FHM2 Simmed Season

[greenOak]

So I simmed a season, and thought I'd compare the statistical results to previous seasons in the NHL.

LEAGUE LEADERS

Points
FHM2: 93,89,87,86,85
2014-2015: 87,86,84,81,81
2013-2014: 104,87,86,84,82

# of players with 80+ Points / # of players with 70+ Pts / 60+ Pts / etc
FHM2: 8,18,51,95,173
2014-2015: 5,19,53,89,154
2013-2014: 7,18,49,92,158

Goals
FHM2: 51,45,41,41,40
2014-2015: 53,43,42,38,37
2013-2014: 51,43,41,39,37

# of players with 40+ Goals / # of players with 30+ Goals / 20+ Goals / etc
FHM2: 5,27,87,255
2014-2015: 3,15,92,267
2013-2014: 3,21,101,258

Assists
FHM2: 63,58,58,54,51
2014-2015: 60,59,56,56,55
2013-2014: 68,65,61,58,56

# of players with 60+ Assists / # of players with 50+ Assists / 40+ Assists / etc

FHM2: 1,6,34,118,245
2014-2015: 1,6,37,105,234
2013-2014: 3,9,35,110,226

PIM
FHM2: 172,158,143,134,130
2014-2015: 238,191,175,172,148
2013-2014: 213,211,209,166,153

Shots
FHM2: 351,338,320,314,306
2014-2015: 395,304,302,292,280
2013-2014: 386,313,305,294,285

Save Percentage (Min 35 Starts)
FHM2: .923,.921,.920,.920,.917
2014-2015: .933,.929,.928,.926,.925
2013-2014: .930,.927,.927,.926,.924

SCORING/SHOTS BY DEFENSEMAN

Points
FHM2: 71,69,67,59,59
2014-2015: 66,60,60,56,55
2013-2014: 74,61,56,56,55

# of players with 60+ Points / # of players with 50+ Pts / 40+ Pts / etc
FHM2: 3,10,33,66,108
2014-2015: 3,11,27,53,96
2013-2014: 2,9,26,52,90

Goals
FHM2: 20,18,17,17,16
2014-2015: 23,21,18,17,16
2013-2014: 23,22,20,20,17

# of players with 20+ Goals / # of players with 15+ Goals / 10+ Goals /
FHM2: 1,8,40,117
2014-2015: 2,10,32,82
2013-2014: 4,6,28,90

Shots
FHM2: 224,222,219,216,214
2014-2015: 292,264,245,238,237
2013-2014: 257,256,245,241,208

MISCELLANEOUS

ATOI - Forwards
FHM2: 23:04,22:56,22:51,22:30,22:30
2014-2015: 20:40,20:38,20:33,20:31,20:19
2013-2014: 21:58,21:48,21:17,21:14,21:03

ATOI - Defenseman
FHM2: 24:56,24:43,24:43,24:43,24:37
2014-2015: 29:03,28:59,27:15,26:28,26:22
2013-2014: 29:24,27:04,26:57,26:54,26:25

Plus-Minus
FHM2: 47,42,42,39,37
2014-2015: 38,38,33,31,30
2013-2014: 39,38,36,34,34


CONCLUSIONS

Long story short, FHM2's sim engine excels in some areas, and not in others. When it comes to goals, assists, and points, FHM2 is sublime, generating not just accurate league leaders, but what appears to be a very accurate distribution of how points are scored in the NHL. Moreover, the top forwards and top defenseman should expect to score a similar amount of points as they would in the modern NHL. The only thing that seems to be off is that it would appear that mid-bottom pairing defenseman produce more offense in FHM2 than their real life counterparts do. Maybe too many 30+ goal scorers? This may be explained by the fact that the top defenseman in FHM2 don't get as much ice time as they do in real life (and thus depth defenseman get more playing time). It should be noted that global scoring, although not explicitly mentioned above, was extremely close to modern NHL totals.

One disappointing result I found, was the low save percentages - a problem that existed in the original FHM. The league wide save percentage in FHM2 was .908 as opposed to the .915 and .914 observed the past two years in the NHL. The culprit seems to stem from too few shots per game. The previous two years in the NHL have averaged 29.8 and 30.0 shots per team. The simulated season in FHM2 averaged just under 29.0 shots per team. This may not seem like a lot, but it explains the low save percentages in FHM2.

As for penalty minutes, the leaders seem to get fewer penalty minutes in FHM2 than in the real NHL. We can see whether or not this is due to too few minor penalties or too few major penalties (or both).

Minor Penalties
FHM2: 31,29,28,28,28
2014-2015: 48,42,39,39,39
2013-2014: 47,45,41,41,38

Major Penalties
FHM2: 10,10,7,7,7
2014-2015:19,17,16,15,13
2013-2014:19,16,15,15,15

So it would seem as there are both too few minor and major penalties among the PIM leaders in the FHM2 universe.

Finally, top forwards seem to get more ice time than they do in the NHL, while top defenseman get less.

The last thing I want to mention, is that the sim engine appears to give similar results for all leagues, so don't expect realistic results from say the major junior leagues for example.

