Saved Game Editor: Player Attributes Conversion

[DannL]

Hey all. I've decided to start my own EHM 2007 project. My goal is to generally uncover the attribute ranges that the saved game editor uses.

When using the saved game editor, everyone notices that some attributes are listed in the normal range of 1-20 while others are listed in ranges that go from 1-255.

From my surveying, I've noticed that these attributes follow the normal 1-20 range and are easy to get the desired attribute rating in-game:
- Aggression
- Bravery
- Flair
- Leadership (Influence)
- Teamwork
- Workrate
- Determination (staff attribute)
- Consistency (shown on a player's Information screen in-game under Details)
- Acceleration
- Agility
- Speed (Pace)
- Stamina
- Strength

And these attributes follow the weird and non-linear 1-255 range which makes it difficult to achieve the desired attribute rating in-game:
- Anticipation
- Creativity (Vision)
- Balance
- Checking
- Deflections
- Deking
- Faceoffs
- Hitting
- Getting Open (Movement)
- Passing
- Pokecheck
- Positioning
- Slapshot
- Stickhandling
- Wristshot
- ALL Goalie Attributes (Blocker, Glove, Rebounds, Recovery, Reflexes)

These attributes seem to be hidden in-game and I'm still unsure how they correlate in-game:
- Decisions (1-255, hidden in-game)
- Dirtyness (1-20, hidden in-game)
- Imp. Matches (1-20, hidden in-game)
- Fighting (1-255, hidden in-game)
- Injury Prone (1-20, hidden in-game)
- Natural Fitness (1-20, hidden in-game)

My method of doing this is basically surveying the non-hidden 1-255 attributes in-game with what the editor shows.

i.e. players A, B, and C have anticipation of 15 in-game. In the save game editor, Player A's anticipation is 52, Player B's is 58, and Player C's is 65. So if you wanted to give a player 15 anticipation in-game, you'd have to give him anticipation of 52 to 65 in the editor.

Now obviously I'm not just surveying 3 players out of each 1-20 attribute rating. I've been using the in-game Player Search, with no fog of war, sorting by the various 1-255 attributes, and going through EVERYONE. It's tedious but I've been recording my findings in a spreadsheet with a general pattern developing.

As I went along, I realized that age plays a factor in the save game editor ranges. For example, 14 year olds with 15 anticipation will have a slightly different range than 41 year olds with 15 anticipation. Now I've only gotten to 18 year olds so far, but I've generally noticed that as age increases, the ranges decrease for every 1-20 rating. There is an observed pattern so this is giving me a purpose to finish.

I hope you guys get what I'm saying here. But the ultimate point I'm trying to make is that when I'm done, I'll have the full spreadsheet completed. This spreadsheet will be a referencing tool for the 1-255 attributes so that people can just look up a player's current age, determine what they want the attribute to be in-game (from 1-20), and see the approximate number they should input into the saved game editor to get the desired 1-20 rating in-game. Then people won't have to start a new game to edit players exactly how they want.

The spreadsheet's axis is Age from 14-42 on the Y-axis and Rating from 1-20 on the X-axis.

I'm not entirely sure how long this is going to take me. Some days I'm not so lazy and I can get through an entire age group in a day. But I'm guessing I should have the full spreadsheet done by the time the 2012-2013 NHL season starts. I'm a 20-year old 3rd year university student enjoying summer so I'm trying to get it done before September when my school starts again.

[DannL]

On second thought, I'm starting to think this referencing tool won't help very much. Funny how I just posted that I would be able to find a correlating pattern and I'm already giving up. Seems like the EHM gods don't want me to fully uncover the secrets of the 1-255 range.

There's just too many ranges that overlap each other so, even if you used my spreadsheet, you'd still have to do some guesswork as to which number you should use to get the desired attribute rating.

However, I will say that attributes in the 213-255 range will give your player an attribute rating of 1-10. Attributes in the 213-216 range will always give out an attribute rating of 1 in-game. So there is a pattern. However, recording all of this manually into a spreadsheet could take me a REALLY long time and I really just want to carry on with my Caps season.

