testing data - NHL forwards Attributes 2018/2050

Post by **nino33** » Tue Jun 25, 2019 1:59 am

Greetings!

While compiling data for some recent Player Role/Attribute testing I've been doing https://www.ehmtheblueline.com/forums/v ... 6&start=50 I got more and more curious about
• Attributes and their range (lowest/highest)
• Attributes and the frequency of players getting “low” Attribute values (under 10) and the frequency of getting “high” Attribute values (16-20)

What follows is from my test save using TBL 11.2, comparing the results from 2018 (startup) and in 2050

I took a look at the top 372 forwards in the NHL by CA (12 forwards x 31 teams = 372 forwards). Very late in the data compiling process I discovered that for 2050 I had accidentally included 6 defensive finesse defensemen, but I decided not to redo anything as I don’t think the small number significantly affects the results

What is evident is with the game created values in 2050 (when any human inputted values are long gone) there are some really significant differences in the frequency of lower/higher Attribute values compared to startup

As mentioned, I sorted by CA to get the top 372 forwards in the NHL. Here’s a list of the number of players for the different Player Roles. The first number is how many in 2018 and the second number is how many in 2050

All-Around 70 – 53
Defensive 14 – 12
Defensive Finesse 12 – 7
Defensive Physical 15 – 8
Finesse 30 – 46
Grinder 12 – 9
Playmaker 44 – 53
Playmaker Finesse 59 – 49
Playmaker Physical 7 – 11
Power Forward 45 – 30
Sniper 26 – 46
Sniper Finesse 25 – 23
Sniper Physical 11 – 19
Enforcer 2 – 0

And here's the some basic data on the two groups

the 2018 players had a 141.5 CA avg (126-190 range) and a 146.8 PA avg (126-198 range)
the 2050 players had a 148.7 CA avg (128-198 range) and a 156.5 PA avg (135-198 range)
the CA avg was higher in 2050!

And here's the table of data...

columns 3 and 4 show the Attribute range...for many Attributes the 2050 lowest range is often all the way down to 1 (or 2-3) where 2018 is significantly higher
the percentage with a value below 10 is what columns 5 and 6 show; for many Attributes the amount of players with a value below 10 is significantly more in 2050 compared to 2018

columns 8 and 9 show the percentage of NHL with an Attribute value above 15; for most Attributes the number of players in this range is greater in 2050 than in 2016, in some cases much greater (though it's the opposite with Determination, Flair, Agility and Natural Fitness)

	2018 avg	2050 avg	2018 range	2050 range	2018 % below 10	2050 % below 10	2050 - 2018 =	2018 % 16-20	2050 % 16-20	2050 - 2018 =

Def Role	12.7	11.8	6-18	2-20	7.5%	25.3%	17.8%	9.7%	13.4%	3.7%
Off Role	15.5	16.0	7-20	2-20	0.3%	6.7%	6.4%	53.2%	68.3%	15.1%

Adaptability	11.7	11.5	5-18	1-20	12.1%	38.7%	26.6%	3.0%	25.0%	22.0%
Ambition	13.8	13.1	6-20	3-20	1.1%	24.5%	23.4%	20.2%	30.1%	9.9%
Determination	15.0	11.8	9-20	1-20	0.5%	31.7%	31.2%	37.1%	23.7%	-13.4%
Loyalty	13.2	11.3	6-20	1-20	3.0%	36.6%	33.6%	13.7%	21.8%	8.1%
Pressure	13.1	11.8	6-20	1-20	4.8%	29.0%	24.2%	10.5%	19.9%	9.4%
Professionalism	13.1	12.1	4-18	2-20	5.1%	31.5%	26.4%	12.6%	21.5%	8.9%
Sportsmanship	13.3	12.3	3-20	1-20	14.0%	30.4%	16.4%	22.6%	29.6%	7.0%
Temperament	14.1	14.4	4-18	1-20	33.6%	32.8%	-0.8%	9.4%	18.3%	8.9%

