The Celtics got Hosed
The Celtics are one of the least awarded teams of their quality ever. The question is, by how much, and why?
I’ve written three previous posts about awards. The first two outline what the path to an award could look like via historical analogues, and the third is my awards ballot where I make and defend the cases for each player. Now that the full slate of awards is published, it is clear the Celtics got hosed. This article will cover all historic analogs and make the case that this was one of the most brutal snubbings ever. A note on sampling; while much of the data in this piece is from 1977 - the first year after the merger - to the present day, some goes from 1989 - the year a third All-NBA team was added to the awards slate.
Let’s begin by situating the Celtics' regular season in history. They were the 5th post-merger team to win exactly 64 games and the 22rd to win 64 or more, tying the record for 18th most wins in a season. They had a net rating of 11.7, the third highest in NBA history behind the 96 and 97 Bulls, just ahead of the 17 Warriors and 16 Spurs. They set a new record for lead within their Conference - 14 games - and led the entire league by 7 games. They had more 20-point wins than they had losses and were also top-5 in both clutch win percentage and clutch net rating. Perhaps the most impressive stat is that after losing games 6 and 7 of the season, the Celtics did not lose two games in a row again until after they had mathematically clinched the top seed in the Eastern Conference. It is safe to say they were the best team in basketball this season and no one else came particularly close.
Their exact placement relative to other top-tier teams is a little complicated. Net rating is generally considered more indicative of a team’s “true” strength than record, but they are typically very aligned. The other top 5 net rating teams are all also top-5 in overall record. In fact, to find a team other than the Celtics with fewer than 66 wins, one would have to go all the way down to the 13 Thunder, who were +9.8 (10th) and won 60 games. So why the discrepancy? The 17 Warriors were +11.6 and won 67 games. Interestingly, they had roughly the same win percentage in the clutch as the Celtics (64% to the Celtics' 63.6%) and a very similar win percentage in non-clutch games as well (89.5 to the Celtics’ 87.8). The main difference in record is that the Celtics played 8 more clutch games than those Warriors did. The Celtics also had a better net rating in wins, losses, and clutch minutes than those Warriors. Perhaps this is explained as a product of the era. Even in 2017, teams were not shooting 3s to the same extent they are now. In 2017, the Rockets shot 40.3 3s per game, similar to the Celtics’ 42.5 this season, but the second-highest numbers in each season were 33.9 to 39.5, the lowest was 21 to 31.2, and the average jumped from 27 to 35.1. This variance could explain why greater dominance doesn’t translate to wins at the same rate. For some context, between 2015 and 2018, the lowest win total a top overall seed had was 65. Since then no team has won more than 64. Even writing this off as mostly the Warriors, from 03-09, another time of great parity, there were 13 60-win teams. Even adjusting win rates for the shortened seasons, there have been 6 since 2018, a drop off of almost 50%. This may be a lull in top-end talent, but conventional wisdom says this is the most talented era in NBA history. If this is the case, the difference in record is likely explained by general higher parity and greater variance allowing teams to take games off superior teams they wouldn’t have been able to even as recently as 5 years ago.
If this is the case and 64 is the new 67, these Celtics should absolutely be considered a top-5 regular season team in history, as high as third is possible. However, it would also be pertinent to identify a low-end in order to build a range. Rejecting the hypothesis that the 3 extra games can be explained by era, the low end is probably that this Celtics team is the 18th greatest ever. They are clearly superior to the other 64-win teams, but under this hypothesis, even one extra win should be considered a mark of superiority. This would place them squarely at 18.
What does this have to do with awards? Well, the Celtics received a rather small share of awards this season. What I am going to do is separate the all-time great teams into three categories, clearly better, arguably better, and clearly worse, and compare their returns to the Celtics. I’ll close by addressing why we might see this discrepancy and possible explanations.
Clearly Better:
This will be the shortest section. There are only two teams that had inarguably better regular seasons than the 2024 Celtics, the 96 and 97 Bulls. These teams both had 2 All-Stars, 2 All-NBA players, 2 All-Defensive players, and Toni Kukoč won 6MOY both years. In 96, Jordan was MVP, Pippen was top-5 in MVP voting, second in DPOY voting, and Phil Jackson was COTY. Of course, this was the greatest team of all time, they needed to be rewarded and they clearly were. I’ll note that while generally, we might see more awards going to a wider variety of players, an MVP or COTY trophy has generally been considered sufficient to explain extreme levels of team success without rewarding other players substantially. Getting both marks this team. Interestingly, the 97 Bulls saw far fewer awards. Jordan finished second in MVP voting, Pippen second in DPOY, and Phil Jackson was not a COTY finalist despite their stunning 69 wins and +12 net rating. The best explanation for this is probably just voter fatigue. Jordan had won 4 MVPs at this point and this was obviously a step down from the 96 season, leading voters to reluctance in giving the same level of awards. Compounding this, the Jazz had exploded onto the scene with 64 wins, with Karl Malone winning the MVP trophy. The combination of an exciting new team combined with a clear step back from their success the prior season led to one of the worst ratios of team success to awards in history. However, I’d like to note that even with this absurd snub, they still had finalists for two of the three biggest awards, the 6MOY winner, and 2 All-NBA players to go with their 2 All-Defensive players.
