Monday Tip-Off: How To Rate a Fading Star

We’re at midcourt, and the ball is about to go up…it’s Monday Tip-Off! Join me as I begin the week here at the NLSC with my opinions and commentary on basketball gaming topics, as well as tales of the fun I’ve been having on the virtual hardwood. This week, I’m tipping things off with some thoughts on how challenging it can be to accurately rate a fading star in basketball video games.

Since they became a visible part of basketball video games, player ratings have been a contentious issue. We’ve rarely been completely happy with the ratings in the official rosters, sometimes for very good reasons; particularly when it comes to the historical players! Developers have also shared stories about players arguing about their ratings when they come in for motion capture or face scans. On a lighter note, when Hassan Whiteside returned to the NBA for the 2015 season and began turning in some good performances for the Heat, he joked about doing it to get his NBA 2K ratings up.

As someone who used to maintain current roster updates for NBA Live on PC, there were certainly ratings in the default rosters that I disagreed with. At the same time, I also sympathised with the producers who were responsible for those official rosters, because it’s impossible to please everyone. We all have different views on how certain players should be rated – with bias often being a factor – as well as how player ratings should be handled across the board. All roster makers end up settling on a system of some kind, me included. Even so, some players have proved particularly challenging to rate over the years, and a fading star will generally be among the most difficult.

I was reminded of this challenge while preparing a screenshot for my Friday Five suggesting five minimalist modding challenges. To provide an example of how we could have some fun challenging the community to spot the minimalist modding, I replaced one of the placeholders on the 1986 Boston Celtics in NBA 2K11 with Bill Walton. Although I assigned his leftover face and portrait from NBA 2K10, I didn’t actually need to do so, since the point was to make a subtle change for the eagle-eyed to spy. However, to ensure that he appeared alongside the other real players on the roster, I needed to edit his attributes so that he had a more appropriate Overall Rating.

Of course, by the time he joined the Celtics for the 1986 season, Bill Walton was no longer at the peak of his game. He was in his thirties, but more pertinently, multiple foot injuries had taken their toll. Indeed, the Celtics’ doctors originally weren’t eager to clear him to play, but Walton’s desire for one more chance was good enough for Red Auerbach. He went on to play a career high 80 games and was named Sixth Man of the Year. His 7.6 ppg, 6.8 rpg, 2.1 apg, and 1.3 bpg aren’t eye-popping numbers on the surface, but Walton only played 19.3 mpg in his sixth man role. Per 36, his 1986 season averages were a more impressive 14.1 ppg, 12.7 rpg, 3.8 apg, and 2.5 bpg.

In other words, while his minutes, role, and athleticism had diminished since his peak in Portland, his skills hadn’t. It’s why additional metrics such as Per 36 are at least worth considering when we’re rating players in rosters. A player averaging around seven rebounds in less than half a game isn’t the same as someone snaring the same number of boards in 38 or 40 minutes per night. Even going back to the early NBA Live games in the 90s, it’s clear that the developers took players’ roles and minutes into account. I can’t prove that they were using Per 36 specifically, but they definitely recognised production that was impressive for limited minutes, and rated accordingly.

It’s an approach that I and many other roster makers have adopted, and it’s served us well for the most part. It gets tricky when you have a fading star, though. After my first pass at replacing the placeholder player’s ratings with attributes more suitable for Bill Walton, his Overall Rating was too high. Now, I’ll be the first to admit that the importance of Overall Ratings can be overstated, but they’re not completely irrelevant, either. In this case, Walton’s Overall Rating made him the second-best player on the team behind Larry Bird, ahead of Kevin McHale and Robert Parish. I’d argue that Walton in his prime was superior to The Chief, but that doesn’t look right for 1986.

Once again though, the problem is that while Walton’s production had dipped, his skills and basketball IQ hadn’t evaporated. Per 36, he averaged more rebounds and blocks than Parish that year, and only four fewer points per game while shooting a slightly better percentage from the field. In fact, in around 16 fewer minutes per game, he was averaging more assists (2.1 to 1.8) and basically the same number of blocks (1.3 to 1.4). To that point, while rating Walton according to his Per 36 stats from 1986 feels as though it’s erroneously portraying him as still being at the top of his game, it’s arguably an accurate representation of his skill and relative production.

