testing data - new player (age 13-14) Attributes

Discuss all aspects of editing the data and databases in EHM here. Have a question about the EHM Editor, EHM Assistant, editing the .cfg files, hex editing the .dat or .db files? Want to tweak the EHM exe file to change league rules/structure, start date etc? This is the place!
Forum rules
This is the forum to discuss all aspects of editing the EHM data and tweaking the game.

Have a bug or feature request for the EHM Editor? Post them in the EHM Editor thread. Please start a new thread or post in another thread if you have a question about how to use the EHM Editor.

Given the large number of questions on similar topics, we ask that you start a new thread for a new question unless you can locate a similar question in an existing thread. This will hopefully ensure that similar questions do not get buried in large threads.

Useful links: EHM 1 Assistant (Download) | EHM 1 Editor (Download) | EHM 1 Editor Tutorials | Editing Rules & Structures Guide | Converting EHM 2004 / 2005 DBs to EHM 1 | Converting an EHM 2007 DB to EHM 1 | Extra_config.cfg | Import_config.cfg | Player Roles
Post Reply
User avatar
nino33
Mr. Goalie
Posts: 6088
Joined: Sat Aug 07, 2010 3:37 am
Custom Rank: Retro Rosters Specialist
Favourite Team: 1970s hockey

testing data - new player (age 13-14) Attributes

Post by nino33 »

This thread is about the youngest players ingame (those age 13-14) and what the average Attribute value for each Attribute is when these players first appear in game.

Upon starting a game, if you immediately do a search for the youngest players available, you will see the youngest players available are age 13-14.

EHM allows a database to contain players younger than this, but they won't show up ingame until age 13-14.
You can even have players in a database that haven't even been born yet! They will appear "ingame" when they reach age 13-14.
The Attributes do not begin to change/develop until the player is actually ingame.

EHM can potentially change Attributes at startup! There are limitations, or checks and balances.
Here's some examples https://1drv.ms/x/s!AvTHo4MBaphkgQzwJKc ... x?e=oEL1ja

I don't know the exact formula(s) used or what the exact reasoning/pathways are for what gets altered and what doesn't.....but now I have some baseline numbers for new players entering the game (a helpful thing for a retro/historical database, or a "seeded" modern or fantasy database)


  • ImActuallyAnOtter created a Prospect Generator, a tool that "seeds the database with future players" - Here's a thread about it https://www.ehmtheblueline.com/forums/v ... 97#p240397
  • The current version is used by both TBL and ECK database (FYI there's also versions without the added prospects) - 100 years of prospects!
  • The game determines the player Attributes for all players (including those that don't show up yet ingame) at startup. Using the EHM Assistant I exported these Attributes and compiled the data for players born between 2010 and 2099 (over 16K players); I exported the data at the Select GM stage.



Here’s a link to the spreadsheet https://1drv.ms/x/s!AvTHo4MBaphkgQrKNAd ... s?e=13ycBJ

The spreadsheet has 3 tabs - one each for Center, Defensemen and Goalie.
My understanding is wingers and centers are essentially the same except for the Faceoff Attribute.
Ideally I wanted to compile the winger data as well, but didn't; the goaltender tab just has the butterfly data.

From left to right there are three sections – the first covers CA/PA data, the second the Physical-Technical-Mental Attributes (with a couple columns averaging them), and the third section covers Offensive/Defensive Role and the Staff Attributes

Across the very top line I indicate which Attributes are permanent/never change, and which change very little (in comparison to the rest)

There are three different rows of data for each Player Role – one that includes all, one that includes just those whose starting CA was 70-80, and one for those whose starting CA is 40-50. For each Player Role, the average Attribute value for each Attribute is shown.






I hope to have the time to comment more in this thread in the future; I welcome anyone's comments/observations


One thing that stood out to me is how the center's skating Attributes (especially Speed and Acceleration) can be significantly different at startup depending on Player Role
  • low/high range for Agility 10.1 - 12.8
  • low/high range for Balance 8.1 - 12.7
  • low/high range for Speed 3.1 – 12.6 (with all-around, defensive, defensive finesse, finesse, playmaker and sniper all starting at the low end, grinder in the middle, and defensive physical and enforcer at the high end!)
  • low/high range for Acceleration 3.2 - 12.7 (with all-around, defensive, defensive finesse, finesse, playmaker and sniper all starting at the low end, and defensive physical and enforcer at the high end!)
User avatar
TBIF
Top Prospect
Posts: 106
Joined: Sun Jan 01, 2017 8:38 pm
Custom Rank: 2000/01 Rosters Author
Favourite Team: Canucks

testing data - new player (age 13-14) Attributes

Post by TBIF »

This will help greatly with a project I'm working on!