My advice is if you wan't to edit a player through the Saved Game Editor exactly how you wan't him to be you should follow these steps:

1) Go into the in-game Player Search.
2) Configure the filter to ONLY show the players that have the same age as the player you wan't to edit.
3) Sort that list by the attribute that you wan't to edit. (Or really, sort it by any 1-255 attribute. Doesn't matter.).
4) Take a small sample of 10-20 players with the desired in-game attribute rating and record their Saved Game Editor values for that attribute.
5) Use the same Saved Game Editor value as some of those 10-20 players or even take the average of those values.
6) Use that value on the attribute for the player you wish to edit.

Note that every attribute that uses the 1-255 range follows the same pattern. So if a player has two or more "1-255" attributes with the same in-game rating, they'll use the same number in the Saved Game Editor.

For example, Ryan Getzlaf is rated a 17 in Anticipation, Creativity (Vision), Stickhandling, and Wristshot. In the SGE, those attributes all have a value of 27. He's rated a 16 in Deflections, Faceoffs, Hitting, and Getting Open. All of those attributes have a value of 18. So you can see the pattern there.

I've also noticed that if a much younger player is really good at certain attributes already (i.e. 17-20 in-game), the values for those certain attributes will be quite high 75-85. But once you get into prime aged players (i.e. mid-20 year olds) the values for attributes that are 17-20 in-game drop down to mid-20s and 30s in the SGE.

My theory on this is that as a player ages, the 1-255 attributes start to decline in value in the SGE. This decline in SGE value keeps in check (based on age) with the actual in-game rating. So if a really young player has an attribute in the 17-20 range in-game, his SGE value starts quite high initially (i.e. 75-85). As this player ages, this 75-85 value gradually declines with each year. By the time this player is say 26 years old, the SGE value will be in the mid-20s and 30s. Yet, this value is still really good for his age group, and the player still retains his attribute rating of 17-20.

Another thing that shows this fact is that if a very young player's SGE value is already in the mid-20s and 30s, his in-game attribute rating will be around 10-13. And by the time he's 26, his SGE value will be at 8-14 but he'll still have the same in-game rating that he had when he was really young (10-13 in-game).

So as you can see, the SGE value declines as the player ages, yet the in-game rating remains the same and the SGE value remains relatively the same as it was when the player was young. Now by relatively the same I mean the player's SGE value remains the same relative to the other SGE values of players in the same age group.

Not sure if this has already been discovered by others but I hope you guys get what I'm saying anyways.

But yea, if you want to achieve a desired in-game attribute rating for a player, just copy the SGE value of a player in the same age group. More likely than not, this will give you the desired in-game rating. If it doesn't, take a bigger sample size but 10-20 should be more than enough.

This also applies to hidden 1-255 attributes like Fighting and Decisions. In the case of those, just figure out the 1-255 SGE value that correlates to whatever in-game 1-20 rating for that age group and use that number for Fighting and Decisions.

I hope this helps some people out and that I'm not reiterating what someone else has already discovered. Feel free to PM me for any questions.

[archibalduk]

I don't have much experience of using the SGE - I never knew it used a 0-255 scale. I wonder why this is - the database (and I presume also the saved game files) uses a 0-20 scale...

[Mordrehl]

I don't have much experience of using the SGE - I never knew it used a 0-255 scale. I wonder why this is - the database (and I presume also the saved game files) uses a 0-20 scale...

From what I've seen ... the SGE and thus the game itself works as follows, though I could be wrong

some attributes use the straight value (ie SPD ... whatever is entered, is what shows up in game)

A great many attributes however use a relational value. CHK, PAS, DEK, WRS, SLA are all examples.

A value between 0-127 produces a value equal to or greater than the players current ability value/10. So if CA is currently 100, we would expect checking to be close to 10 for low values in that field (ie 0-20ish). The higher the value, the higher checking will be in relation to CA/10.

For high values, (ie 200-255) it seems to give a value lower than the CA. There also seems to be a magical number of 136, which is entered in One on Ones for almost all players. So I assume the "lower than CA" range is from 137-255

This means as a players CA rises, their stat will rise proportionally..

What I am currently checking, is how the PRACTICE schedules relate to this design. What I suspect, is that a practice area that raises a stat, does not in fact raise that stat ... but changes the relative value itself. Thus sliding from 255 down to 137 or from 0 to 127 ... and as a by product raising the value in ralation to CA.

