Element's True Ratings
-
Some of you might remember my True Rating calculations from last year. Some of you might have zero idea what it is. What ever happened to that?
In case you didn't know, they never left! True Ratings are calculated on 2KDB and possibly somewhere else... (stay tuned). True Ratings are a calculation based on the attributes of a card, weighted on importance per position (passing accuracy matters a lot for a PG, but not as much for a center, rebounding vice versa).
For example, here is DM Jayson Tatum's:
@jdealla and I really want to push this even further to get other factors involved, like badging, animations, height, etc. Those are a little tricky because some of them are subjective (e.g. 6'3" is too short for a PG? 6'10" too short for a center?) but I recognize that purely attribute-based calculation leaves out some important details. However, it is a great baseline to see past the overall rating / card tier that 2K assigns as they can be incredible misleading (in both ways -- cards are underpowered for their tier or are budget monsters who play like opals/DM's).
I just wanted to get the conversation going again and give you a hint of things to come. For now, based on the traditional True Rating calculation, here are the top 5 cards per position as of today:
Point Guards
Russell Westbrook - 99 - IDOLS Series II - 98.07
Stephen Curry - 99 - Warped Reality - 97.29
Derrick Rose - 99 - Retro - 96.42
Grant Hill - 99 - Out of Position - 96.39
John Wall - 98 - All Star Spotlight Sim Completion Rewards - 96.18Shooting Guards
Michael Jordan - 99 - Pantheon All Star - 97.39
Tracy McGrady - 99 - Radioactive - 96.5
Carmelo Anthony - 99 - IDOLS Series II - 96.03
J.R. Smith - 99 - Season Ultimate Rewards - 95.91
DeMar DeRozan - 99 - Glitched - 94.45Small Forwards
Larry Bird - 98 - IDOLS Series II VIP - 93.95
Jayson Tatum - 99 - Boston Celtics - 93.93
Billy Owens - 98 - IDOLS Series II VIP - 93.91
Rudy Gay - 98 - Warped Reality - 93.87
Chris Webber - 98 - Out of Position - 93.16Power Forwards
Zion Williamson - 99 - Pantheon All Star - 96.54
Giannis Antetokounmpo - 98 - Radioactive - 94.26
Paul Millsap - 98 - Glitched Market Reward - 91.45
Anthony Davis - 97 - IDOLS Series II - 91.35
Rui Hachimura - 98 - IDOLS Series II VIP - 91.33Centers
Kareem Abdul-Jabbar - 99 - Radioactive - 97.02
Blake Griffin - 99 - IDOLS Series II VIP - 96.88
Shaquille O'Neal - 99 - Warped Reality - 96.83
Wilt Chamberlain - 99 - Glitched - 96.66
Kevin Garnett - 99 - IDOLS Series II - 96.13In this quick and dirty look, players are only categorized in their higher True Rating position, so someone like Zion who is the best SF, is only listed in PF because his rating is higher there. Which is why Larry Bird, generally not considered an elite card, is the #1 rated SF.
Let me know your thoughts, if you have any questions or comments about how to think about the best ways to calculate the True Rating, we'll be building this into something even better in the near future
-
Pinned!
-
could you also possibly factor in dribble sigs for PG eligible cards?
-
Just a couple that jumped out to me --
98 Danny Ferry
PF: 91.85% SF: 92.56%97 Tony Kukoc
PF: 89.14% SF: 90.47%97 Don Ohl
PG: 93.94% SG: 92.99%96 Alex Caruso
PG: 93.82% SG: 93.37%97 Moses Malone
C : 93.72% PF: 93.08%95 Bol Bol
C : 90.11% PF: 89.78%You want to be aiming for 93+ at this point. As you can see model size, animations, badges can play a huge role in the card's actual in game ability (Ferry, Bol, Kukoc, etc) which is what we're looking to incorporate.
-
@11wakawaka We're going to try something like that, yes. The more subjective something is, the harder it is to quantify, but some of the dribble sigs are obviously 1000 x better than others so we don't want to exclude them. Same with shot release speed.
-
-
@element yeah I look at Ferry and Kukoc and Bol and I'm like... this are quite literally the Top 3 cards in the game.
