Power play % and other thoughts

[jwr38]

But, saying there is a problem after making 20 test seasons isn't the best way to find these kind of problems, because if you use all the auto functions for your roster and lines, you may get 3 guys with high SC on the 1st line, and it's no secret that they won't be in the top scorers because they have all 33% chances of scoring a goal.

I agree with you here. tests seasons that I have done aren't the best way of seeing if the best scoring guys reach the top in scoring. But since I created a player creator I am able to have a range of players and skill sets. I am not only doing just 10 seasons. I do 10 seasons with each set and then redraft all the players and do another 10 seasons. Then create more players and start all over again. So far by doing this there is nothing but randomness. One season a player who scores points get a +/- +20 while the next -20. Nothing has changed, same line mates, same schedule, all extra randomness I can taken out, same on ice time. This is not just happening to one player either but many in the sim. While in any sim you can't expect them to be the same I do expect to see some consistency. What is making a player go from one season as a +20 to a -20 the next? This is not only with the -/+ stat but I have seen this with others as well. This is the reason I have asked simon if he is still tweaking the engine. If he is not it then there is no point is relating all the things I am finding out. But since this is still a beta and I am sure Simon wants version 2 to be better than version 1 .......

[GHL Commish]

I'm curious how you find V2 more realistic than V1 when focusing on the overly insane level of randomness. Simply put, V2 has better features added into it, however, the confusing level at which powerhouse teams are screwed over and teams that should be decent at best are winning makes V1 more realistic in that sense.

No one is saying that Simon's product is crap. But the fact remains, V2 needs to be corrected so that it is a much improved version of his original product. Not a more flaky product with a better number of features.

[Foo]

I also finished my first season (44 games) with V2.

2 85 SC guys are on top while Ovechkin struggles with 8.. yes 8 goals in 35 games. (31 points)

Crosby? well 12 goals, 30 points in 38 games. 97 of PA and 80 of SC.

This season was simed with 0.9.9 version.

This week Simon released the 0.9.10 version with those modifications:

•Lower the PP by a little.
•Small modification regarding players scoring abilities.

While the first one is clear, I'm wondering if the small modification is the result of this topic, I would like to ear Simon about that.

I really like V2, love the random thing and all, but I have to agreed with 90% of the guys here, randomness take too much place in the result of the sim, regarding mostly the high rated players in scoring.

Keep randomness at a reasonable level, and let the superstar score some goals!!!

[sportsceo]

I just bought the V2 because of the Quality of the V1. The best part of Simon's product is the Realism and I can only go from what everyone is saying but if the V2 Sim really is as random as people are saying I will be staying with the V1 for many obvious reasons. Players and Teams underachieving is one thing but I believe in the thinking that if you as a GM make good trades, make the proper lines and adjustments throughout the season taking care of your team that your team should do prosper more than a team that doesn't deal with their lines or makes bad trades. If it really is so random what's the point of being a good involved GM?

I love all the new features especially where you get to control the minutes played per line and how the defense and offense are separate but none of it makes any difference if the Sim doesn't allow the GM's who actually care for their team to get an advantage.

Maybe in a future version if you could actually have an Option of how random you want the Sim to run would work best so those of us who want our GM's to be rewarded for taking care of their team can have it one way and those who love the crazy randomness can have it another.

Regardless of what happens i'm just happy that theirs someone out their like Simon with the Sim and Michael with the Portal who have put such superior products out their on the market for all of us.

[bond007]

I agree with most of you.

I tried so many ratings with differents settings during the past 8 months and I have never been able to get realistic statistics. Guys like Crosby, Ovechkin,... are never the best players. I am not sure if it's randomness since I see some average players (on average team but with quality ice time.... like Matt Stajan on 1st line and 1st PP) getting season after season better statistics that the superstars. Also, even if I lowered their EN, they keep performing really well. They might miss games because of exhaustion but the P/20 minutes is always way too high, especially for an exhausted player.


Version 0.9.10 seems a bit better, but in my opinion it still can be improved. I really want to use V2 with all the new features, but with results like those I get, I have to stay with V1.