I'd love to figure out the science behind the CA - attribute distribution. Like an auto generated 100 CA - Centre : Power forward and have the "max" available stats to distribute.

If my rig was better, and I had a fraction more time and organization skills I could probably figure out the best way to test it.
Smirnov2Chistov
Prospect
Posts: 80
Joined: Mon Jul 10, 2017 10:06 pm
Favourite Team: Anaheim Ducks

testing data - new player (age 13-14) Attributes

Post by Smirnov2Chistov »

Nino, this topic regarding these ‘Dead Era’ years, could they be avoid by expanding a DB? As I repeated before, I probably play with a larger DB than anyone else (with only the NHL selected). It’s because through your expansive testing that more ‘better’ players do appear if you have a large database with extra players included.
User avatar
nino33
Mr. Goalie
Posts: 6088
Joined: Sat Aug 07, 2010 3:37 am
Custom Rank: Retro Rosters Specialist
Favourite Team: 1970s hockey

testing data - new player (age 13-14) Attributes

Post by nino33 »

Smirnov2Chistov wrote: Sat May 16, 2020 12:12 amNino, this topic regarding these ‘Dead Era’ years, could they be avoid by expanding a DB? As I repeated before, I probably play with a larger DB than anyone else (with only the NHL selected). It’s because through your expansive testing that more ‘better’ players do appear if you have a large database with extra players included.
I'm not sure about database size and the possible effect on drafts (I always use the maximum sized database); I wouldn't think there'd be a significant difference in regards to the number of NHL quality players available
Smirnov2Chistov
Prospect
Posts: 80
Joined: Mon Jul 10, 2017 10:06 pm
Favourite Team: Anaheim Ducks

testing data - new player (age 13-14) Attributes

Post by Smirnov2Chistov »

nino33 wrote: Thu Apr 16, 2020 6:38 am You can even have players in a database that haven't even been born yet! They will appear "ingame" when they reach age 13-14.
The Attributes do not begin to change/develop until the player is actually ingame.
Nino, looking at some old threads got me thinking about something.

Could you hypothetically have a Sidney Crosby/Connor McDavid regen but not even be born yet 'within' the game? I thought when a player retired in EHM1, you would still have to wait 2-3 years before they appear.

I must admit - I'm potentially one of the few that doesn't care for the Prospect Generator. In some testing that I've done (simple 5-8 year sim) feel like many of the players from it are a bit wonky. I remember a save that someone uploaded from HFBoards and the realism was lost for me when you had top line players from China, Japan, Uzbekistan. There would be players that are really HUGE, and others super tiny lol.
User avatar
nino33
Mr. Goalie
Posts: 6088
Joined: Sat Aug 07, 2010 3:37 am
Custom Rank: Retro Rosters Specialist
Favourite Team: 1970s hockey

testing data - new player (age 13-14) Attributes

Post by nino33 »

Smirnov2Chistov wrote: Mon Nov 02, 2020 3:47 pm Could you hypothetically have a Sidney Crosby/Connor McDavid regen but not even be born yet 'within' the game? I thought when a player retired in EHM1, you would still have to wait 2-3 years before they appear.
You can't have a regen without the player actually existing and retiring (and taking a few years to regen as you noted), so you couldn't have such a regen


You can "seed a database" which is what the prospect generator does (with fake players), and what a retro database does too (with real players).

Any retro database that starts before 1997 could/would have a Sidney Crosby/Connor McDavid in it who had not been born yet (but they wouldn't be regens, they'd be "future players" who would appear for the first time at age 14-16 ingame). And anyone could edit such players into any other database if desired.


P.S. Something related that's coming: the upcoming 1.5 update will allow the PA limitation for a set time for all regens (for example, the regens PA levels can be limited to 100 or 120). This allows a retro database to have only real NHL level players, so past seasons can be played without regens affecting things.
Post Reply