	2018 avg	2050 avg	2018 range	2050 range	2018 % below 10	2050 % below 10	2050 - 2018 =	2018 % 16-20	2050 % 16-20	2050 - 2018 =
AGGRESSION	10.2	4.4	1-18	1-20	47.3%	89.5%	42.2%	5.1%	3.0%	-2.1%
ANTICIPATION	13.1	13.7	7-20	1-20	6.2%	10.5%	4.3%	14.0%	16.7%	2.7%
BRAVERY	12.3	13.0	6-18	1-20	18.0%	17.2%	-0.8%	10.8%	27.4%	16.6%
Consistency	11.9	10.5	7-18	2-20	15.3%	43.5%	28.2%	4.8%	13.7%	8.9%
Decisions	12.6	10.1	6-18	2-20	9.1%	44.1%	35.0%	9.4%	5.1%	-4.3%
Dirtiness	7.1	10.5	2-17	1-20	80.4%	42.7%	-37.7%	1.1%	16.7%	15.6%
FLAIR	13.2	9.3	6-20	1-20	15.3%	52.4%	37.1%	26.9%	8.3%	-18.6%
Important Matches	11.6	13.8	6-18	1-20	19.1%	9.9%	-9.2%	5.4%	30.1%	24.7%
INFLUENCE/LEADERSHIP	11.9	13.9	5-20	1-20	28.2%	4.3%	-23.9%	16.1%	26.6%	10.5%
Pass Tendency	11.2	9.2	7-15	1-20	15.3%	50.4%	35.1%	0.0%	12.6%	12.6%
TEAMWORK	14.5	13.8	8-20	3-20	0.8%	11.3%	10.5%	26.0%	34.4%	8.4%
VISION/CREATIVITY	13.0	13.5	6-20	1-20	12.9%	11.8%	-1.1%	19.1%	27.7%	8.6%
WORK RATE	15.1	14.5	9-20	2-20	0.3%	7.5%	7.2%	41.9%	41.7%	-0.2%

	2018 avg	2050 avg	2018 range	2050 range	2018 % below 10	2050 % below 10	2050 - 2018 =	2018 % 16-20	2050 % 16-20	2050 - 2018 =
ACCELERATION	13.4	14.4	6-20	5-20	4.0%	11.6%	7.5%	14.0%	44.9%	30.9%
AGILITY	13.3	9.2	8-20	3-16	4.0%	53.0%	48.9%	13.2%	0.5%	-12.6%
BALANCE	12.2	10.1	7-17	5-20	6.7%	54.8%	48.1%	3.0%	7.3%	4.3%
Fighting	7.0	9.5	1-19	5-20	80.6%	62.6%	-18.0%	3.0%	7.8%	4.8%
HITTING	10.8	10.2	3-20	3-20	42.5%	55.9%	13.4%	11.0%	15.6%	4.6%
Injury Proneness	7.6	10.2	2-18	1-20	78.2%	42.7%	-35.5%	1.1%	15.1%	14.0%
Natural Fitness	13.4	9.5	7-20	2-15	3.5%	50.3%	46.8%	18.0%	0.0%	-18.0%
SPEED	13.5	14.6	7-20	5-20	2.4%	12.1%	9.7%	11.6%	46.0%	34.4%
STAMINA	13.3	15.3	7-18	6-20	3.2%	2.4%	-0.8%	15.9%	53.8%	37.9%
STRENGTH	11.4	10.2	5-19	2-20	27.2%	52.2%	25.0%	5.6%	13.2%	7.5%
Agitation	7.3	10.4	1-19	1-20	81.2%	41.1%	-40.1%	2.7%	13.2%	10.5%

	2018 avg	2050 avg	2018 range	2050 range	2018 % below 10	2050 % below 10	2050 - 2018 =	2018 % 16-20	2050 % 16-20	2050 - 2018 =
CHECKING	13.2	11.3	5-20	3-20	9.1%	41.7%	32.5%	15.9%	19.6%	3.8%
DEFLECTIONS	11.7	15.9	6-18	8-20	17.7%	1.9%	-15.9%	5.1%	55.9%	50.8%
DEKING	13.2	15.8	3-20	7-20	8.1%	3.0%	-5.1%	18.8%	57.0%	38.2%
FACEOFFS	10.8	12.2	4-18	1-20	39.5%	44.9%	5.4%	5.6%	38.7%	33.1%
MOVEMENT/OFF-THE-PUCK	12.5	16.4	6-20	5-20	13.2%	4.3%	-8.9%	9.9%	69.1%	59.1%
PASSING	14.4	14.8	11-20	7-20	0.0%	1.6%	1.6%	25.5%	38.2%	12.6%
POKECHECK	13.3	10.0	4-20	4-20	5.6%	48.1%	42.5%	13.2%	7.5%	-5.6%
POSITIONING	11.6	13.0	6-19	4-20	21.5%	16.4%	-5.1%	5.1%	25.5%	20.4%
SLAPSHOT	12.5	13.5	7-20	5-20	4.6%	9.4%	4.8%	6.5%	27.2%	20.7%
STICKHANDLING	13.8	13.4	8-20	7-20	2.4%	5.1%	2.7%	20.7%	21.8%	1.1%
Versatility	10.4	13.8	6-15	2-20	26.1%	5.6%	-20.4%	0.0%	26.3%	26.3%
WRISTSHOT	13.8	14.5	8-20	7-20	1.1%	3.8%	2.7%	17.7%	40.1%	22.3%

