EHM:EA Attribute testing

[nino33]

I did some testing at the end of July with EHM:EA and shared my results with some of the researchers...I thought I'd share them publicly

EDIT (July 2016) links are now to OneDrive on my new EHMtesting account
Test 1 https://1drv.ms/u/s!AvTHo4MBaphkbRylZwEyHMn26VY
Test 2 https://1drv.ms/u/s!AvTHo4MBaphkbv5lWa3RP6bCe40


I used the TBL Rosters and looked at 16 goalies & 112 skaters (age 13-23 at start-up).
I looked at every Attribute and it's development on a year by year basis for all 128 players.
I did actually did three tests of 16-13-13 years, but (because it takes a lot of time to do) in the end I just compiled the data for the first 12 years for tests 1 and 2

I used the EHM:EA Assistant to export the saved game data, and to compile/display the data I used slightly modified versions of the Attribute Progression Tracker spreadsheet http://www.ehmtheblueline.com/forums/vi ... te+tracker

For each test there are 8 Attribute Tracker Spreadsheets
- one each for Hidden/InGame Atts from 2014-2018
- one each for Hidden/InGame Atts from 2018-2022
- one each for Hidden/InGame Atts from 2022-2026
- one each for Hidden/InGame Atts showing just 2014 Atts and 2026 Atts (showing development over 12 years)

The Attribute Tracker Spreadsheets have a "comparison" tab for both skaters and goalies were you can see the growth/regression.

The initial data (2014) is from right after game start-up.
Each year I looked at the data on June 27th. In addition to all Attributes I included CA and PA too.
I included a few bits of additional information on the 2014-2018 InGame spreadsheet (like Offensive/Defensive Role)

[nino33]

Posting in the researcher forum on July 26th

Over the last couple weeks I've been running 3 test games, using 112 skaters and 16 goalies (age 13-23 at start-up) and I looked at all Attributes on a yearly basis.....I've just finished compiling the first 12 years of data from Text 1 and Text 2, and I've decided I've got enough "data" to answer the questions I wanted to answer.

Essentially I wanted to know
- what Attributes never change
- what Attributes rarely change/only change a little bit when they do change
- what Attributes change "normally"

I've only just this afternoon completed compiling the data, and a few others will be looking at it, so a more detailed breakdown on some Attributes or combination of Attributes is likely forthcoming (for example, at first glance it looks like the skating Attributes develop differently than they did in EHM07...in EHM:EA they seem to develop more "normally").




SKATERS
For skaters the following Attributes are always rated as “1” and never change/grow – One On Ones, Blocker, Glove, Rebound, Recovery and Reflexes.

For skaters, the following Attributes never change, they always maintain the same value – Aggression, Flair, Consistency, Dirtiness, Pass Tendency, Injury Proneness, Agitation and Adaptability.

Ambition, Loyalty, Pressure, Professionalism, Sportsmanship and Temperament are somewhat more likely to change (but still don’t change much).

For skaters, the following Attributes change/develop – Acceleration, Agility, Balance, Speed, Stamina, Strength, Anticipation, Bravery, Creativity, Influence, Teamwork, Work Rate, Checking, Deflections, Deking, Faceoffs, Getting Open, Passing, Pokecheck, Positioning, Slapshot, Stickhandling, Wristshot + Decisions, Important Matches, Fighting, Natural Fitness and Versatility.




GOALIES
For goalies the following Attributes are always rated as “1” and never change/grow – Hitting, Checking, Creativity, Deflections, Deking, Faceoffs, Getting Open, Slapshot and Wristshot.

For goalies, the following Attributes never change, they always maintain the same value – Aggression, Flair, Consistency, Dirtiness, Pass Tendency, Injury Proneness, Agitation and Adaptability.

For goalies Ambition, Loyalty, Pressure, Professionalism, Sportsmanship and Temperament are somewhat more likely to change (but still don’t change much). Fighting is usually a low rating and unlikely to change/doesn’t change much.

For goalies, the following Attributes change/develop – Acceleration, Agility, Balance, Speed, Stamina, Strength, Anticipation, Bravery, Influence, Teamwork, Work Rate, Blocker, Glove, Passing, Pokecheck, Positioning, Rebounds, Recovery, Reflexes, Stickhandling + Decisions, Important Matches, Natural Fitness, Versatility and One On Ones.




For both skaters & goalies Determination rarely change/changes very little.