-
Any system that puts Owens above Rudy is SUS until further notice.
-
@jase agreed
Owens is great but Rudy is better
-
@element said in Element's True Ratings:
Billy Owens - 98 - IDOLS Series II VIP - 93.91
Rudy Gay - 98 - Warped Reality - 93.87@Sssnaps
alert -
!2ku c billy owens
-
98 Billy Owens
Golden State Warriors - Premium IDOLS Series II VIP
HT: 6'8, WING: 7'2"
Key Stats89Shot 3pt97Driving Dunk96Speed96Acceleration93Speed With Ball86Ball Handle96Lateral Quickness94Perimeter Defense96Strength90Interior Defense94Block92Offensive Rebound90Defensive Rebound94Pass PerceptionKey BadgesGoldDeep ThreesGoldHot Zone HunterNoneBlindersHOFPosterizerHOFFearless FinisherHOFSlithery FinisherHOFQuick First StepNoneDimerHOFClampsHOFIntimidatorGoldRim ProtectorKey AnimationsDamian LillardLower BaseMichael JordanUpper ReleaseVery QuickRelease TimingQuickDribble StylePro 2Size Up Escape PackagesPro 3Moving Behind The Back -
@element I personally have never even used the card, I just know Sssnaps hates it
-
!2ku c rudy gay 98
-
Key Stats93Shot 3pt98Driving Dunk95Speed95Acceleration93Speed With Ball90Ball Handle95Lateral Quickness94Perimeter Defense86Strength90Interior Defense90Block87Offensive Rebound89Defensive Rebound92Pass PerceptionKey BadgesHOFDeep ThreesGoldHot Zone HunterNoneBlindersHOFPosterizerHOFFearless FinisherHOFSlithery FinisherHOFQuick First StepGoldDimerHOFClampsHOFIntimidatorGoldRim ProtectorKey AnimationsRudy GayLower BaseRudy GayUpper ReleaseVery QuickRelease TimingQuickDribble StylePro 2Size Up Escape PackagesPro 4Moving Behind The Back -
Oladipo is probably 00.01 @Sssnaps
-
HOF range and gold dimer on Rudy is the key info for me. Plus Rudy Gay shot is superior to Jordan in every way. Is Pro 3 BTB better than Pro 4? I can't recall. That might be a tick in Billy's column if so.
-
@jase said in Element's True Ratings:
Any system that puts Owens above Rudy is SUS until further notice.
You spelled OBJECTIVE wrong.
-
J.R. Smith - 95.91 (7th)
Trevor Ariza - 90.07 (equal 125th)I appreciate the effort so much @element, but would love to know how much of the ratings is based behind pure stats, and how much is based behind other factors? Because I can tell you right now Ariza > JR
And this is coming from a standpoint where I though most of your ratings were relatively accurate last year
-
@sp1ash said in Element's True Ratings:
I appreciate the effort so much @element, but would love to know how much of the ratings is based behind pure stats, and how much is based behind other factors? Because I can tell you right now Ariza > JR
It's all based on pure stats right now as @element said:
True Ratings are a calculation based on the attributes of a card, weighted on importance per position (passing accuracy matters a lot for a PG, but not as much for a center, rebounding vice versa).
I think a better question to help @element out in the future is what makes you say that Ariza is better than J.R.? that would be helpful in order to start thinking about what to focus on for quantifying factors that aren't immediately quantifiable.