Post by **nino33** » Fri Jun 28, 2019 2:11 pm

FYI - there's an issue with the Temperament Attribute (below); some game created players and some database players (who have a "0" for Temperament) are receiving values in excess of 20 at startup. At first I thought it was an issue with the EHM Assistant not showing the Temperament data correctly, but I now know that's not the case.

I had adjusted the numbers to account for the 534 players who had a Temperament rating above 20 (534 total players, 67 with a CA above 100, 16 above 130) but now I now the adjustments I made are not correct. For example, column 4 was actually 4-23

Temperament

14.1

14.4

4-18

1-20

33.6%

32.8%

-0.8%

9.4%

18.3%

8.9%

P.S. I'm assuming "game created players" too because I couldn't find some of who I looked for in the database (I just went though a couple/few dozen players though); with the various characters involved from other languages I'm not always sure name searches are 100% trustworthy (does what the Assistant shows/what the Editor shows always exactly match what shows ingame?)

Post by **nino33** » Sun Jul 14, 2019 12:16 am

Last month I posted some testing results (above) using TBL 11.2
This quote gives a basic overview of what I did

nino33 wrote: Tue Jun 25, 2019 1:59 am What follows is from my test save using TBL 11.2, comparing the results from 2018 (startup) and in 2050

I took a look at the top 372 forwards in the NHL by CA (12 forwards x 31 teams = 372 forwards). Very late in the data compiling process I discovered that for 2050 I had accidentally included 6 defensive finesse defensemen, but I decided not to redo anything as I don’t think the small number significantly affects the results

I did the same thing for my testing using ECK 3.2
Here's the results...

	2018 avg	2049 avg	2018 range	2049 range	2018 % below 10	2049 % below 10	2049 - 2018 =	2018 % 16-20	2049 % 16-20	2049 - 2018 =

Def Role	13.3	13.5	6-20	3-20	3.2%	7.8%	4.6%	16.7%	22.6%	5.9%
Off Role	16.0	16.5	10-20	6-20	0.0%	2.2%	2.2%	66.4%	73.9%	7.5%

Adaptability	11.9	12.8	2-28	1-20	12.1%	24.2%	12.1%	4.0%	29.6%	25.5%
Ambition	13.5	14.4	6-20	4-20	2.7%	12.6%	9.9%	14.2%	44.9%	30.6%
Determination	14.8	13.1	6-20	1-20	0.5%	18.8%	18.3%	33.3%	32.3%	-1.1%
Loyalty	13.2	12.1	2-20	1-20	2.4%	29.3%	26.9%	15.6%	26.9%	11.3%
Pressure	12.8	12.9	4-20	3-20	5.4%	20.7%	15.3%	7.5%	26.9%	19.4%
Professionalism	12.7	13.2	5-20	4-20	8.1%	18.8%	10.8%	8.9%	31.2%	22.3%
Sportsmanship	13.1	13.3	5-20	1-20	12.9%	21.5%	8.6%	20.7%	36.6%	15.9%