[nino33]

And then a few hours later...

Ambition, Loyalty, Pressure, Professionalism, Sportsmanship and Temperament are somewhat more likely to change (but still don’t change much).

Some examples, looking at the 29 test players (skaters) age 17 or 18 at start-up in Test 1 (most with a -8 or -9 PA).....and looking at the change over 12 years (from 2014 to 2026)

Ambition - 12 saw no change, 4 went down one, 9 went up one, 3 went up two & 1 player went up three
Loyalty - 12 saw no change, 7 went down one, 1 went down two, 8 went up one & 1 player went up three
Pressure - 18 saw no change, 3 went down one, 3 went down two, 1 went down three, 3 went up one & 1 player went up two
Professionalism - 17 saw no change, 6 went down one, 1 went down two, 1 went down four, 4 went up one
Sportsmanship - 19 saw no change, 2 went down one, 1 went down two, 5 went up one, 2 went up two
Temperament - 16 saw no change, 6 went down one, 2 went down two, 1 went down three, 3 went up one, 1 went up two


And continuing on, looking at the same 29 players...

For skaters, the following Attributes change/develop – Acceleration, Agility, Balance, Speed, Stamina, Strength, Anticipation, Bravery, Creativity, Influence, Teamwork, Work Rate, Checking, Deflections, Deking, Faceoffs, Getting Open, Passing, Pokecheck, Positioning, Slapshot, Stickhandling, Wristshot + Decisions, Important Matches, Fighting, Natural Fitness and Versatility.

Acceleration - all but 3 players saw change, 1 player went down, 25 players went up (average increase/growth of 2.72, range of +1 to +7)
Agility - all but 2 players saw change/growth (average increase/growth of 2.22, range of +1 to +3)
Balance - all but 5 players saw change/growth (average increase/growth of 3.33, range of +1 to +5)
Speed - all but 3 players saw change, 1 player went down, 25 players went up (average increase/growth of 3.00, range of +1 to +6)
Stamina - everyone saw an increase (average increase/growth of 5.24, range of +3 to +7)
Strength - all but 5 players saw change/growth (average increase/growth 4.39, range +1 to +9)
Anticipation - all but 2 players saw an increase (average increase/growth 2.96, range +1 to +6)
Bravery - all but 7 players saw change, 3 players went down, 19 players saw an increase (average increase/growth 1.26, range +1 or +2)
Creativity - all but 3 players saw an increase (average increase/growth 2.58, range +1 to +5)
Influence - everyone saw an increase (average increase/growth of 4.38, range of +1 to +6)
Teamwork - all but 3 players saw change, 1 player went down, 25 players saw an increase (average increase/growth 1.60, range +1 to +3)
Work Rate - all but 4 players saw change, 3 players went down, 22 players saw an increase (average increase/growth 1.23, range +1 or +2)

Checking - all but 4 players saw an increase (average increase/growth 3.04, range +1 to +8)
Deflections - everyone saw an increase (average increase/growth of 4.34, range of +1 to +8)
Deking - all but 2 players saw an increase (average increase/growth 3.56, range +1 to +9)
Faceoffs - all but 5 players saw an increase (average increase/growth 4.25, range +1 to +10)
Getting Open - everyone saw an increase (average increase/growth of 4.48, range of +1 to +9)
Passing - everyone saw an increase (average increase/growth of 5.03, range of +1 to +8)
Pokecheck - all but 4 players saw change, 1 players went down, 24 players saw an increase (average increase/growth 3.00, range +1 to +8)
Positioning - everyone saw an increase (average increase/growth of 5.86, range of +3 to +9)
Slapshot - all but 2 players saw an increase (average increase/growth of 4.15, range of +1 to +8)
Stickhandling - everyone saw an increase (average increase/growth of 4.31, range of +1 to +8)
Wristshot - everyone saw an increase (average increase/growth of 4.00, range of +1 to +8)

Decisions - all but 3 players saw an increase (average increase/growth of 2.38, range of +1 to +5)
Important Matches - all but 4 players saw an increase (average increase/growth of 3.04, range of +1 to +10)
Fighting - all but 3 players saw change, 1 players went down, 24 players saw an increase (average increase/growth 2.28, range +1 to +6)
Natural Fitness - all but 4 players saw an increase (average increase/growth of 2.00, range of +1 to +4)
Versatility - everyone saw an increase (average increase/growth of 4.86, range of +1 to +6)


