryan White is better.

Subban slapshot at 14 is low, should be a good 17-18 I think.

Bournival is only 3 in determination and only 9 in hard work. From what I've seen so far of him this season he should be at least 15-15 in both, easily. Acceleration is a little bit low and the right column is weak I think.

Gorges as a nice rating except the without the puck attribute, that's where is is the best. Though I don't know if this is only for good foward.

The Avs now.

Alex Tanguay 10 in wristshot should be a 15-16

O'reilly should be way better.

Giguere is really weak as well. He's not the best goalie ever, especially at his age but could be better.

PA Parenteau skating attributes are low. I never watched him play that much but from what I've heard he's fast on his feet.

Hope it helps! If I see something else I'll let you know. Keep up the good job!

[Manimal]

...and the right column is weak I think.

If the right column is weak doesn't that make it the wrong column?

You are also making a common fault. That is debating specific attributes without looking in the database first.
Most of Bournival's attributes are set at random so they turn up differently each time you start a game.
No big deal but it would save us researchers some time if users checked this first before posting.
Now, you are telling us of where the players can be improved and that will be looked upon by whoever will be updating the forwards.

And what is this 'without the puck' attribute? Never heard of it

Thanks for your input!

[philou21]

You are also making a common fault. That is debating specific attributes without looking in the database first.
Most of Bournival's attributes are set at random so they turn up differently each time you start a game. No big deal but it would save us researchers some time if users checked this first before posting.

I absolutely don't know how you guys work or what you've done so I don't know. It's the first time I install one of your DB so I didn't know it was common that we should check the DB at the same time.

And what is this 'without the puck' attribute? Never heard of it

I don't know how to put it in english sorry. Ok I found it's the "getting open" attribute. Wow the french traduction is goddamn awful it's like totally another meaning.