Arguably Better
There are 16 teams who had more wins but a lower net rating than the Celtics. They range from 65 to 67 wins and from +7.4 to +11.6. First, 10 of these teams had 2 All-NBA players. If you include teams who played before the third team was introduced but count the players who finished 11th to 15th in voting (an admittedly suspicious analytical move - Borda Count is a tricky voting system and adding more places could affect outcomes), this number becomes 12. However, even if we do not expand the number of teams, every team with a single All-NBA player also has the MVP. Only 4 of these teams lacked an MVP; the 17 Warriors, 16 Spurs, 08 Celtics, and 09 Lakers. In 16 and 09, the MVP was won by a team with even more wins. The Spurs and Lakers also had the MVP runner-up while the Celtics had the third-place finisher. The Celtics, Spurs, and Warriors also received the DPOY as a consolation prize. As previously noted, all four of these teams had at least 2 All-NBA players: the Spurs had 2, 2 All-Stars, and 1 All-Defensive player, the Celtics had 2, 3 All-Stars, and 1 All-Defensive player, the Lakers had 2, 2 All-Stars, and 1 All-Defensive player, and the Warriors, the only team without an MVP finalist, had 3, 4 All-Stars, and 1 All-Defensive player.
How does this relate to the Celtics? Well, this is the range of teams that they feel roughly comparable to in the scheme of history. If we treat awards as a way of explaining what causes teams to succeed, then we can understand this to be the range of awards the Celtics should have expected to receive. Within this scope, their awards were basically unprecedented. 75% of the time these teams had an MVP, 94% of the time they had an MVP finalist, and the one team that had neither had 4 All-Stars and 3 All-NBA players, as well as the DPOY. The Celtics’ return was a paltry 1 All-NBA player and an extra All-Defensive player over most of these teams. The conclusion is that among teams they were comparable to, they got by far the least awards recognition and it wasn’t even remotely close.
Clearly Worse:
For the remaining sample, I’ve capped our list at any team with at least 62 wins or a net rating higher than 9. Our cut-off for this section is all teams that satisfy one of those two conditions while having a lower net rating than the Celtics and the same or fewer wins. This sample includes 31 teams who were all dominant in their own right but clearly not to the extent the Celtics have been. These teams generally saw lower rates of awards. Only 38% had the MVP, 14% had the DPOY, and 16% had the COTY. They averaged 2.2 All-Stars, 1.4 All-NBA players, and 1.3 All-Defensive players. Immediately jumping out is that while the Celtics were within normal ranges on most of these, they are still distinctly below average. Compounding this, 8 of the 31 teams played when there were only two All-NBA teams, which could possibly lead this to being even more pronounced.
Immediately, we can say that the 14 teams with 2 or more All-NBA players had a better awards return than the Celtics but we’ll return to this later. In total, 15 of these teams had 1 All-NBA player and 2 had 0. Starting with the 0 All-NBA teams, we have the 12 Bulls and 89 Pistons. The Bulls were on a 62-win pace and were +9.1 and the Pistons were 63, +6. In both situations, these teams are not really in the same stratosphere of dominance as the Celtics were this year. Derrick Rose was the reigning MVP and it is likely that voters were a little put off by giving him the MVP for team success the prior season and then getting absolutely destroyed in the playoffs by the player who had the best individual stats. The Pistons I am not totally clear what happened and won’t be doing a deep dive into media coverage from that era, but it was likely a combination of people simply disliking them and the fact that their net rating was pretty unimpressive, closer to a 3 seed than an all-time great team.
For teams with exactly 1 All-NBA player, the Celtics still look relatively snubbed. Adjusting for shortened seasons, these teams ranged from 58 to 63 wins and +5.9 and +9.5. The closest any of these teams get to the Celtics is the 2020 Bucks, who won 56 of their 73 games (a 63-win pace) and were +9.5. This team is clearly still measurably worse than the Celtics but was better rewarded. Their 1 All-NBA player was also the MVP and DPOY, they had the COTY runner-up, and 3 All-Defensive players. Looking at broader trends, 47% had an MVP, 60% had an MVP finalist, and only the 09 Celtics had no finalist for any of MVP, DPOY, and COTY. In fact, across all 48 teams in all 3 categories we are addressing, the 09 Celtics are the only team to have no finalist for any of these three categories.