P.S. In this case the 29 skaters were made up of 7 defenseman and 22 forwards. Ideally such testing/review would be done looking at just defensemen, or just wingers, or just centermen.....I also think, in addition to Position, the Player Role has an impact on what Attributes develop/how much

But the above still gives a reasonable indication of how Attributes that change do so.....based on the Attribute the lower/higher ranges are indicative of the different Positions/Player Roles. I think if you looked at all defensive D and compared them with all pointman D you'd see the patterns even clearer. Another such example would be comparing all Power Forwards with all Playmakers, etc etc


EDIT - Looks like I forgot to include Hitting in the post/data above

[Tasku]

This is my first attempt ever at sharing files, hope it works...
Test 1 http://www.megafileupload.com/lQd3/EHME ... Test_1.zip
Test 2 http://www.megafileupload.com/lQd1/EHME ... Test_2.zip
Apparently they'll be available to download for a week, so get 'em while you can if you're interested

I suggest using Google Drive or some other Cloud Service, where the files will remain indefintively. If you have a Google email address (gmail), you can use the same account while uploading files to Google Drive.

[deknegt]

TL;DR for anyone who doesn't like reading really interesting stuff about attributes. (really, read it, it's pretty interesting!)

Technical attributes improve the most over a player's career.
Physical attributes improve pretty quickly too
mental attributes vary wildly between decent improvements, to little improvement, to none at all. This reinforces the fact that most intangibles are nature and only really get better through wisdom and experience.

---

My take on the test:

What this picture also paints that regardless of talent a player has, they need very high base attributes to be actually able to become a master at a skill. The highest improvement was +10 in that whole test, and that one were faceoffs and the hidden attribute ''important matches'', the latter one makes sense due to players becoming more cool as they gain experience. The former also makes sense in that same respect. But that also means a player must start out with a 10 for Faceoffs to even have a chance to reach a perfect 20 in faceoffs.

Naturally perfection is unrealistic for every player but Wayne Gretzky, really. And I wouldn't expect the best to have a clean 20 for skills. But that poses an issue due to newly generated players being 13/14 years old, since no player starts out with technical, mental, or physical skills at such a high level unless it's a ''personality'' skill e.g. determination, work rate, and most hiddens.

What it means that many if not all players wont be able to reach a high level of attributes, which explains the issues many people have been having with long-term games of EHM:EA where the talent level gets lower and lower, it's simply the game regressing to it's mean and that average is a league filled with little to no ''superstar'' level players.
The issue also isn't as easily fixed as simply increasing start-up skills, because that means players are simply too strong for the level they're supposed to be playing at... A better solution might be tweaking attribute growth levels for U21 players, with play in the CHL and other junior leagues having a distinct effect on player growth.

Naturally, the test is inherently flawed due to only running for 12 seasons, and having a bit too broad a focus including players from age 13 to 23, so it's not SUPER accurate. And it might also be better to run more focused (and longer) tests in the future. But what Nino rightfully said, it takes a lot of time to simulate such tests.
For parity we'd need tests that have a big batch of players starting at age 13/14, and then viewing their progress over their entire careers (Talking bout 27-30 years roughly speaking). Growth, Peak, Regression, etc.

But that'll take lots and lots of time and effort, so I can't blame for Nino or others to undertake that endeavor.

Regardless of the test being flawed, lots of interesting information can be taken from this test. And even with it's broad spectrum of players used, we can safely say that attribute changes are a bit... on the low side. Or at the very least, it needs to be front-loaded more and slow down considerably as they move into their mid-20s

[archibalduk]

What this picture also paints that regardless of talent a player has, they need very high base attributes to be actually able to become a master at a skill. The highest improvement was +10 in that whole test, and that one were faceoffs and the hidden attribute ''important matches'', the latter one makes sense due to players becoming more cool as they gain experience. The former also makes sense in that same respect. But that also means a player must start out with a 10 for Faceoffs to even have a chance to reach a perfect 20 in faceoffs.

Naturally perfection is unrealistic for every player but Wayne Gretzky, really. And I wouldn't expect the best to have a clean 20 for skills. But that poses an issue due to newly generated players being 13/14 years old, since no player starts out with technical, mental, or physical skills at such a high level unless it's a ''personality'' skill e.g. determination, work rate, and most hiddens.