[SimonT]

Quick and Dirty Answer :
Players with 99 PA and 99 of SC won't be your best scorer in your league because he'll pass as much that he shot.
Players with 50 PA and 99 of SC might be your best scorer in your league because he'll not pass and just shot.

[Foo]

Well that's a problem at first.... and that's good to know.

so what's the difference between an 80 PA and 80 SC with an 99 PA and 99 SC ?

I do my own ratings, and my SC doesn't have any link with the shots on goal... but just pure natural goals... 99 SC should have better accuracy then a 75 SC, I hope it is the case. SC mean shot control in fact.

same thing with the PA... the rating title says it by itself, Pass Accuracy... he his more probable to have an assist on a goal then a 55 PA who will miss the pass often or be intercepted more... I hope it is not only 99 = pass often, 50= don't pass the puck at all...

... if so, I wonder how all the ratings are interpreted by the sim.

Here's my top 5 pointer:

PA -- SC

80 -- 85
89 -- 72
75 -- 71
74 -- 74
78 -- 88

No Thornton, No Ovy, No Crosby, No Malkin, No Datsyuk, etc. etc. etc.

[SimonT]

Players make decision on what to do with punk regarding a formula. As you can guest easily, SK is a important part of Skating Deicision, PA is a important part of the Passing Decision and SC is a important part of Shooting Decision.

[Foo]

Well, indeed, but a 99-99 should be the top scorer of the team cause is the total package! that seem a nonsense to have a 50-99 better then a 99-99.. I talk about point totals and not only goals scored.

[SimonT]

In my test, a 99-99 will get a lot of point. But won't be "top scorer" (Goal only speaking) for the league.

Quick and Dirty Answer :
Players with 99 PA and 99 of SC won't be your best scorer in your league because he'll pass as much that he shot.
Players with 50 PA and 99 of SC might be your best scorer in your league because he'll not pass and just shot.

I was speaking goal only! Not Point!

[hockeytown5]

I also find the sim much too random. There seems to be no consistency in the standings as teams finish all over the map from season to season. I just purchased this program after using FHLSim in our league for a long time. I still intend to use it eventually but I'm leaning towards putting it off until this problem gets fixed.

[Foo]

Sincerly it's not that bad, the scoring of star player must be corrected for sure, but the engine is still the greatest . Top teams should be in playoffs if you have a good mix of players. In my league, only one team (the ovechkin one) that was supposed to be part of playoffs missed it.

[jwr38]

I think good teams should have a chance not to make the playoff as well but it should be based of how a GM runs the team not how the sim runs as it does currently. If all the extra random things are removed such as coaches, injuries, Morale, lines changing, suspensions etc are removed then it should be pretty consistant. Shouldn't it be how the GM sets the lines, ice time, deals with injuries/suspensions, if the coach suits the team, how the GM deals with low morale?

I also believe the engine will do the above, only reason I am here. If I didn't believe that I wouldn't have made a player creator, I wouldn't be testing. Rome wasn't built in a day and I don't expect Simon to fix all issues in a day. The only thing we can do is offer our support and ideas.

[LUF]

Add a "randomization" meter in the league settings. That way, the leagues can determine themselves if they want a realist simulation or a randomized simulation.

For example, let's say it'd be a slider with a value of 1 to 100. The lowest the value, the lowest luck would be factored in.
1 = No random at all. Favorite teams always win.
25 = Easily predictable, but once every couple of years, a good team could not make the playoffs.
50 = Average result. Good teams arent that sure of making the playoff. Big teams could could miss them every once in a while.
75 = Sure, some of the skills are considered, but there's at least one big surprise every year.
100 = Totally random, no skill required!

-----

About the 99PA/99SC thing. How do you recreate "Ovechkin" then? He always have a lot of assists and a lot of goals every years in the NHL. Set the PA too low, and he'll only shoot, set the PA too high and he'll only pass. I was under the understanding that a 99/99 player would do BOTH a lot.

[Foo]

I think that's the main thing here, the scoring of big offensive players. It is not really the random thing... cause if star players score goals like they should do, results in the standings will be automaticly better.