League File | Standings | Schedule | League Leaders | Free Agents | Coaches | CSLO

New Ratings Explained

Moderators: Rizzo, KW

User avatar
Rizzo
Chairman of the Board
Posts: 6182
Joined: Fri Jul 11, 2014 2:01 am
Contact:

Re: New Ratings Explained

Post by Rizzo »

Sorry that we haven't had this information given to you guys yet. In full disclosure, Brett's responsibility when we did this testing was to test the different shooting ratings for each position. Kyrus and I did a ton of research using BasketballReference.com on historical players and their shooting percentages from different areas of the floor.

What we were seeing from a lot of guys - especially the fictional game created players - was that a lot of them were shooting 70+% from inside the restricted area. For example, Giannis in his MVP season in 2018-2019 shot 76.9% from 0-3 feet on the floor. That is one of the all-time highest numbers in NBA history, up there with some vintage Shaq seasons and of course Wilt. We shouldn't be seeing that all that often at all, let alone for a vast majority of players.

A huge reason why a lot of the interior scoring ratings FG_RA and FG_ITP are lower than you would expect to see is because we are taking that into account when creating these players. The truly elite should be scoring in the high 60% at the rim while others should be lower. I'm pretty sure two other players we used were Gilbert Arenas and Kemba Walker to compare their shooting percentages by distance from the hoop to what we were seeing in our created class.

On top of this, it was much easier for guards or perimeter players to shoot a higher percentage at the rim than power forwards and big men. To have these percentages fall in line, the guards had their FG_RA and FG_ITP ratings tuned down a bit to relfect that and keep them from all shooting 70% at the rim.

We put a TON of time into this testing and research and believe we came up with ratings that are:
- realistic to what we should see on an NBA court
- tuned to the DDS19 engine
- in line with prior draft classes

If Kyrus or Brett have anything else they can add to this, I would love for them to share their thoughts as well. Sorry it took so long to get some of this out to you.
CSL: Brooklyn Nets (2018-present)

User avatar
KW
CSL Champ 2017
Posts: 7696
Joined: Fri Jul 11, 2014 5:18 am
Location: The 704
Contact:

Re: New Ratings Explained

Post by KW »

It was actually crazier than 70%. In my tests I was regularly seeing 75-80% at the rim on 50+ FGA from pretty unassuming players. Adam and Brett did way more testing than I did, but we all saw the need to have lower than expected interior scoring numbers.

Now, I'm sure some of those exorbitant numbers will get mitigated by our GMs with strats to keep players away from the rim, as numbers have normalized in DDS3 since the early days of the CSL. But the decreases were necessary imo. We probably won't know whether we overcorrected or not, till they're dropped into the league file next season. I'm hoping that the conversion team saw what we did and are rating accordingly.
Image
Plus Joe Jackson / Jabari Brown / Oshae Brissett / Ed Davis / Kenneth Faried

User avatar
emplep7
Commissioner & CSL Champ 2018
Posts: 5468
Joined: Fri Jul 11, 2014 5:06 pm
Contact:

Re: New Ratings Explained

Post by emplep7 »

As always, thank you guys for all the work you put into these Draft classes. Its one of the reasons the league is so successful.

I do think that there's going to be an adjustment period for the new engine and that could come with some minor adjustments throughout the first season to make sure the long term success of the league is still in line.

Thanks again guys!
Image

User avatar
Rizzo
Chairman of the Board
Posts: 6182
Joined: Fri Jul 11, 2014 2:01 am
Contact:

Re: New Ratings Explained

Post by Rizzo »

emplep7 wrote: Fri Jan 24, 2020 6:05 pmI do think that there's going to be an adjustment period for the new engine and that could come with some minor adjustments throughout the first season to make sure the long term success of the league is still in line.
I think this is going to be very important for all GMs to realize. This is a HUGE transition and if we notice that a draft class or the transition is off in some way - be it low FG_RA/FG_ITP or something else - we should all be open to allowing the admins to make the necessary tweaks to get the league in line with what we need to see for the best of the league. It wouldn't benefit the league at all if the draft class comes in either super overpowered or super weak. We want everything aligned as planned.
CSL: Brooklyn Nets (2018-present)

User avatar
Marcos_Beck
CSL Champ 2019
Posts: 5998
Joined: Fri Jul 11, 2014 5:15 am
Location: Brazil
Contact:

Re: New Ratings Explained

Post by Marcos_Beck »