What it means that many if not all players wont be able to reach a high level of attributes, which explains the issues many people have been having with long-term games of EHM:EA where the talent level gets lower and lower, it's simply the game regressing to it's mean and that average is a league filled with little to no ''superstar'' level players.

This is a really interesting point and something we should probably flag with Riz (indeed I will do this now). I think this is something that can only be properly addressed by adjusting the game rather than adjusting the DB.

It goes to show how useful these sorts of long term tests are. Thanks Nino and deknegt!

[nino33]

I'd sure like some actual data rather than just anecdotal evidence about the possible issue of low quality regens
Despite the existence of the EHM:EA Assistant I've noticed no one seems to have ever looked up some of the top NHL Draftees to check PA or used the EHM:EA Assistant to give average NHL CA levels

Having done testing like this for a number of years I can say that sometimes the actual results don't support the anecdotal beliefs at all
I'm not saying a lowering of the talent is not true, but I wish I had more than anecdotal evidence...

If it is true another possibility is the talent level in the TBL database is to high, and people are thinking normal talent distribution is to low (I had significant push back regarding the lower talent level in the 1974 DB)

When I read "regardless of talent a player has, they need very high base attributes to be actually able to become a master at a skill" I immediately wonder what you think as mastery...for me 12-14 is super talented/professional, 15-17 is mastery level/superstar, 18-20 is elite/generational - I've always thought/said players ratings are to high and they get way to many ratings in the 19-20 range

At any one time I think worldwide there should only be maybe 6-12 players with 20 in any given Attribute
And maybe a few dozen in the 18-20 range, and the rest 17 and under.....


I actually didn't run the tests to actually look at player/Attribute development, but rather to determine some basic facts for editing purposes.....I really just wanted to know what Attributes never change, what Attributes rarely change/only change a little bit when they do change & what Attributes change "normally"

A better solution might be tweaking attribute growth levels for U21 players, with play in the CHL and other junior leagues having a distinct effect on player growth.

I suggested something like this before.....if a player plays Major Junior hockey successfully for a reasonable length of time (40 games? 50 games? one season?) his Attributes should raise up to a "minimum" based on this (there IMO should never be players in Major Junior with ratings of 1-3 or 1-5 in basic skating/hockey skills)

[nino33]

I'm trying to look into the poor regens issue to see if I can find some statistical support, and thus far I think I've not found it...



I'm looking at Test 1 in the year 2030 (16 years after start-up), and looking at the top 600 CA players in the NHL in 2030...

• average CA is 144.5
  21 players with a 180-189 CA
  33 players with a 170-179 CA
  63 players with a 160-169 CA
  108 players with a 150-159 CA
  129 players with a 140-149 CA
  102 players with a 130-139 CA

At first glance it doesn't seem "low quality" to me.....



Looking at (553) individual skater Attributes...

• average of Aggression 7.5, range of 1-20, 37 players 15-20 (7 are 19-20)
  average of Anticipation 13.8, range of 3-20, 293 players 15-20 (21 are 19-20)
  average of Bravery 12.6, range of 2-20, 165 players 15-20 (31 are 19-20)
  average of Consistency 11.7, range of 1-20, 132 players 15-20 (12 are 19-20)
  average of Decisions 11.9, range of 5-20, 159 players 15-20 (5 are 19-20)
  average of Dirtiness 9.4, range of 1-20, 80 players 15-20 (15 are 19-20)
  average of Agitation 8.8, range of 1-20, 54 players are 15-20 (6 are 19-20)
  
  average of Flair 10.2, range of 1-20, 74 players 15-20 (4 are 19-20)
  average of Important Matches 13.5, range of 1-20, 222 players are 15-20 (29 are 19-20)
  average of Influence 14.8, range of 4-20, 331 players 15-20 (9 are 19-20)
  average of Teamwork 14.5, range of 4-20, 267 players 15-20 (35 are 19-20)
  average of Creativity/Vision 13.6, range of 1-20, 256 players 15-20 (36 are 19-20)
  average of Work Rate 14.8, range of 5-20, 303 players 15-20 (47 are 19-20)
  
  average of Acceleration 12.0, range of 4-20, 146 players 15-20 (32 are 19-20)
  average of Agility 11.1, range of 4-20, 92 players 15-20 (7 are 19-20)
  average of Balance 10.7, range of 4-20, 97 players are 15-20 (17 are 19-20)
  average of Speed 12.4, range of 4-20, 151 players are 15-20 (28 are 19-20)
  