[dave1927p]

Very good summary. Remember that you can tweak the goals per game for playable leagues. So if the OHL, WHL, or QMJHL point totals are too low. Then just increase it in your game but not too much as you don't want the goalies save percentage in the juniors to get too low.

[JeffR]

Originally Posted by greenOak

The last thing I want to mention, is that the sim engine appears to give similar results for all leagues, so don't expect realistic results from say the major junior leagues for example.

Not sure how you got that idea, but the CHL leagues in the game typically range from at least 0.5 to 1.5 goals per game higher than the NHL, with close to twice as many penalty minutes in some cases. If you check the league histories, you'll see they follow their recent historical averages, and the same for the non-junior leagues.

[Markus Heinsohn]

This looks good to me. A bit tweaking to the number of (saved) shots and the stats are spot-on.

[sprague]

This is a good summery.

When it comes to penalties, you should also look at is it low because the main penalty players are not getting ice time by the AI hence the low totals for leaders (which then is an issue of not getting those players playing time)

or are those players dressed and playing just not recieving penalties, that then is a bit of an engine tweek of how penalties are handled. You might want to take a look at who the top 10 penalty players in the game, and then list the top 10 players from last year and what their game totals were. That could help the developers I would think.

[Cod]

Great summary; however, I think you should've simmed more than one season before making a conclusion. One season just doesn't give enough insight into what's going on behind the scenes.

[Sebastian Palkowski]

Originally Posted by Cod

Great summary; however, I think you should've simmed more than one season before making a conclusion. One season just doesn't give enough insight into what's going on behind the scenes.

This!

But I like the original post. I hope we get a lot more stats analyses over time so we can further improve the Match Engine. Our researcher so a lot of analyses as well but the more people look, the better the game becomes. So keep up the good work!

[greenOak]

Well I've simmed through the 2018-2019 season now, and although I didn't deeply analyze each seasons stats, I still stand by most of what I said in the original post.

I questioned whether or not there were too many 30 goal scorers, but this does not appear to be true, as the next few seasons had ~20 30+ goal scorers, which is accurate.

League wide save percentages have ranged from .907 to .911.

The TOI issues I talked about were seen in every season.

PIM leaders have been low in every season.

Great summary; however, I think you should've simmed more than one season before making a conclusion. One season just doesn't give enough insight into what's going on behind the scenes.

Just to get technical here, but having X number of seasons isn't really a good criteria to use. Stuff like short handed goals, plus minus would likely require a lot of seasons to infer anything meaningful, but certain stats should converge very fast (ie save percentage).

Not sure how you got that idea, but the CHL leagues in the game typically range from at least 0.5 to 1.5 goals per game higher than the NHL, with close to twice as many penalty minutes in some cases. If you check the league histories, you'll see they follow their recent historical averages, and the same for the non-junior leagues.

This is correct. I didn't look at anything in major junior except for CHL league leaders which were low (league leader for points in all three leagues had 90-95 points). I don't know whether or not that season was an abberation, or was a result of the initial database, but the results look a lot better for the 2018-2019 season. For those who are interested, here is a look at scoring distribution in the major junior league for 2018-2019:

# of players with 110+ Points / # of players with 100+ Pts / 90+ Pts / etc

WHL

FHM2: 1,6,8,21,35,58
2014-2015: 1,3,6,19,29,49
2013-2014: 2,5,13,20,38,61

OHL

FHM2: 3,4,9,16,27,54
2014-2015: 4,9,12,18,30,47
2013-2014: 1,3,11,21,36,50

QMJHL

FHM2: 4,7,9,15,26,42
2014-2015: 1,4,12,21,35,44
2013-2014: 1,5,7,10,22,39

FHM2's results seem very good in this regard. FWIW the league leader in this sim for points in the WHL, OHL, and QMJHL, had 122, 123, and 136 points respectively.

or are those players dressed and playing just not recieving penalties, that then is a bit of an engine tweek of how penalties are handled. You might want to take a look at who the top 10 penalty players in the game, and then list the top 10 players from last year and what their game totals were. That could help the developers I would think.

From my 2018-2019 season:

LEAGUE LEADERS IN PIM

FHM2

1. 191, 78GP
2. 129, 82GP
3. 128, 82GP
4. 122, 82GP
5. 116, 82GP

2014-2015:

1. 238, 72GP
2. 191, 82GP
3. 175, 79GP
4. 172, 67GP
5. 148, 80GP

2013-2014:

1. 213, 77GP
2. 211, 76GP
3. 209, 81GP
4. 166, 58GP
5. 153, 67GP

So it doesn't seem like games played is the issue here. If anything, it looks like the NHL players are getting more PIM in fewer games than their FHM2 counterparts. The FHM leaders appear to be getting more TOI as well.

[ebersbach]

How do I simulate a season?

[dave1927p]

Originally Posted by ebersbach

How do I simulate a season?

under the PLAY tab on the top, you will see a title called "AUTO-PLAY UNTIL..." ---You can select specified date there. Or just the right of that, select the bookmarked date (still under that same PLAY tab) and it will simulate to that point.