Rizzo wrote: Fri Jan 24, 2020 6:18 pm
emplep7 wrote: Fri Jan 24, 2020 6:05 pmI do think that there's going to be an adjustment period for the new engine and that could come with some minor adjustments throughout the first season to make sure the long term success of the league is still in line.
I think this is going to be very important for all GMs to realize. This is a HUGE transition and if we notice that a draft class or the transition is off in some way - be it low FG_RA/FG_ITP or something else - we should all be open to allowing the admins to make the necessary tweaks to get the league in line with what we need to see for the best of the league. It wouldn't benefit the league at all if the draft class comes in either super overpowered or super weak. We want everything aligned as planned.
I’m on board with that, as long as it’s a league-wide change and not a player related change I’m good with anything.
The #1 priority should be remembering that we all play by the exact same rules so if someone is nerfed everyone should be as well.
But it’s reallly important to keep the balance of the league for the transition to be good
#1 L.Ball #2 J.Murray #23 B.Bowen #35 M.Wagner #44 B.Adebayo
#16 J. Evans #38 M.Gbijine #52 Caleb Swanigan

18-19 Champions Image
#1 D.Rose #11 J.Holiday #21 J.Butler #42 A.Horford #13 J.Noah #0 L.Stephenson

User avatar
Rizzo
Chairman of the Board
Posts: 6182
Joined: Fri Jul 11, 2014 2:01 am
Contact:

Re: New Ratings Explained

Post by Rizzo »

Yes of course. I believe any mid-season edits should be global to either the entire league or to the rookie class.
CSL: Brooklyn Nets (2018-present)

User avatar
Flaming Homer
Commissioner
Posts: 2998
Joined: Tue Feb 20, 2018 6:31 pm
Contact:

Re: New Ratings Explained

Post by Flaming Homer »

emplep7 wrote: Fri Jan 24, 2020 6:05 pm As always, thank you guys for all the work you put into these Draft classes. Its one of the reasons the league is so successful.

I do think that there's going to be an adjustment period for the new engine and that could come with some minor adjustments throughout the first season to make sure the long term success of the league is still in line.

Thanks again guys!
This

Great work and thanks for all this time testing!!

mgtr81
CEO
Posts: 2370
Joined: Wed Jul 06, 2016 6:14 am
Contact:

Re: New Ratings Explained

Post by mgtr81 »

First of all, thanks for the explanation.

Second, of course, thanks for all the work put into this.

Thirdly, those of us like Ryan, myself or others who have run leagues for so many years and then are "normal" GM's in other leagues (like the CSL), may appear as having a bit of a "Commissioner syndrome" or something like that, and it may seem that we are telling others how to do their jobs. That's not at all my intention.

That being said, I would like to say that there is no need for mid-season adjustments if the transfer is done properly. And with that I'm not saying that the transfer is not done properly, it's a ton of work and I'm sure you guys are doing it the right way. But since you mentioned that possibility, if once you have the ratings transferred you take rosters from this season (with the new ratings in the new engine) and sim this season a few times, making the necessary tweaks until you get the same results (or close to the same results), then that's it, no need for mid-season adjustments. That's what Ryan did in the NLL and what I did in the GAH and I guess you already have done that here, too.

Again, I hope this is not taken the wrong way. I fully respect the work you do, because I know how many hours that implies. So thanks again for that.
ImageImageImageImageImage

User avatar
andrei
Chairman of the Board
Posts: 5173
Joined: Fri Jul 11, 2014 8:26 am
Location: Dublin, Ireland
Contact:

Re: New Ratings Explained

Post by andrei »

The way I understand it: dunk is a subset of rim which is a subset of paint. We have RIM rate and accompanying RIM shot ratings for efficiency. Same for paint. What determines DUNK efficiency?
So what do you expect when you draft Grabby McHolderson to play your right tackle?

User avatar
Rizzo
Chairman of the Board
Posts: 6182
Joined: Fri Jul 11, 2014 2:01 am
Contact:

Re: New Ratings Explained

Post by Rizzo »

andrei wrote: Sat Jan 25, 2020 4:22 pm The way I understand it: dunk is a subset of rim which is a subset of paint. We have RIM rate and accompanying RIM shot ratings for efficiency. Same for paint. What determines DUNK efficiency?
Two different sets: shot ratings and rates. Shot ratings consist of FG_RA, FG_ITP, FG_MID, FG_COR, FG_ATB. For this discussion, we are only talking about FG_RA and FG_ITP. There is one zone considered the rim area/restricted area. The percentage a player shoots in that area is determined by their FG_RA shot rating. There are about 7-9 other zones that are considered the paint, not including the the rim/restricted area. Those ratings are determined by the FG_ITP shot rating.

Now there are the two rates: dunk rate and rim rate. Rim rate dictates what percentage of a player's shots that come in the paint happen inside the rim/restricted area. So if a player attempts 100 shots in the paint and has a rim rate of 35, we would assume close to 35 of those 100 shots will come in the rim/restricted area.