  average of Fighting 6.6, range of 2-20, 37 players are 15-20 (7 are 19-20)
  average of Hitting 9.7, range of 1-20, 125 players are 15-20 (60 are 19-20)
  average of Injury Proneness 8.9, range of 1-20, 57 players are 15-20 (5 are 19-20)
  average of Natural Fitness 12.3, range of 4-20, 131 players are 15-20 (25 are 19-20)
  average of Stamina 15.3, range of 4-20, 358 players are 15-20 (125 are 19-20)
  average of Strength 12.2, range of 4-20, 189 players 15-20 (102 are 19-20)
  
  average of Checking 10.9, range of 1-20, 146 players 15-20 (59 are 19-20)
  average of Deflections 11.8, range of 2-20, 183 players are 15-20 (61 are 19-20)
  average of Deking 12.4, range of 3-20, 187 players 15-20 (80 players are 19-20)
  average of Faceoffs 8.8, range of 1-20, 132 players are 15-20 (68 players are 19-20)
  average of OffThePuck/GettingOpen 13.9, range of 3-20, 274 players are 15-20 (110 are 19-20)
  average of Passing 15.7, range of 7-20, 368 players are 15-20 (156 players are 19-20.....90 players at 20!)
  average of Pokecheck 10.6, range of 2-20, 142 players are 15-20 (50 players are 19-20)
  average of Positioning is 13.4, range of 2-20, 272 players are 15-20 (81 players are 19-20)
  average of Slapshot 12.4, range 2-20, 166 players 15-20 (47 are 19-20)
  average of Stickhandling 13.8, range 5-20, 249 players 15-20 (69 are 19-20)
  average of Wristshot 13.3, range of 2-20, 245 players 15-20 (93 are 19-20)

First glance thoughts.....some of the low ranges connect to them being not connected to the position (i.e. Defenseman and the Faceoff Attribute), but there are cases where that doesn't seem likely (i.e. most every other hockey skill) and yet the ranges are still extremely wide; I definitely think in these cases there should be more 'near the middle" (so those at the bottom need increasing, and some at the top need decreasing)

The different averages of the skating Attributes seems a bit low to me (like Speed/Acceleration, but I'm only thinking a bump of 1 or 2).....the range for all is 4-20, and I think the range should be maybe 8-20 & then most should be nearer the average (so some of the higher will need a bit of a reduction to offset the boost to those with lower ratings)

For the technical/hockey skills IMO there is no way an NHL player should ever have ratings of 2 and 3 and 4 and 5 and...



The biggest thing that stands out to me - IMO the game actually seems "top heavy" for many Attributes, and I think there should be fewer players in the 17-20 range (they'd be lowered) and far fewer players under a 10 (they'd be increased).....but the averages in most cases seem OK at least

But the low ranges are see for all the hockey skills is significantly off IMO (there's no way there's any NHLer with a 2 or a 3 or a 4 in any of those skills.....
No NHLer should have a bantam level shot, skating ability, positioning/etc...nothing should be that low IMO: nothing should be as low as Junior B or Junior A...it's only at a Major Junior level that maybe we see some comparable skills; since EHM:EA simulates Bantam-Midget-Junior B-Junior A below Major Junior, I think the 1-5 or 1-6 range is for those Leagues & while I can see Major Junior and low level pro players have some Attributes below 7, I think in most cases/for most Attributes for NHL and Elite League players such low values should be rare/nonexistent

I'm all for limiting the higher values! I think there are to many players to high in some Attributes (like Passing and the number of players who have a 20).....and I think the bottom end/threshold needs to be raised



Final thoughts...while it is true the data would be way more interesting/helpful if at least split into defenseman and forwards (better yet, by Player Role too)...but I didn't do this to look at the data in general, I did it looking for support for the idea that low quality regens are being produced, and I don't think the data above supports that

[Alessandro]

Great job Nino

[deknegt]

Interesting stuff Nino.
Just a question, how are you parsing your data? Are you going through every NHL player and seeing what their skills are, or are you using something else to find it?
Because if I knew how it worked, I might do a test for the CHL and see if the CHL skill levels might be lower than they should be. It's generally easier to check that due to the maximum player stint is 4-5 years...