It would seem that CA is purely affected by game experience.

[Mordrehl]

Hmm just read the part where player age seems to change the values as well. First we need to distinguish that this is NOT how the SGE stores the info but rather this is how EHM does.

So the sliding scale is probably 0-135 for higher than CA and 137-255 for lower than. Don't ask me what 136 is, the only place I've seen it is One on Ones

What I'm curious about, and I'll check when I'm home ... is if we have any prime aged players (27-33 for Forwards, 29-35 for Defence) with attributes rates 137-255 in the saved game file.


Btw, the sliding scale for attributes explains why you can scout a player who is rated as a superstar (potential ability > 180) and yet has the worst stats possible and never developes NHL level attributes. Scouting seems to be based purely off the potential ability. What you want to combine that with, is the player they compare him to ... since the final attributes can be closer to that player. The "similar player" probably looks at the combination of abilities and how much higher or lower than the Current Ability they will be and compares them to a known player with identical or similar values.

[DannL]

One-on-Ones is a goalie stat. So all non-goalie players have 136 in this attribute which probably equals 0. That's why pretty much all non-goalie players also have 136s in Blocker, Glove, Rebounds, Recovery, and Reflexes.

[Mordrehl]

One-on-Ones is a goalie stat. So all non-goalie players have 136 in this attribute which probably equals 0. That's why pretty much all non-goalie players also have 136s in Blocker, Glove, Rebounds, Recovery, and Reflexes.

Actually, that makes sense ...

Pity we don't have a utility to do a full dump of the saved game to an excel sheet for every player.

[Mordrehl]

These attributes seem to be hidden in-game and I'm still unsure how they correlate in-game:
- Decisions (1-255, hidden in-game)
- Dirtyness (1-20, hidden in-game)
- Imp. Matches (1-20, hidden in-game)
- Fighting (1-255, hidden in-game)
- Injury Prone (1-20, hidden in-game)
- Natural Fitness (1-20, hidden in-game)

Decisions, while not apparent I think the higher the value the better a player is
Dirtyness has a direct impact on Penalty Minutes and Suspensions
Imp. Matches has a direct impact on playoff game performance or any game that would be considered "important" ... you'll see players with a high value here having a 8+ rating in important games
Fighting - propensity to win fights
Injury Prone - the lower the better. A high value results in a player who is made of glass
Natural Fitness - seems to affect how quickly a player recovers from being fatigues after games

[DannL]

Pity we don't have a utility to do a full dump of the saved game to an excel sheet for every player.

I would kill for that data lol.

[Mordrehl]

Hmmm who created the SGE and is the source code kicking around anywhere?

[archibalduk]

This is really interesting - keep up the good work!

Unfortunately SI cannot publish the saved game structure because it contains all sorts of proprietary stuff. However, they did say that users are more than welcome to take a look and post whatever they find. This is one reason why I haven't taken a look at the saved games yet - i.e. the structure isn't freely available (whereas the structure is available for the database files).

The saved game is essentially a number of files joined together into one file (assuming you're not using the "saved compressed" option in EHM). Presumably these are similar to the database files, but with additional data. If you have EHM running with a loaded game, you can find the files that make up the saved game by going to:

Windows Vista/7: C:\Users\<USERNAME>\AppData\Roaming\Sports Interactive\EHM 2007\temp
Windows XP: C:\Program Files\Sports Interactive\NHL Eastside Hockey Manager 2007\temp

Hmmm who created the SGE and is the source code kicking around anywhere?

I think mne2 created it. The source code isn't available, but you could ask whether he has any "structs" (C/C++ style data structures) or any other structure info that he might be able to share. This would give us a big head start in considering how to dump the player attributes into Excel. I'm really interested in this because I want to export player stats from EHM into Excel (the data export options in EHM are too basic IMHO) for the purposes of evaluating players' in game performance with their CAs (because this would help with perfecting player ratings when updating the rosters).

[DannL]

A value between 0-127 produces a value equal to or greater than the players current ability value/10. So if CA is currently 100, we would expect checking to be close to 10 for low values in that field (ie 0-20ish). The higher the value, the higher checking will be in relation to CA/10.

For high values, (ie 200-255) it seems to give a value lower than the CA. There also seems to be a magical number of 136, which is entered in One on Ones for almost all players. So I assume the "lower than CA" range is from 137-255

This means as a players CA rises, their stat will rise proportionally..

You sure this is full-proof?

I'm looking at Ovechkin right now and his CA is at 188 so divide that by 10 and we get 18.8. For attributes that are above this rating (i.e. 19 or 20) these SGE values work with what you're saying:

- Anticipation is 19 in-game and 35 in the SGE
- Hitting is 19 in-game and 35 in the SGE
- Balance is 20 in-game and 63 in the the SGE
- Slapshot and Wristshot are 20 in-game and 49 in the SGE

However, when looking at attributes that are lower than 18.8, your logic doesn't apply:
- Deflections is 17 in-game, 24 in SGE
- Deking is 17 in-game, 26 in SGE
- Passing is 16 in-game, 10 in SGE
- Pokecheck is 15 in-game, 4 in SGE

By your logic those should all be in the 137-255 range.

So I'm not sure that theory works. I'd be hard-pressed to find a player with a 15+ in-game rating in any of the "special" 255 attributes and have it have a SGE value in the 150-255 range.

But I do see a sliding scale pattern.

It really is tough to figure this all out without being able to compare large amounts of players' CAs, in-game ratings, and SGE values all in one place. A utility to export the SGE data into a spreadsheet would be amazing.

Another sample of a "crappier" player I took Luke Schenn. CA is 138 so /10 you get 13.8.
- Balance 15 in-game, 29 SGE
- Anticipation 12 in-game, 4 SGE (close!)
- Creativity 12 in-game, 255 SGE (now we're getting somewhere)
- Checking and Positioning 17 in-game, 43 SGE
- Deflections 11 in-game, 253 SGE (ok)
- Deking and Getting Open 12 in-game, 0 SGE
- Passing and Pokecheck 15 in-game, 27 SGE
- Slapshot 13 in-game, 6 SGE
- Stickhandling 13 in-game, 10 SGE
- Wristshot 11 in-game, 253 SGE

So as you can see your theory works better with this player. Although some of the attributes under the CA/10 aren't in the 137-255 range, although they came close. Also, my theory that attributes that share the same value in-game also share the same SGE value for the same player doesn't always work, but it also comes quite close.

[Mordrehl]

hmmm might be problematic asking mne2 ....

Last visited:
Thu Dec 29, 2011 9:53 am

[Mordrehl]

A value between 0-127 produces a value equal to or greater than the players current ability value/10. So if CA is currently 100, we would expect checking to be close to 10 for low values in that field (ie 0-20ish). The higher the value, the higher checking will be in relation to CA/10.

For high values, (ie 200-255) it seems to give a value lower than the CA. There also seems to be a magical number of 136, which is entered in One on Ones for almost all players. So I assume the "lower than CA" range is from 137-255

This means as a players CA rises, their stat will rise proportionally..

You sure this is full-proof?

I'm looking at Ovechkin right now and his CA is at 188 so divide that by 10 and we get 18.8. For attributes that are above this rating (i.e. 19 or 20) these SGE values work with what you're saying:

- Anticipation is 19 in-game and 35 in the SGE
- Hitting is 19 in-game and 35 in the SGE
- Balance is 20 in-game and 63 in the the SGE
- Slapshot and Wristshot are 20 in-game and 49 in the SGE

However, when looking at attributes that are lower than 18.8, your logic doesn't apply:
- Deflections is 17 in-game, 24 in SGE
- Deking is 17 in-game, 26 in SGE
- Passing is 16 in-game, 10 in SGE
- Pokecheck is 15 in-game, 4 in SGE

By your logic those should all be in the 137-255 range.

Try comparing using the alternative attribute range turned on (ie 1-100) and see if there's a correlation to the attribute vs CA/2?

[Mordrehl]

Another sample of a "crappier" player I took Luke Schenn. CA is 138 so /10 you get 13.8.
- Balance 15 in-game, 29 SGE
- Anticipation 12 in-game, 4 SGE (close!)
- Creativity 12 in-game, 255 SGE (now we're getting somewhere)
- Checking and Positioning 17 in-game, 43 SGE
- Deflections 11 in-game, 253 SGE (ok)
- Deking and Getting Open 12 in-game, 0 SGE
- Passing and Pokecheck 15 in-game, 27 SGE
- Slapshot 13 in-game, 6 SGE
- Stickhandling 13 in-game, 10 SGE
- Wristshot 11 in-game, 253 SGE

So as you can see your theory works better with this player. Although some of the attributes under the CA/10 aren't in the 137-255 range, although they came close. Also, my theory that attributes that share the same value in-game also share the same SGE value for the same player doesn't always work, but it also comes quite close.

Hmmm wonder if each attribute has it's own sliding scale.

[archibalduk]

CA doesn't work as simply as you're suggesting. The CA sets an expected average total of the technical attributes, but this isn't concrete because it will allow some variation so long as each attribute is within a certain range - e.g. for a CA of 150, each technical attribute for a player should not be lower than 5 or greater than 17 and the average should be roughly 11.5.

[Mordrehl]

CA doesn't work as simply as you're suggesting. The CA sets an expected average total of the technical attributes, but this isn't concrete because it will allow some variation so long as each attribute is within a certain range - e.g. for a CA of 150, each technical attribute for a player should not be lower than 5 or greater than 17 and the average should be roughly 11.5.

So .... we have the complication of this, added in with the value ranges in the Save Game file of 0-255 where 0-135 appears to be higher than average and 137-255 appears to be lower than average maybe


do we know what the ranges are for CA? is it a calculation? I thought it was always the average was CA/10?

[Mordrehl]

NVM, found it ... hmm I'm going to do some playing. I don't think this chart is hard and fast because I've definately edited Super Duper stars by setting their values to 64ish and had players with a CA of 150 but all attributes 20

CA AVR MIN MAX
10 4,5 1,0 10,5
20 5,0 1,0 11,0
30 5,5 1,0 11,5
40 6,0 1,0 12,0
50 6,5 1,0 12,5
60 6,0 1,0 13,0
70 7,5 1,5 13,5
80 8,0 2,0 14,0
90 8,5 2,5 14,5
100 9,0 3,0 15,0
110 9,5 3,5 15,5
120 10,0 4,0 16,0
130 10,5 4,5 16,5
140 11,0 5,0 17,0
150 11,5 5,5 17,5
160 12,0 6,0 18,0
170 12,5 6,5 18,5
180 13,0 7,0 19,0
190 13,5 7,5 19,5
200 14,0 8,0 20,0

[archibalduk]

I think it is a relatively hard and fast rule. One of the functions of the CA is to moderate a player's attributes. Try creating a new player in the database and setting all of his attributes to 20 and his CA to something like 100 or 150. You'll see his technical attributes won't all be 20 when you start a new game.

The Saved Game Editor might allow you to alter the attributes in spite of the CA's normal limits, but this is because you're bypassing the normal rules EHM uses and are directly modifying the game's data (whereas the data modified in the database is processed by EHM upon starting a new game).

[Mordrehl]

I think it is a relatively hard and fast rule. One of the functions of the CA is to moderate a player's attributes. Try creating a new player in the database and setting all of his attributes to 20 and his CA to something like 100 or 150. You'll see his technical attributes won't all be 20 when you start a new game.

The Saved Game Editor might allow you to alter the attributes in spite of the CA's normal limits, but this is because you're bypassing the normal rules EHM uses and are directly modifying the game's data (whereas the data modified in the database is processed by EHM upon starting a new game).

I think it's hard and fast when it comes to how it creates the initial attributes based on the Database ... but as far as the editor and post-save game I have clearly seen that it's not in effect. So it's simply a matter of figuring out how the 0-255 range translates to the current attribute value as seen in game.

[Mordrehl]

I think it is a relatively hard and fast rule. One of the functions of the CA is to moderate a player's attributes. Try creating a new player in the database and setting all of his attributes to 20 and his CA to something like 100 or 150. You'll see his technical attributes won't all be 20 when you start a new game.

The Saved Game Editor might allow you to alter the attributes in spite of the CA's normal limits, but this is because you're bypassing the normal rules EHM uses and are directly modifying the game's data (whereas the data modified in the database is processed by EHM upon starting a new game).

I think it's hard and fast when it comes to how it creates the initial attributes based on the Database ... but as far as the editor and post-save game I have clearly seen that it's not in effect. So it's simply a matter of figuring out how the 0-255 range translates to the current attribute value as seen in game.

Actually to clarify, it's a hard and fast rule on how a players attributes are distrubuted by the game engine (whether created at game start from the DB or a player who is created/regen'd during a game).

But the value's listed in the SGE, which presumably are what was pulled from the saved game data ... can be edited with no regards to that CA chart.

When I get home, going to pick a team with 20 players ... set them all to CA 100 and set their attributes to increasing value's starting at 0 and see what happens

Then I'll increase and decrease the CA for all players and see what happens.

Then I'll set the CA back to 100 and increase age to see what happens.

EDIT : Another monkey wrench in the mix is the players Offensive Role and Defensive Role stats ... going to have to test those as well

[archibalduk]

Actually to clarify, it's a hard and fast rule on how a players attributes are distrubuted by the game engine (whether created at game start from the DB or a player who is created/regen'd during a game).

This is exactly what I was trying to say.

But the value's listed in the SGE, which presumably are what was pulled from the saved game data ... can be edited with no regards to that CA chart.

That's correct. The SGE is essentially a hex editor. The data can be edited irrespective of what hardcoded rules/processes may be present in EHM. I wonder if EHM will self-correct the altered attributes at one of the in game data update dates (i.e. 1 January and 1 July).

I'm curious as to what is ultimately hoped to be achieved through editing the saved game. It feels like anything that can be edited using the SGE can more easily be edited in the database using the Pre-Game Editor before starting a new game. Please don't take this as criticism; I'm just interested to know what the ultimate aim is.

[Mordrehl]

Actually to clarify, it's a hard and fast rule on how a players attributes are distrubuted by the game engine (whether created at game start from the DB or a player who is created/regen'd during a game).

This is exactly what I was trying to say.

But the value's listed in the SGE, which presumably are what was pulled from the saved game data ... can be edited with no regards to that CA chart.

That's correct. The SGE is essentially a hex editor. The data can be edited irrespective of what hardcoded rules/processes may be present in EHM. I wonder if EHM will self-correct the altered attributes at one of the in game data update dates (i.e. 1 January and 1 July).

I'm curious as to what is ultimately hoped to be achieved through editing the saved game. It feels like anything that can be edited using the SGE can more easily be edited in the database using the Pre-Game Editor before starting a new game. Please don't take this as criticism; I'm just interested to know what the ultimate aim is.

EHM does not correct the players ... one of my "pure fantasy" games I edited the top prospect for the QMJHL entry draft and he developed into all 20's in every attribute with a CA of 170
I think the CA rule is only enforced upon game start and when a player is regen'ed

My hope, is simple ... the ability to realistically fix prospects on occasion who are simply well .. broken. Nail Yakupov for example in Manimal 4.0 (think it's 4.0) ... good Potential Ability, horrible attributes even when his CA reaches close to his PA. Having said that, I'm trying to stay away from the editor as much as possible since it tends to ruin my enjoyment.

[archibalduk]

Ah I see! Please do also report any broken prospects to the Manimal Mistakes Thread so we can check them out and perhaps tweak things if needs be.

[DannL]

I'm curious as to what is ultimately hoped to be achieved through editing the saved game. It feels like anything that can be edited using the SGE can more easily be edited in the database using the Pre-Game Editor before starting a new game. Please don't take this as criticism; I'm just interested to know what the ultimate aim is.

I like to use the SGE to edit upcoming draft picks as I get to know more about them in real life as the ISS rankings and whatnot are released. I also use it to edit players here and there when I feel like they are greatly under-performing or just need a boost in some attributes. I try to keep my save file file realistic with how players are performing in real life. I try not to tinker too much because then I'd feel like I be cheating. I know, in-game, some players are bound to have their ups and downs but some times I'll tinker to balance it out if I deem it necessary for my own enjoyment.

Obviously I'm not yet too familiar with the players that are eligible for the 2013 NHL Draft or any other upcoming draft. But I can assume a bunch of those players are potentially underrated as their stock will rise as the next NHL season rolls along and the 2013 ISS rankings come out. I don't want to completely restart my seasons in order to edit players how I want.

That's my reasoning.