-
@jdealla Very fair point. Some aspects of what makes (in a vacuum) Ariza > JR:
- Quicker release timing, and possibly a better base full stop
- Slightly better dribble sigs
- Height
- Wingspan
- Very, VERY slightly better defensive tendencies
Not sure if you take into account importance of less relevant stats (such as post game). Obviously I don't know exactly how your ratings work, so I could be barking up the wrong tree, but those are just some aspects that I think could be helpful
apologies if I came off accusatory
-
@sp1ash said in Element's True Ratings:
apologies if I came off accusatory
nope not at all! feedback like this is exactly what will help @element make the ratings more robust
-
Maybe this helps. These are the attribute weights per position. Sorry the format is janky but you can see what's happening here
PG SG SF PF Shooting C Non-Shooting C Shot_close 2 2 3 4 5 5 Shot_mid 4 4 4 3 3 1 Shot_3pt 4 5 4 3 3 0 Free_throw 4 4 4 3 3 2 Offensive_consistency 1 1 1 0.75 0.25 0.5 Driving_layup 4 3 3 2 2 2 Standing_dunk 0 0 0 3 5 5 Driving_dunk 2 3 5 3 3 3 Draw_foul 3 3 4 4 3 4 Post_moves 0 0 3 4 5 5 Post_hook 0 0 1 2 5 5 Post_fade 1 1 3 4 5 5 Hands 0 0 0 0 0 0 Speed 5 4 4 3 3 3 Acceleration 4 3 3 2 3 3 Vertical 2 2 3 4 4 4 Strength 1 1 3 4 5 5 Stamina 0 0 0 0 2 1 Hustle 0 0 0 0 0 0 Speed_with_ball 5 3 4 2 2 1 Ball_handle 5 2 2 1 1 1 Passing_accuracy 5 2 2 1 0 0 Passing_vision 0 0 0 0 0 0 Interior_defense 1 1 2 4 5 5 Perimeter_defense 5 5 4 3 2 2 Help_defense_IQ 0 0 1 3 1 4 Lateral_quickness 5 5 4 3 3 3 Pass_perception 0 0 0 0 0 0 Steal 3 3 3 1 2 1 Block 0 0 1 3 5 5 Defensive_consistency 3 4 3 4 4 4 Offensive_rebound 0.5 0.75 1 4 5 5 Defensive_rebound 0.5 0.75 1 4 5 5
-
@element listen idk what it says for bol bol his true rating is a 99 lol
-
I think jumpers should factor into this. Like for example, a really really good jumper like 98 or ray allen could be 5 points, and a really bad jumper could be 0, or even negative points. It could also depend on position. You don’t need the absolute best jumper on your center
-
@nd13 malik sealy’s rating would skyrocket with this formula
-
@element said in Element's True Ratings:
John Wall - 98 - All Star Spotlight Sim Completion Rewards - 96.18
The OG under 18's know.
-
@thekingcartii yes we do
-
@thekingcartii really a 99
-
@thekingcartii should be an 9999 overall
-
Top ten True Ratings after today's cards (all DM:)
10 Shaquille O'Neal 97.29%
9 Derrick Rose 97.33%
8 Wilt Chamberlain 97.35%
7 Kareem Abdul-Jabbar 97.49%
6 Blake Griffin 97.61%
5 Zion Williamson 97.67%
4 Vince Carter 97.71%
3 Stephen Curry 97.74%
2 Michael Jordan 97.85%
1 Russell Westbrook 98.69% -
@element Blake is great and is flying under the radar a bit bc of lack of curry and short arms, but he still manages to grab boards at an elite level and grab passing lane steals. I might keep him around
-
@jdealla yeah I'm not selling my Blake yet. Gotta see if Thon is actually better first. Might not be.
-
@quantumkat yea I'm coming to the same conclusion. nothing in this drop is making me want to splurge and get the card right away. I'm gonna wait and see.
-
@jdealla I don’t really like his jumper but he has hof range
-
@atgalex on next? I've ended up liking it a lot and I didn't expect to
-
@jdealla yeah next Im fully next now but I felt like he had a hitch
-
True Rating 2.0 test, PG, with height, wingspan, badges, and tendencies factored
10 Penny Hardaway 96.38%
9 J.R. Smith 96.56%
8 Kobe Bryant 96.67%
7 Rajon Rondo 96.90%
6 John Wall 96.96%
5 Derrick Rose 97.39%
4 Ben Simmons 97.43%
3 Stephen Curry 97.65%
2 Grant Hill 98.28%
1 Russell Westbrook 98.64% -
True Ratings 2.0 Beta
Top 20 at each position:
-
@element said in Element's True Ratings:
True Ratings 2.0 Beta
Top 20 at each position:
This is fantastic.
-
@shinostique There are some notable absences in this list (Danny Ferry, Bol Bol, etc) but this does take badging, tendencies, height and wingspan into consideration... I think most people even if you're NMS or budget should have a lineup consisting of most of these players
-
@element i'm interested in looking into it and building code to run it.