If CHL levels are (too) low, and the NHL levels are (too) high, then maybe it means that the development curve is not tuned rightly enough. Sure, most CHL players will never be NHL players, let alone stars. And NHL talent should be around 20 years of age to even be ready for that level (unless it's generational talent), but still it might be interesting to find out how the attribute levels are among CHL players.

[zbguy]

Thanks for the analysis Nino

[nino33]

Just a question, how are you parsing your data? Are you going through every NHL player and seeing what their skills are, or are you using something else to find it?

The EHM:EA Assistant spits out Attributes from a saved game in a spread sheet! Every player in the world (including identifying/bio data) http://www.ehmtheblueline.com/forums/vi ... &start=250

A huge improvement from the days of looking at EHM07 saved game data is the EHM:EA Assistant shows all Attributes in a 1-20 scale

I've kept all my yearly saves from my tests, so today I loaded up test 1 year 2030 and used the EHM:EA Assistant to get the data.....it needs to be cleaned up a little (Short Position doesn't work well, some players have multiple positions/columns and it throws off the data summaries.....there's a team named after a radio station that messes things up too, and a player having Jr. at the end of his last name + there's no League column, so you have to go through the Team Playing For or Team Contracted To column to select the teams you want)

As a cheat tool I'm against the concept of something like the EHM:EA Assistant, but as a research tool it's spectacularly helpful! HaHa

[nino33]

Great job Nino

Thanks!

Thanks for the analysis Nino

You're welcome!

[Alessandro]

As a cheat tool I'm against the concept of something like the EHM:EA Assistant, but as a research tool it's spectacularly helpful! HaHa

Indeed!

[nino33]

FYI I created another thread for Major Junior testing results/comments http://www.ehmtheblueline.com/forums/vi ... 10&t=14811

Also, I started another long term test (as there have been a couple/few updates since I did the testing being referenced thus far).....but it takes many days to sim (as I always sim with all Leagues selected, a Full Database and everything at Full Detail; started yesterday in September 2014, currently twentyone hours later and the sim's in February 2017)

[visualdarkness]

I think you are on to something nino. It is not like the average atts are bad, it is the distribution.

You get regen guys going top five in the draft with something like 17-20 in bravery, hitting and checking but 1-4 in speed, stamina and positiong. Or an offensive winger with 18 in stickhandling, 3 in deking, 17 slapshot, 4 writer, 16 speed and 1 in agility (always the agility for speedy guys). This kind of kills the realism and usability for drafts in the future. Worst case is the goalies where almost all "good" goalies suck in stats like recovery.

[Alessandro]

I think you are on to something nino. It is not like the average atts are bad, it is the distribution.

You get regen guys going top five in the draft with something like 17-20 in bravery, hitting and checking but 1-4 in speed, stamina and positiong. Or an offensive winger with 18 in stickhandling, 3 in deking, 17 slapshot, 4 writer, 16 speed and 1 in agility (always the agility for speedy guys). This kind of kills the realism and usability for drafts in the future. Worst case is the goalies where almost all "good" goalies suck in stats like recovery.

Yes, this is my biggest complaint thus far, it looks like the new system didn't manage to correct the old system's problems. I even have to say that in this field things are worse now (IMHO). But I am sure that all these numbers will help Riz in better modelling the attributes system in future fixes.

[nino33]

I'm not saying there's not the issue you guys are noting (it sure was there in EHM)7!), but it sure would be nice to have some "proof" - there's been no review of first round picks or top 10 picks, no screenshots/saved games uploaded, nothing to support this wide discrepancy in Attributes that anecdotal evidence says is occurring

So far my attempts at supporting anecdotal evidence with my testing has shown the anecdotal evidence to be wrong/unsupported

I don't think my data addresses this issue. I would think if those that see it upload saved games with examples Riz will address it, but if not he may not see it looking at the data (I don't see the data above showing this problem......there's nothing above that separates a first round draft choice from an undrafted player or that shows players have such wide/unrealistic variances)

[Alessandro]

So far my attempts at supporting anecdotal evidence with my testing has shown the anecdotal evidence to be wrong/unsupported

You wrote yourself:

But the low ranges are see for all the hockey skills is significantly off IMO (there's no way there's any NHLer with a 2 or a 3 or a 4 in any of those skills.....

