Let's Test NHL 16

[AdamJones113]

How much does X affect Y? How will the gameplay be affected if X turns into X+1? This thread will channel my series of (mostly gameplay) tests on NHL 16, in the hope that people use it to improve their own game play and perhaps even more than that.

Slider Effectiveness List
Injuries: Works (Tested CPU & Human)
Wrist/Slap Shot Power: Works (Tested CPU and Human)

Test 1: Passing Accuracy Slider

Spoiler

Let's use tonight's matchup, Senators (CPU) at Leafs. For this test, we'll play a series of three games on all default AS/Hardcore sliders (5 minute periods), with one change: in game 1, the passing accuracy sliders will be minimized; game 2 will have the passing accuracy slider maxed; game 3 will have the passing accuracy slider at default.

Game 1: Passing Slider Accuracy Minimized (CPU and Human)
Senators Passing %: 76
Leafs Passing %: 61
Notes: I played poorly. Tried to force some passes, as I fell down early and stayed down (lost 4-0). CPU missed some passes they probably should have made, but also made some of their trademark incredible passes.

Game 2: Passing Slider Accuracy Maximized (CPU and Human)
Senators Passing %: 81
Leafs Passing %: 54
Notes: I played about the same skill level in this one, losing 5-1. My main defensive troubles came with my piss-poor defensive AI, which left men open in the slot left and right... and a shaky, shaky netminder.

Game 3: Passing Sliders in the Middle (CPU and Human)
Senators Passing %: 71
Leafs Passing %: 62
Notes: I played well this game, winning 2-1. It was a solid game overall, but the Senators' lone goal came on a hella-lucky pass through 3 defensemen.

Test 2: Player Separation Over an 82-Game Season

Spoiler

Test 2: Player Separation Over the Course of a Season

I think most of us dislike the in-game player separation. It's hard to tell, on-ice, that X player is better than Y, when that same comparison is so clear in real life. Yet at the same time, it's been pointed out that the right players get the higher point totals through an 82-game simulated season. The counter to that has been that those players are getting the higher point totals because of their high ice time. Let's see what happens when the Philadelphia Flyers play one season with normal lines, and one season with the lines completely reversed.

I'm not focusing on NHL results (I will note the Stanley Cup winner each time), but on individual player statistics. No playoff comparison; that's hardly different enough in game to merit a comparison. Injuries are off to maintain the exact same lines through the course of the season. Assistant Coach Edit Lines is set to 'No' for the same reason.

2015/2016 Philadelphia Flyers: Normal Lines
Click 'Spoiler' to see all the Flyers' lines.

Spoiler

Offense (Overall):
Simmonds (88)—Giroux (90)—Voracek (90)
Raffl (82)—Schenn (86)—Read (83)
Lecavalier (83)—Couturier (86)—Umberger (82)
Bellemare (79)—Gagner (83)—White (82)

Defense:
Streit (87)—MacDonald (85)
Del Zotto (83)—Schenn (84)
Schultz (82)—Gudas (83)

PP Line 1:
Voracek—Giroux—Simmonds
Streit—Del Zotto

PP Line 2:
Couturier—Schenn—Gagner
MacDonald—Schenn

4 Man PP1 (and 4 on 4 #1):
Giroux—Voracek
Streit—Del Zotto

4 Man PP2 (and 4 on 4 #2):
Schenn—Simmonds
MacDonald—Schenn

3 on 3:
Giroux—Streit—MacDonald
Couturier—L. Schenn—Del Zotto
B. Schenn—Gudas—Schultz

XAttacker:
Giroux
Voracek

Let's take a look at the player stats, in descending order of points.

Philadelphia Flyers Stats - 2015/16
Skaters	GP	G	A	PTS	+/-	GW	PIM	SOG	SPCT
J. Voracek	82	31	40	71	7		14	266	11.65
C. Giroux	82	27	40	67	4		10	257	10.51
M. Streit	82	7	41	48	0		14	165	4.24
S. Couturier	82	23	21	44	17		40	149	15.44
W. Simmonds	82	21	22	43	9		42	193	10.88
B. Schenn	82	19	23	42	0		67	199	9.55
S. Gagner	82	16	19	35	-13		4	138	11.59
M. Del Zotto	82	2	32	34	-3		8	142	1.41
V. Lecavalier	82	12	20	32	5		24	125	9.60
R. Umberger	82	9	20	29	8		34	127	7.09
A. MacDonald	82	2	24	26	5		38	161	1.24
M. Read	82	11	14	25	-4		8	118	9.32
M. Raffl	82	10	12	22	-6		2	118	8.47
P. Bellemare	82	11	10	21	-8		10	98	11.22
R. White	82	6	13	19	-2		36	109	5.50
R. Gudas	82	3	13	16	10		46	51	5.88
L. Schenn	82	4	11	15	6		54	117	3.42
N. Schultz	82	1	7	8	2		44	50	2.00

This team made the playoffs as a wild card, swept Detroit in the first round, beat Montréal in 5 games in the second round, and fell to Pittsburgh in 6 games in the ECF. Pittsburgh, in turn, fell to the Vancouver Canucks in 7 games in a very good SCF.

2015/2016 Philadelphia Flyers: Reversed Lines


Click 'Spoiler' to all the Flyers' lines.

Spoiler

Offense (Overall):
Bellemare (79)—Gagner (83)—White (82)
Lecavalier (83)—Couturier (86)—Umberger (82)
Raffl (82)—Schenn (86)—Read (83)
Simmonds (88)—Giroux (90)—Voracek (90)

Defense:
Schultz (82)—Gudas (83)
Del Zotto (83)—Schenn (84)
Streit (87)—MacDonald (85)

PP Line 1:
Bellemare—Gagner—White
Schultz—Gudas

PP Line 2:
Raffl—Couturier—Read
Del Zotto—Schenn

4 Man PP1 (and 4 on 4 #1):
Gagner—White
Streit—Del Zotto

4 Man PP2 (and 4 on 4 #2):
Couturier—Bellemare
Del Zotto—Schenn

3 on 3:
Bellemare—Schultz—Gudas
Couturier—L. Schenn—Del Zotto
B. Schenn—Gudas—Schultz

XAttacker:
Bellemare
White

Here are the team stats for the last-place Philadelphia Flyers:

Philadelphia Flyers Stats - 2015/16
Skaters	GP	G	A	PTS	+/-	GW	PIM	SOG	SPCT
S. Couturier	82	19	36	55	-22		35	162	11.73
M. Read	82	23	23	46	4		2	152	15.13
S. Gagner	82	18	25	43	-19		17	192	9.38
R. White	82	18	20	38	-17		88	165	10.91
B. Schenn	82	23	14	37	7		57	142	16.20
M. Raffl	82	10	27	37	5		10	158	6.33
J. Voracek	82	16	15	31	9		6	122	13.11
M. Del Zotto	82	6	24	30	-24		7	150	4.00
P. Bellemare	82	11	19	30	-23		6	173	6.36
R. Gudas	82	2	27	29	3		105	89	2.25
C. Giroux	82	9	20	29	6		2	118	7.63
R. Umberger	82	14	15	29	-16		41	119	11.76
A. MacDonald	82	3	21	24	-5		22	101	2.97
V. Lecavalier	82	13	10	23	-17		16	122	10.66
W. Simmonds	82	7	16	23	3		32	91	7.69
N. Schultz	82	2	15	17	5		60	109	1.83
M. Streit	82	4	11	15	-14		2	88	4.55
L. Schenn	82	1	13	14	-21		67	122	0.82

In conclusion, the simulation engine works pretty well for player and stat differentiation. Increased ice time for worse players saw their stats jump and lowered ice time for better players led to a decrease in their stats, but the overall point totals were much lower, so that's good.

Let me know what you think I should test next.


Disclaimer: by all means, take these tests with a grain of salt. I encourage you to do your own tests and post them here. It's perfectly acceptable to disagree with what I've posted, and don't hesitate to post those disagreements. Remember, however, that I am not making these tests with an agenda. I, like you, want the best possible NHL game.

[Smallville102001]

Quote: Originally Posted by AdamJones113

How much does X affect Y? How will the gameplay be affected if X turns into X+1? This thread will channel my series of (mostly gameplay) tests on NHL 16, in the hope that people use it to improve their own game play and perhaps even more than that. Test 1: Passing Accuracy Slider Let's use tonight's matchup, Senators (CPU) at Leafs. For this test, we'll play a series of three games on all default AS/Hardcore sliders (5 minute periods), with one change: in game 1, the passing accuracy sliders will be minimized; game 2 will have the passing accuracy slider maxed; game 3 will have the passing accuracy slider at default. Game 1: Passing Slider Accuracy Minimized (CPU and Human) Senators Passing %: 76 Leafs Passing %: 61 Notes: I played poorly. Tried to force some passes, as I fell down early and stayed down (lost 4-0). CPU missed some passes they probably should have made, but also made some of their trademark incredible passes. Game 2: Passing Slider Accuracy Maximized (CPU and Human) Senators Passing %: 81 Leafs Passing %: 54 Notes: I played about the same skill level in this one, losing 5-1. My main defensive troubles came with my piss-poor defensive AI, which left men open in the slot left and right... and a shaky, shaky netminder. Game 3: Passing Sliders in the Middle (CPU and Human) Senators Passing %: 71 Leafs Passing %: 62 Notes: I played well this game, winning 2-1. It was a solid game overall, but the Senators' lone goal came on a hella-lucky pass through 3 defensemen. Let me know what you think I should test next. Disclaimer: by all means, take these tests with a grain of salt. I encourage you to do your own tests and post them here. It's perfectly acceptable to disagree with what I've posted, and don't hesitate to post those disagreements. Remember, however, that I am not making these tests with an agenda. I, like you, want the best possible NHL game.

Interesting while I know sports games not ever game is going to be the same has the game before the fact that u had a high pass % with the pass slider at zero then it did with the slider max just shouldn't happen ever! With the slider at zero passes should be all over the place. With the slider at max they should be like tap to tap. Do they even keep track of pass % in real hockey? I really only watch the sharks in hockey more causal hockey guy but I know the puck is miss handled way more in real life then you see in this game.

[jake19ny]

This one single test is further evidence that sliders do virtually nothing. Look at CPU passing:

76% on minimum
81% on maximum
71% on default

It's an overall differential of 10% and does not reflect the sliders from lowest to highest. The differential appears to come from how the user plays as the the game where user played best the CPU had the lowest %.

[Vikes1]

It seems to me that the base programming decides when and just how much your slider input will have an effect. Ie, sliders ignored at times to try to force a certain outcome. Just a thought.

[AdamJones113]

I ran a second test for more of a direct comparison between Max and Min passing accuracy. 4 minute periods, same other settings as before.

Game 1: Leafs @ Blue Jackets (Passing Accuracy Minimized)
Leafs (Human) Passing %: 59
Blue Jackets (CPU) Passing %: 75
Notes: A solid 3-1 win by me, a very nice dirty-area goal from Arcobello to seal it.

Game 2: Leafs @ Blue Jackets (Passing Accuracy Maximized)
Leafs Passing %: 62
Blue Jackets Passing %: 79
Notes: Also a solid game, a 2-1 loss. I had some really good cycling but couldn't capitalize enough.

[jake19ny]

OK you inspired me to do a test of my own and learned some very interesting things. I lowered all but a few sliders to the minimum (0) and I'm posting the results of what I saw different as opposed to the regular sliders I use and default sliders;

First lets start with the sliders I used.

Everything at zero except the following

Attribute Effect 5
Game Speed 2
Player Acceleration 3
Pass assist 5
Pass speed 2
Faceoff Difficulty 3
Shot Power 3
Slap shot power 4

Game results of NYR (Me) vs Flyers (CPU) 10 minute periods Superstar/Custom

Score: Flyers 10 NYR 2 (The CPU was unstoppable juggernaut)
Shots: Flyers 56 NYR 30
Hits: Flyers 14 NYR 26
Pass% Flyers 74% NYR 66%
PP Flyers 2/4 NYR 0/0

The other stuff was irrelevant

Summary

The game had a few good things to like. there were a lot of loose pucks and missed passes, but the CPU still had those Superhuman nothing you can do to stop us moments.

Sliders that had absolutely zero to little difference on minimum as opposed to default or the sliders I use

Fatigue Effect: This seemed to do nothing to the performance of the CPU but had a slight effect on human controlled players. Mainly a placebo slider

Puck Control: HUMAN ONLY, even at zero the CPU still knocked you right off the puck same as default or my sliders. CPU was tougher but I saw more loose pucks overall.

Hitting Assistance and Hit Power: I was able to hit exactly like any other game I played on Default or my usual sliders. As did the CPU

Aggression: At zero I saw very little decline in the Aggressiveness of the CPU players on both sides.

Poke Effectiveness: Saw no difference in poke checking on zero as a well timed poke still knocked the puck away. Spamming however did not increase your chances.

Stick Lift Effectiveness: Same as Poke Check except I noticed the CPU just didn't really attempt it too much.

Pass Accuracy: CPU ONLY, tape to tape passing was no problem at zero for the CPU and when they went into superhuman mode it was tic tac toe goal....every time

Pass Interception: CPU ONLY, even at zero the CPU picked off a ton of my passes.

Pass Reception Ease: CPU ONLY, I did see more loose pucks and some miss handled passes but as with pass accuracy the CPU would put together plenty of tape to tape magnetic passes.

Goalie Reaction Speed: Saw very little difference in how the goalies played. I did give up 10 goals but I think that had to do with the huge differential you can see above in how sliders effect the Human over the CPU. I only scored 2 goals on 26 shots so the CPU goalie did just fine.

As you can see a large amount of very important sliders have very little to no impact on the game, mainly how the CPU plays and performs.

Sliders that actually had an impact on zero:

Fatigue Recovery: Both sides fatigue levels went down and on zero to low red for all lines and were unable to go back up, however this seemed to have no impact on the performance of the CPU team what so ever so it could easily have been listed in sliders that do nothing CPU ONLY

NEXT UP: All the sliders I tried at zero I will try at MAX and I will post my results and I suspect the list of sliders that do nothing to little in this game will increase. This game basically has a built in CPU AI which I think are the EA tuners and all the tweaking in the world will net you little results with sliders that are merely their to try and make us feel warm and fuzzy. Were on to you EA......

[AdamJones113]

Good stuff Jake! One thing I heavily recommend, though, is that experiments have more weight with more controlled variables. We don't know how X variable affects Y, so that's why experiments only changing one variable (slider) are a little more soundproof.

But by all means, keep posting that experiment! Looking forward to see what you get.

[jake19ny]

Quote: Originally Posted by AdamJones113

Good stuff Jake! One thing I heavily recommend, though, is that experiments have more weight with more controlled variables. We don't know how X variable affects Y, so that's why experiments only changing one variable (slider) are a little more soundproof. But by all means, keep posting that experiment! Looking forward to see what you get.

I get what your saying and I can almost tell which sliders interact with each other by doing this but your right a more accurate way is one slider at a time. Not sure I have the patience for that LOL. Over all I saw very little difference in the CPU in my game played with sliders minimized and I suspect I will see little difference with sliders maximized meaning they do very little to nothing. Your passing experiment is showing the same. Looking forward to your next one.