I think one thing that sticks out to me is that AI is above the rest, which is sort of a "code smell" so to speak.
Another thing is the non VQ releases seem to be higher than cards in the same tier with those that are on quick.
I guess it all depends on intent of what the ratings are measuring, but I think a good way to look at it is "meta consensus", ie who people with all the options are running at certain positions, and I don't think AI, Dame, Russ, or Rose are filling in that slot for PG
-
@jdealla I agree, there's two challenges to this
1 - Determining whether something like a dribble package, shot release, "ideal" position height is objectively better or not.
2 - Once identifying something that is objectively better, what's the appropriate way to scale it.To address the Iverson thing, it could be something as simple as scaling the height penalty up way more. Right now it's simply that you get a 1% boost if you're a PG over 6'6". Very basic. But of course, height matters - every facet of the game is contingent upon height, so it's far from a subjective difference. So Iverson vs Wade at PG, height matters an equal amount, but obviously A.I. being 6'0 is objectively worse than Wade at 6'4". I'll find a better way to address it.
I've had much more trouble thinking of ways to deal with the animations. There's just inevitably going to be someone who thinks Pro 1 is the best and Pro 2 is overrated, etc. And I'm not a dribble god so I don't know the intricacies of which packages are better than others. But as you said, we should try to attack the notion of the 'meta consensus', i.e. in game mechanics, unique to each installment of NBA 2K, that are harder to counter, and if you don't do these things in competitive online play, you are objectively at a disadvantage. So if the curry slide is considered one of these things, we have to calculate it's importance and factor it in.
Same thing with shot releases. Base 11 from 2K19 was such a major advantage for cards - if you mastered the shot timing, you could go 24/25 from 3PT unless the opponent knew exactly how to stop it. But on the other hand, some people prefer a smooth, easy to green shot release vs the fastest release possible. How do we account for that? I'm not totally convinced that the fastest shot release is critical to objectively necessary in-game mechanics, even though they certainly are for me personally.
The purpose is to give everyone, regardless of play style, preferences, skill level, etc a way to evaluate how good a card really is in-game. Anyone can look at a card and say 'this card is ass' and they might be right, but when you're looking at 2 elite cards it's kind of tricky to say which one is truly better. The card overall rating provided by 2K is absolutely meaningless. There are really people out there that think that this year, a Dark Matter will ALWAYS be better than a Galaxy Opal card no questions asked, which is just wrong.
-
@element said in Element's True Ratings:
I'll find a better way to address it.
maybe you could do it like this
The average height for a PG is 6’2
So the lower they go under that height
The lower the true rating goes
Do that for every position
Here are the average heights for each position in 2k
Average pg height: 6’3
Average Sg height: 6’5
Average sf height: 6’7
Average pf height: 6’9 1/2
Average c height: 6’11 1/2So for example, let’s take Paul millsap’s opal
He’s 6’8
At PF, hes 1 1/2 inches below the average height
So his rating would go from 91.5 to 90 at pf
But at SF he’s an inch taller than average
So his rating at SF would go from 91.2 to 92.2just my suggestion
-
@element I think tiers are probably the only way to handle animations- the base 11 example you bring up is a good one thst makes the case for tiers since at some point it really is subjective.
I think for the most part we could bucket animations for all 2Ks - there are always elite animations (you get points), neutral animations (no change), and busted animations (negative points).
This would likely cover all cases. I think the weighing of each of these is a little more subjective about play style, but at the end of the day weighing anything is somewhat subjective. Although we can mitigate that with polling by getting data on what users think should be weighted more heavily
-
@jdealla Made some tweaks to penalize lower height and add points for taller players. I also added a very preliminary calculation to favor VQ & certain shot base/uppers.
PG
C
-
True Rating 2.0 mini update incoming
-
I was like something is wrong with the data, why is Invincible Magic having C as secondary position.
I was today years old when I realized he actually is PG/C
-
Top 25 POINT GUARDS (new rating system, factors height, wingspan, badges, tendencies, animations):
-
Top 25 SHOOTING GUARDS
You'll notice some cards go over 100%. WHAT IS THIS, THE SHOW? It's because I've boosted the impact of height quite a bit and it's not quite a round 100 at the moment.