[nino33]

I did no testing on the variance between Attributes.
I wasn't saying I disagreed, I was saying indirectly "it sure would be nice if others used the EHM:EA Assistant to produce actual evidence/data"
Even screenshots would be better than no support/evidence IMO

The testing you're referring to (above, in this thread) was supposed to show (based on anecdotal evidence) poor regens - and it did not.
The Staff testing I did yesterday was supposed to show all Staff with coaching Attributes below 8 (but the NHL averaged 10-11).....so the testing I've done has not supported the anecdotal evidence


EDIT - Another comment about "random results" - as has already been shown, the game overwrites the values entered by a researcher if it doesn't fit the game mechanics.....I think there's less chance of random with EHM:EA than any other version of EHM (and I'd think there'd be coding for Staff, and it seems there is/though it may need tweaking, but IMO it's certainly not "random")

[Manimal]

"random" is probably the wrong wording here. What it is is "relative" IMO.
"relative" to me here means "random with parameters"

[nino33]

"random" is probably the wrong wording here. What it is is "relative" IMO.
"relative" to me here means "random with parameters"

Agreed

I think the issue is the parameters need tightening up, and I think we're all agreed on this...in some/many cases the low value/threshold needs to be raised (with some Attributes like the coaching Attributes that's all I'd do as I think the NHL averages are a little low, with some other Attributes where the average is already at a desired level then reduction for some is required to compensate for raising others to a minimum/reasonable level)

I don't understand why some think I don't recognize/accept the problem of "to low" Attributes! HaHa
I'm the one that long ago suggested "League minimums" for Attributes when looking at Major Junior/potential NHL draftees & Staff (and I think that minimum success in a League, as shown by Games Played/Games Coached, should result in rapid growth to the minimum Attribute level IMO)

[visualdarkness]

I havent saved too many screens sadly.

Nr 1 ranked:

"Interesting" distribution:

I will try to get some better examles when I got the time.

[nino33]

Looking at Test 1 in the year 2030 (16 years after start-up), and looking at the top 600 CA players in the NHL in 2030...

• average CA is 144.5
  21 players with a 180-189 CA
  33 players with a 170-179 CA
  63 players with a 160-169 CA
  108 players with a 150-159 CA
  129 players with a 140-149 CA
  102 players with a 130-139 CA

Here's looking at Test 1 in the year 2030, and looking at the top 600 CA players in the AHL...

• average CA is 101.7
  0 players with a 180-189 CA
  0 players with a 170-179 CA
  1 players with a 160-169 CA (162)
  8 players with a 150-159 CA
  18 players with a 140-149 CA
  18 players with a 130-139 CA
  74 players with a 120-129 CA
  102 players with a 110-119 CA
  67 players with a 100-109 CA
  125 players with a 90-99 CA
  124 players with a 80-89 CA
  62 players with a 70-79 CA
  1 players with a 69 CA

Here's looking at Test 1 in the year 2030, and looking at the top 560 CA players in the KHL
average CA is 96.8

• 2 players with a 180-189 CA
  1 players with a 170-179 CA
  2 players with a 160-169 CA
  9 players with a 150-159 CA
  11 players with a 140-149 CA
  17 players with a 130-139 CA
  52 players with a 120-129 CA
  58 players with a 110-119 CA
  90 players with a 100-109 CA
  101 players with a 90-99 CA
  88 players with a 80-89 CA
  67 players with a 70-79 CA
  62 players under 70 CA (range 56-69)

Here's looking at Test 1 in the year 2030, and looking at the top 240 CA players in the SHL
average CA is 102.0

• 0 players with a 180-189 CA
  0 players with a 170-179 CA
  1 players with a 160-169 CA (160)
  3 players with a 150-159 CA
  7 players with a 140-149 CA
  8 players with a 130-139 CA
  27 players with a 120-129 CA
  28 players with a 110-119 CA
  56 players with a 100-109 CA
  44 players with a 90-99 CA
  45 players with a 80-89 CA
  21 players with a 70-79 CA

[visualdarkness]

Here are all the goalies from one draft. As you can see despite having one year of development after the draft all goalies except one got bad recovery OR rebound control.

As you can see there seems to be a pattern in which one of these two attributes are nerfed bad by the engine and sort of cripples the goalies. This is even worse the following years but I wanted to use goalies that had some development and this was a save I had. The next year 7 is the best of any goalie in the "nerfed att" with loads of 2-5s.

http://s787.photobucket.com/user/Visual ... e%20regens