However, appearances do matter when it comes to ratings in basketball video games. Moreover, when he’s on the virtual hardwood, you want him to be appropriately good, but not feel like prime Trail Blazers Bill Walton when you play with or against him. Obviously, it didn’t really matter in this case, since I was just setting up a scenario for a screenshot. If I nerfed him too much or got a few ratings wrong to make his Overall Rating appear accurate, no harm done. In a proper roster mod though, it’d be vital to ensure that I was making the right changes. It’s a great example of how it’s not easy to accurately portray a decline in role and production without underrating skilfulness.

There are some ratings that can be lowered organically. As players get older, their athleticism and durability naturally declines, even as their skills remain sharp. Some players remain tremendous athletes into their thirties and even forties, but if you were to actually compare them to their younger selves, the decline will become obvious. As such, it’s usually easy to shave a little (or if necessary, a lot) off the athletic ratings of a fading star. If they’re even a step slower, speed and quickness may come down. If they don’t dunk as often or save the fancier stuff for special occasions, dunking and jumping should be reduced. If they’re a bit banged up, durability ratings can take a hit.

These reductions do usually make a difference on the court, and help whittle down their Overall Rating to represent how we perceive their current rank in the NBA. It’s not always enough though, especially when their skills and production – either their raw statistics or Per 36 numbers – haven’t similarly declined. This is where it becomes folly to focus too heavily on the Overall Rating of a fading star, because lowering it to where it “should” be usually requires them to take a massive hit to attributes such as offensive and defensive awareness, or ratings connected to stats that are only lower due to reduced minutes. This is where you can begin to underrate their skills and IQ.

That isn’t something you want to do, and yet, an Overall Rating that’s a bit too high for a fading star is impossible to ignore. From representing the players’ rank in the league at a glance, to the tangible effect on their value in franchise mode trades, you want to accurately portray that a fading star is no longer at their peak. Sometimes it’s a bit easier if a fading star has gone downhill rapidly and their production has fallen off a cliff – perhaps because they’ve suffered a few injury-plagued years – but even then, it’s tricky. Again, their skills and basketball IQ haven’t disappeared, but their bodies simply aren’t cooperating. Therefore, we must find ways to make them less effective.

Needless to say, it’s a challenge that producers have also had to tackle in the official rosters. The solution that they and most community roster makers have ended up adopting has been to fudge ratings wherever possible, much as I did to lower Bill Walton’s Overall Ratings for that screenshot. One of the advantages of the 50-99, 0-99, and 25-99 ratings scales in many sim titles throughout the years is that they do provide some wiggle room here. A rating of 54 may produce results that are very similar to a rating of 58, but those four points could be the difference in lowering or raising a player’s Overall Rating another notch. Thus begins a delicate balancing act!

Adding Bill Walton to the 1986 Celtics in NBA 2K11 reminded me of this aspect of roster modding, but obviously, it’s a very old issue. Another prominent example that comes to mind is Michael Jordan in his comeback with the Washington Wizards. As his return was on the brink of becoming official in September 2001, a user created a thread in our old Forum titled “Prepping NBA Live for MJ’s comeback”, discussing what his ratings should be in roster mods for NBA Live 2001. The initial post pointed out that while MJ’s shooting and awareness ratings should remain the same, his speed, quickness, jumping, and dunking should all be lowered due to his age and ailing knee.

When NBA Live 2002 came out, some of the ratings were identical or similar to the ones proposed in that thread. The NBA Live developers did visit our Forum back then, so it’s possible they considered those suggestions! To that end, they dropped his dunking and jumping ratings to 82 as suggested, but weren’t as harsh with his speed and quickness, lowering them to 87 and 89 rather than 82 and 80. They made a bigger adjustment to his field goals rating though, dropping it from 99 to 90 rather than the suggested 94, and made minor reductions to his dribbling and stealing. His offensive and defensive awareness ratings were also lowered to 92 from 98 and 97 respectively.

Those additional changes left him with ratings that were still high, but lower than his prime attributes, resulting in an Overall Rating of 88. This actually tied MJ with six other players for second-best in the game; yes, it was a very different time when it came to Overall Ratings! In fact, despite the Legends version of MJ having the same ratings as in NBA Live 2001 – apart from a slightly higher steal rating- his Overall Rating was calculated as 93 rather than the 95 it had been the previous year. In any case, the point is that the developers lowered a few ratings that shouldn’t necessarily be impacted by MJ’s age and health, in order to depict him as being past his prime.

NBA Live 2003 actually pumped up a couple of Wizards MJ’s ratings – and calculated his Overall Rating at 89 – but lowered his three-point rating and shooting range, as he didn’t go to work beyond the arc very often in the 2002 season. His dunking rating was drastically reduced from 82 to 55 – on a 50-99 scale – and he was assigned the dunk package that was mostly comprised of layups along with a basic one-handed jam and a simple two-handed slam. It was understandable given his lower numbers, but it did underrate his ability to still throw it down during his Wizards years. It also felt wrong to see him rated lower than players like Shareef Abdur-Rahim, even at age 40!

It’s what was expected when it came to his ratings though, as well as those of other players from his generation whose star was fading. For example, it made sense that Scottie Pippen was rated lower than Rasheed Wallace owing to his age, production, and reduced role on the Blazers at that point. Nevertheless, rating Pip 83 Overall to Sheed’s 93 required the former to be underrated in terms of basketball IQ, key skills, and even a couple of his athletic attributes. Like MJ, for the most part Pip’s ratings aren’t so low that he’s completely useless or doesn’t play at all like himself, but they do slightly underrate his enduring skills and smarts, as well as his Per 36 production.

The comparisons to other players are what really make it challenging to rate a fading star, and accurately portray their decline. Sure, it made sense for Michael Jordan and Scottie Pippen to take a hit to their ratings in games set during their Wizards and Trail Blazers years, but how do you reconcile that with other players’ ratings? Should any version of MJ be rated lower than Elton Brand? Is it right that Pip is rated lower than Bonzi Wells in NBA Live 2003, just because he was a fading star while Wells was peaking? In terms of skill and basketball IQ – not to mention legacy – I’d have to say no. In terms of production and athleticism…well, it becomes far more complicated!

After all, if the intention is to accurately rate a fading star in their later years in the NBA, you can’t be too generous with legacy ratings. Wizards Jordan shouldn’t soar like he did with the Bulls. Shaquille O’Neal should no longer be incredibly agile for his size in his twilight years. John Stockton still schooled young guns such as Gilbert Arenas circa 2002 – whether Agent Zero wants to admit it or not – but he was past his prime. Kobe Bryant’s last season and Dwyane Wade in his second Heat run are great examples of the need to nerf a fading star. Still, while games should reflect a player’s decline, it feels wrong to rate them lower than anyone who was never on their level.

Furthermore, it never feels truly satisfying to achieve an accurate depiction of a player whose star is fading by lowering ratings that shouldn’t drastically change, if at all. Even with some wiggle room, fudging the ratings is a finicky exercise: a little off here, as low as possible there, and sacrificing this rating to keep that rating relatively high. When I was adding Bill Walton for the purpose of capturing that screenshot, I was reminded of that necessary evil. As I said, there have been players who were very tough to accurately rate in their prime because their athleticism or deficiencies heavily skewed their Overall Rating, but trying to rate a fading star is always an arduous task.

Perhaps the only comparable situation is when a star is coming off an injury-riddled season. How do you balance representing their most recent statistical performance and the potential lasting impact of their injuries with their established skill level and the chance that they’ll bounce back? Much like rating a fading star, there’ll be guesswork, fudged ratings, and nods to legacy, balanced by sensible nerfs. It can be frustrating, especially since you can’t please everyone. In an odd way though, that’s all part of the fun of roster modding. And so, I’d be lying if I said that briefly encountering the issue and talking about it hasn’t brought on an itch to pick up tools and start tinkering again.

The post Monday Tip-Off: How To Rate a Fading Star appeared first on NLSC.