	2018 avg	2049 avg	2018 range	2049 range	2018 % below 10	2049 % below 10	2049 - 2018 =	2018 % 16-20	2049 % 16-20	2049 - 2018 =
AGGRESSION	10.4	7.0	4-20	1-20	43.0%	68.5%	25.5%	4.8%	10.2%	5.4%
ANTICIPATION	13.9	14.0	7-20	4-20	1.1%	5.6%	4.6%	18.3%	32.8%	14.5%
BRAVERY	13.2	13.4	6-20	2-20	6.7%	16.7%	9.9%	17.2%	30.9%	13.7%
Consistency	13.3	11.6	6-18	3-20	3.0%	25.3%	22.3%	12.4%	14.2%	1.9%
Decisions	13.5	12.0	7-20	5-20	4.8%	15.9%	11.0%	14.5%	16.1%	1.6%
Dirtiness	6.9	9.9	2-20	1-20	82.8%	47.6%	-35.2%	1.6%	11.6%	9.9%
FLAIR	13.4	11.0	6-20	1-20	11.8%	41.7%	29.8%	27.2%	18.8%	-8.3%
Important Matches	12.3	14.7	7-20	5-20	12.6%	3.8%	-8.9%	9.1%	41.1%	32.0%
INFLUENCE/LEADERSHIP	12.3	13.5	6-20	1-20	19.9%	14.0%	-5.9%	16.4%	35.8%	19.4%
Pass Tendency	9.4	9.7	3-17	1-20	62.4%	47.6%	-14.8%	4.3%	13.4%	9.1%
TEAMWORK	15.0	14.7	11-20	1-20	0.0%	1.9%	1.9%	33.6%	40.3%	6.7%
VISION/CREATIVITY	14.0	14.1	7-20	6-20	5.4%	8.6%	3.2%	29.3%	34.1%	4.8%
WORK RATE	15.4	15.1	10-20	6-20	0.0%	4.6%	4.6%	47.6%	47.8%	0.3%

	2018 avg	2049 avg	2018 range	2049 range	2018 % below 10	2049 % below 10	2049 - 2018 =	2018 % 16-20	2049 % 16-20	2049 - 2018 =
ACCELERATION	14.3	15.9	8-20	5-20	1.1%	7.3%	6.2%	27.4%	63.2%	35.8%
AGILITY	14.1	10.6	8-20	5-18	2.4%	33.6%	31.2%	27.2%	3.0%	-24.2%
BALANCE	13.0	11.4	9-18	5-20	1.1%	25.0%	23.9%	7.8%	11.6%	3.8%
Fighting	6.5	10.2	1-17	4-20	83.9%	50.0%	-33.9%	2.2%	11.8%	9.7%
HITTING	10.9	11.1	3-20	4-20	36.3%	39.5%	3.2%	8.9%	16.4%	7.5%
Injury Proneness	7.6	9.2	2-19	1-20	77.2%	53.2%	-23.9%	1.3%	11.8%	10.5%
Natural Fitness	14.0	11.7	6-20	3-20	2.2%	24.7%	22.6%	25.8%	12.4%	-13.4%
SPEED	14.4	16.0	9-20	5-20	0.8%	6.2%	5.4%	27.7%	62.9%	35.2%
STAMINA	13.9	17.1	7-20	9-20	1.6%	0.3%	-1.3%	21.5%	76.3%	54.8%
STRENGTH	12.6	12.1	6-19	5-20	11.0%	29.0%	18.0%	15.9%	22.6%	6.7%
Agitation	7.4	9.8	1-18	1-20	80.6%	42.2%	-38.4%	2.4%	9.7%	7.3%

	2018 avg	2049 avg	2018 range	2049 range	2018 % below 10	2049 % below 10	2049 - 2018 =	2018 % 16-20	2049 % 16-20	2049 - 2018 =
CHECKING	13.0	11.7	7-20	4-20	8.3%	35.5%	27.2%	15.3%	22.6%	7.3%
DEFLECTIONS	11.3	16.2	5-20	9-20	27.4%	2.2%	-25.3%	7.0%	59.4%	52.4%
DEKING	12.6	15.3	4-20	6-20	18.5%	10.5%	-8.1%	19.9%	52.4%	32.5%
FACEOFFS	10.5	12.8	3-20	4-20	45.2%	40.1%	-5.1%	7.3%	38.4%	31.2%
MOVEMENT/OFF-THE-PUCK	12.4	16.8	6-20	9-20	15.3%	0.3%	-15.1%	9.9%	69.6%	59.7%
PASSING	14.0	15.2	9-20	6-20	0.5%	2.7%	2.2%	26.3%	46.2%	19.9%
POKECHECK	13.4	10.4	6-20	2-20	3.0%	37.9%	34.9%	12.9%	6.2%	-6.7%
POSITIONING	12.7	14.1	6-20	4-20	8.6%	12.6%	4.0%	7.8%	39.2%	31.5%
SLAPSHOT	11.4	14.0	5-20	7-20	20.4%	7.0%	-13.4%	2.7%	30.6%	28.0%
STICKHANDLING	13.6	14.2	7-20	6-20	5.6%	5.1%	-0.5%	24.5%	31.5%	7.0%
Versatility	10.2	14.6	5-15	3-20	33.1%	5.1%	-28.0%	0.0%	46.8%	46.8%
WRISTSHOT	13.5	14.8	7-20	7-20	4.8%	5.1%	0.3%	20.7%	41.7%	21.0%

here's the some basic data on the two groups

the 2018 players had a 142.5 CA avg (116-196 range) and a 150.4 PA avg (116-200 range)
the 2050 players had a 159.0 CA avg (127-198 range) and a 166.1 PA avg (140-200 range)
the CA avg was much higher in 2050 (and the CA/PA averages are all higher in ECK, especially the future CA average)

Here's a reminder of what it was with TBL 11.2

nino33 wrote: Tue Jun 25, 2019 1:59 amAnd here's the some basic data on the two groups
the 2018 players had a 141.5 CA avg (126-190 range) and a 146.8 PA avg (126-198 range)

the 2050 players had a 148.7 CA avg (128-198 range) and a 156.5 PA avg (135-198 range)

the CA avg was higher in 2050!

Post by **nino33** » Sun Jul 14, 2019 1:22 am

nino33 wrote: Tue Jun 25, 2019 1:59 am What is evident is with the game created values in 2050 (when any human inputted values are long gone) there are some really significant differences in the frequency of lower/higher Attribute values compared to startup

Here's the Attributes where the AI is significantly (5 or more) lower than the human entered values, from the TBL data above
Defensive Role
Offensive Role
Determination
Loyalty
Pressure
Anticipation
Bravery
Consistency
Flair
Important Matches
Pass Tendency (hmm...that's interesting)
Teamwork
Vision/Creativity
Work Rate
Agility
Natural Fitness

I do wish that such high CA players did not get such extremely low ratings (ratings that human researchers don't give)
The ones that I think would have the most impact are

the Offensive/Defensive Roles
Determination, Anticipation, Work Rate (all visible ingame too)
maybe/probably Consistency too

All thoughts/comments/feedback is welcome and appreciated!

Post by **nino33** » Fri Aug 02, 2019 6:13 pm

nino33 wrote: Tue Jun 25, 2019 1:59 amWhat is evident is with the game created values in 2050 (when any human inputted values are long gone) there are some really significant differences in the frequency of lower/higher Attribute values compared to startup

nino33 wrote: Tue Jun 25, 2019 1:59 amI do wish that such high CA players did not get such extremely low ratings (ratings that human researchers don't give)

There's some good related discussion about these things here (July 15th/16th) https://www.ehmtheblueline.com/forums/v ... &start=525 & also here (from the last few days) https://www.ehmtheblueline.com/forums/v ... 29#p241229

Something that just occurred to me, and I'm surprised it hasn't before (feels like a "d'oh!" moment HaHa)...
Looking at the expected average attribute by CA chart...

CA AVG
200 16
190 16
180 15
170 14
160 14
150 13
140 12
130 12
120 11
110 10
100 10
90 9
80 8
70 8
60 7
50 6
40 6
30 5
20 4
10 4
1 3

In regards to those Attributes that are most connected to CA development, seems to me that quite clearly not every increase of 10 CA will necessarily lead to a noticeable increase in these Attributes

It's interesting that from below NHL average (130) to about NHL average (140) doesn't have an expected Attribute average increase (and of course it happens at other times too)...I wonder if at such times some people might think their players aren't developing or development's broken

P.S. What Attributes are most connected to CA development? Here's what i think...

Technical Attributes for sure.
Physical Attributes too, but it's not necessarily the same as the Technical Attributes (for example, I think skating connects to age too).
Mental Attributes less though (in particular because 7 of the 8 unchanging Attributes are Mental Attributes, as are all 5 of the Attributes that don't change much)

And other things have an impact too (player/team performance, injuries, coaching, etc)