The other rate is Dunk_Rate. This determines what percentage of the shots that come in the rim/restricted area will be dunks instead of layup attempts. So if this same player attempts 30 shots in the rim area and has a Dunk_Rate of 33, we would assume that of those 30 shots in the rim/restricted area, about 10 of them will be dunk attempts.

I hope all that made sense.
CSL: Brooklyn Nets (2018-present)

User avatar
andrei
Chairman of the Board
Posts: 5173
Joined: Fri Jul 11, 2014 8:26 am
Location: Dublin, Ireland
Contact:

Re: New Ratings Explained

Post by andrei »

Got it Adam - I understand the difference between rates and shooting ratings. But you mention Dunk_rate determines how many will be dunks vs layups in RA - this is clear. What is not clear is what is the shooting rating that determines the success of dunks? I.e I see two rates for restricted area: dunks (dunk rate) and layups (every shot in RA outside of Dunk Rate) but only one shooting rating FG_RA. Or am I missing something?
So what do you expect when you draft Grabby McHolderson to play your right tackle?

mgtr81
CEO
Posts: 2370
Joined: Wed Jul 06, 2016 6:14 am
Contact:

Re: New Ratings Explained

Post by mgtr81 »

andrei wrote: Sat Jan 25, 2020 7:27 pm Got it Adam - I understand the difference between rates and shooting ratings. But you mention Dunk_rate determines how many will be dunks vs layups in RA - this is clear. What is not clear is what is the shooting rating that determines the success of dunks? I.e I see two rates for restricted area: dunks (dunk rate) and layups (every shot in RA outside of Dunk Rate) but only one shooting rating FG_RA. Or am I missing something?
You're correct. That's the greatest mystery, as there is nothing like that in the DB.
ImageImageImageImageImage

User avatar
Rizzo
Chairman of the Board
Posts: 6182
Joined: Fri Jul 11, 2014 2:01 am
Contact:

Re: New Ratings Explained

Post by Rizzo »

mgtr81 wrote: Sun Jan 26, 2020 2:52 pm
andrei wrote: Sat Jan 25, 2020 7:27 pm Got it Adam - I understand the difference between rates and shooting ratings. But you mention Dunk_rate determines how many will be dunks vs layups in RA - this is clear. What is not clear is what is the shooting rating that determines the success of dunks? I.e I see two rates for restricted area: dunks (dunk rate) and layups (every shot in RA outside of Dunk Rate) but only one shooting rating FG_RA. Or am I missing something?
You're correct. That's the greatest mystery, as there is nothing like that in the DB.
What Manu said. There is no actual dunk rating. I guess there must be a formula that has to do with different ratings to calculate the rating for how often a player converts a dunk but I would imagine it's pretty high for most players.
CSL: Brooklyn Nets (2018-present)

User avatar
emplep7
Commissioner & CSL Champ 2018
Posts: 5468
Joined: Fri Jul 11, 2014 5:06 pm
Contact:

Re: New Ratings Explained

Post by emplep7 »

mgtr81 wrote: Sat Jan 25, 2020 2:42 pm First of all, thanks for the explanation.

Second, of course, thanks for all the work put into this.

Thirdly, those of us like Ryan, myself or others who have run leagues for so many years and then are "normal" GM's in other leagues (like the CSL), may appear as having a bit of a "Commissioner syndrome" or something like that, and it may seem that we are telling others how to do their jobs. That's not at all my intention.

That being said, I would like to say that there is no need for mid-season adjustments if the transfer is done properly. And with that I'm not saying that the transfer is not done properly, it's a ton of work and I'm sure you guys are doing it the right way. But since you mentioned that possibility, if once you have the ratings transferred you take rosters from this season (with the new ratings in the new engine) and sim this season a few times, making the necessary tweaks until you get the same results (or close to the same results), then that's it, no need for mid-season adjustments. That's what Ryan did in the NLL and what I did in the GAH and I guess you already have done that here, too.

Again, I hope this is not taken the wrong way. I fully respect the work you do, because I know how many hours that implies. So thanks again for that.
Haha believe me, we know we are not perfect and will take any and all advice for anything, especially something as major of an undertaking as this. One of the added benefits of using DDS3 for the off-season is to allow for ample testing once the final rosters are imported into the new game and that will allow us to sim through a season multiple times to ensure that the players transition as close to their current DDS3 counterparts as possible.

I do just want to make GMs aware that since I'm admitting just this once that we're not perfect we will be very open to feedback from the league throughout this first season about "potential" global tweaks to ensure continued success.
Image

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest