Sim Engine needs to be upgraded

[Steve 1,921]

It certainly has flaws. Unfortunately, they are bad flaws that leave a player wondering what the fuck to do with his build. The RNG is incredibly broken and I think if it could be corrected to have some form of accuracy, it would be fine. When things like line strats and line usage isn't even in this version is just an indicator that it is time to upgrade. At least to the newer decade old version. I see your point though and hope you understand what I am seeing in these sims are far from good. If that is the way the league wants it then it is what it is. I am more for a work/reward system with RNG thrown in as it should be. Not the other way around. 

[Corco 1,237]

34 minutes ago, Steve said:

It certainly has flaws. Unfortunately, they are bad flaws that leave a player wondering what the fuck to do with his build. The RNG is incredibly broken and I think if it could be corrected to have some form of accuracy, it would be fine. When things like line strats and line usage isn't even in this version is just an indicator that it is time to upgrade. At least to the newer decade old version. I see your point though and hope you understand what I am seeing in these sims are far from good. If that is the way the league wants it then it is what it is. I am more for a work/reward system with RNG thrown in as it should be. Not the other way around. 

I definitely see where you’re coming from, and I agree that we should be doing more testing with the newer STHS engines; I’ve asked tfong about their reported testing 1.5 awhile ago, but he never got back to me.

[Beketov 9,025]

6 minutes ago, Corco said:

I definitely see where you’re coming from, and I agree that we should be doing more testing with the newer STHS engines; I’ve asked tfong about their reported testing 1.5 awhile ago, but he never got back to me.

FWIW we are constantly testing but a few things with the newer engines are apparent:

 

1. Nothing is less random, even people on the STHS forums have said the results in 2+ are less realistic.

 

2. Going to 1.5 may be possible but 2+ would need huge restructuring of either our starting TPE or update scale or both.

[Corco 1,237]

14 minutes ago, Beketov said:

FWIW we are constantly testing but a few things with the newer engines are apparent:

 

1. Nothing is less random, even people on the STHS forums have said the results in 2+ are less realistic.

 

2. Going to 1.5 may be possible but 2+ would need huge restructuring of either our starting TPE or update scale or both.

Even going to 1.5 would be beautiful.

[Beketov 9,025]

12 minutes ago, Corco said:

Even going to 1.5 would be beautiful.

@Josh can confirm because he’s running the shadow sims but I believe 1.5 is one of the tests.

[Steve 1,921]

40 minutes ago, Beketov said:

FWIW we are constantly testing but a few things with the newer engines are apparent:

 

1. Nothing is less random, even people on the STHS forums have said the results in 2+ are less realistic.

 

2. Going to 1.5 may be possible but 2+ would need huge restructuring of either our starting TPE or update scale or both.

What does less realistic mean. From the STHS forum, it referred to it being more realistic when it came to "high end" players actually playing like good players. I think that part of it would be rewarding if you work hard towards a build and have said player actually play somewhat to their potential. I can see that this isn't an easy thing to do or an upgrade would have been completed long ago.  

[Josh 1,671]

1 hour ago, Beketov said:

@Josh can confirm because he’s running the shadow sims but I believe 1.5 is one of the tests.

Yup

[Beketov 9,025]

1 hour ago, Steve said:

What does less realistic mean. From the STHS forum, it referred to it being more realistic when it came to "high end" players actually playing like good players. I think that part of it would be rewarding if you work hard towards a build and have said player actually play somewhat to their potential. I can see that this isn't an easy thing to do or an upgrade would have been completed long ago.  

It’s more so that it’s too good of a thing. With our current meta of 99’s (aka without needing huge update scale changes) we see players getting 400+ points per season.

[Rayzor_7 741]

5 hours ago, Beketov said:

I’m still okay doing that, just need to get around to make them. There’s nothing really special about our sliders; they are just what works for us.

I'm fascinated by the STHS engine and testing. I would love to get my hands on the sliders and maybe a couple seasons old VHL league file if this is possible. 

[Corco 1,237]

31 minutes ago, Beketov said:

It’s more so that it’s too good of a thing. With our current meta of 99’s (aka without needing huge update scale changes) we see players getting 400+ points per season.

move over Scotty Campbell

 

 but actually, I would be all for a harsher update scale if it meant that we were moving to 1.5

Edited February 2, 2020 by Corco

[Beketov 9,025]

2 minutes ago, Corco said:

move over Scotty Campbell

 

 but actually, I would be all for a harsher update scale if it meant that we were moving to 1.5

1.5 theoretically shouldn’t have the same issues as 2.X in regard to scoring but we’re testing it.

[Esso2264 774]

oh i want league file pls

[Beketov 9,025]

1 hour ago, Esso2264 said:

oh i want league file pls

We’d have to discuss it but I’m a little less open to straight up league files.

[Devise 4,475]

On 2/1/2020 at 8:26 AM, Tagger said:

I agree with the point you make about too many stuff like settings and sliders not being made readily apparent as to what is being used and to what level. I made a similar point in my media article here and Devise mentioned the idea of making a video showing what settings and sliders are currently set, but since then I haven't heard anything more on that.  

 

Still meaning to do that. I meant to do it last year when I was already recording regularly during playoffs, but I'll get to it this playoffs for sure. 

 

There is even a bunch of misinformation in this thread. Simon himself is responsible for all sorts of misinformation mostly because on the whole a lot of his sliders and settings aren't fully tested to work with the older versions/our roster sizes. The biggest one I see is the "sliders mean nothing." Which doesn't explain how you can manually sim 10 exhibition games against the same two teams and have one team have 5 physical and watch them hit more and then go 5 offense for another game and watch them shoot more. Same with minutes. Pushed to the extremes, the sim shows these settings do something on a semi consistent basis. 

 

The issue is we aren't pushed to the extremes and we have so many other variables it comes down to, "how much do they actually do in the regular landscape of the VHL?" That is a question that is sadly much harder to answer. 

 

As for not being allowed to show sliders, no idea where that info came. I've shown them in playoff sims before. VHL and VHLM (to my knowledge unless it got changed or Bek/I accidentally picked up the wrong index to build something off of) should have had the same sliders since the old Jardy days. Even then we were showing the sliders. We have in the past expiremented with slider changes when trying to test newer versions (V2 for example) mostly to try and get better results.

 

That said as @Josh can confirm he has been doing two test shadow sims this season with some update scale changes and one of them has been on V 1.5. I've only seen a snippet of those results, but thus far stats and things in general seem pretty similar to this season. So that is probably a plus that we have some semi consistent results across 3 different test sims. 

 

Anyway @Steve thanks for the thoughts and criticisms, as well to everyone else. While we aren't quite ready to "upgrade sims" we absolutely still are trying to find the right answer as the league continues to grow and evolve. One of the hardest things for us in the old days was trying to use newer versions of STHS that have more things added. But as update scales lower and harshen, player sizes grow and overall league parity increases (I know @Victor I shouldn't hold my breath just because it's here now) we may be able to even use V2 of STHS at some point. Which is probably my biggest counter argument to switching sim engines. Our biggest issue with newer STHS was how "NHL like" it was. However the league itself is becoming "NHL like" itself, so there is a good chance as things continue on this path we'll meet the standards required to make that version of STHS work. 

 

 

Edited February 2, 2020 by Devise