Perhaps the 09 Celtics are a good analog to the 24 Celtics, but even that looks funny in the light. They had a third All-Star and likely would have had 2 All-NBA players had Kevin Garnett not gotten hurt. While he may have still made it if his injury had taken place at a different time during the season, voters tend to penalize late-season injuries at a much higher rate. It is quite likely Garnett would have given those Celtics a 2nd All-NBA player, an MVP finalist, and a DPOY finalist. Despite this, those Celtics still arguably saw a better awards slate than the 24 Celtics due to the third All-Star.
Before moving on to other trends, I want to look at the high end of the “Clearly Worse” category. Starting with the 4 other 64-win teams, we have a range in net rating from +7.5 for the 22 Suns to +9.7 for the 97 Jazz. None of these teams got within 2 points of the Celtics this year. All 4 of these teams had 2 All-NBA players and each had one of the major 3 award winners. 2 DPOYs, 1 COTY, and 1 MVP among them. They had All-Defense numbers of 1, 1, 2, and 3, and three had 2 All-Stars while the 06 Pistons had 4.
On the high end of net rating, there were 5 teams with +9.5 or better. We’ve discussed two, the 20 Bucks and 97 Jazz. The remaining were the 13 Thunder, the 94 Supersonics, and the 91 Bulls with 60, 63, and 61 wins respectively. The Thunder and Sonics each had 2 All-NBA players; the Bulls had 1 but he was the MVP. The Thunder also had the MVP runner-up and a DPOY finalist while the Sonics had 2 top-10 MVP finishers and a COTY finalist. These awards still reflect a much higher degree of value being placed on the players of these teams.
Explaining this:
With the exception of maybe the 12 Bulls or the 89 Pistons, the 24 Celtics received the lowest share of awards of any team we’ve looked at - a sample that basically includes the entirety of the top 50 regular-season teams in post-merger history. I can come up with three possible explanations for this. The first is simply bias by voters. You’ll note that the only other team without a major awards finalist was also a Celtics team, perhaps the explanation is simply that voters do not like the Celtics and don’t like voting for them. This explanation would be sufficient but is unsatisfying and frankly, lazy. While I believe it is a piece of the puzzle, there is still data to be explored. The second explanation is that the Celtics are a totally unique team construction within NBA history. The argument essentially posits that the Celtics have collected the best sub-All-Star collection of talent in their 3-5 of any team in NBA history by a huge margin. I’ve noted that the era-shifting could explain differentials we haven’t seen before, but the absurd differential between the Celtics awards slate and comparable teams in history doesn’t feel satisfied simply by this. The third explanation, which I think is the biggest piece of the puzzle, is that the criteria for awards has shifted over time and we are moving into an era where winning more is not considered prima facie evidence of being better.
I won’t touch on option one beyond what I have already said; Celtics fans should bear in mind that the bias is real, but it does not make for an interesting explanation.
Unique Construction
The second argument is that the Celtics are a totally unique team construction in history - allowing them to have absurd dominance without the level of top-end talent that has been required in almost every single prior season. I want to quickly qualify this claim because generally, it is made in the context of this season. “The Celtics have the best top 6 so will always have a talent advantage somewhere” and while that may get some leverage in explaining why they are the best team this year, it runs into a major problem; it implies that role players might be more important than stars..
Conventional wisdom suggests that top-end talent shares a disproportionate role in determining the outcomes of games. This makes sense logically; the best players will play more minutes and generally have the ball or be central to a play at a much higher rate than their worse teammates. This means that if the Celtics have a weaker star and a weaker second option than another team, the gap between their 3-6 must be even larger than the gap between other team’s 3-6 to make up for the difference. Additionally, this gap must be astronomical. If we take, for example, the Lakers, a team with two All-NBA players, we end up with a funny little calculation. Tatum is better than Lebron and Anthony Davis by some amount based on awards, but both are substantially better than Jaylen Brown, receiving a little more than 3 times and more than 4.5 times the All-NBA votes he did respectively. Obviously, disaggregating the exact impact of each player borders on impossible, but it seems reasonable to suggest Davis plus James is better than Tatum plus Brown based on these results. You must now make up that gap and then exceed it by 17 wins and an 11.1 net rating (which would in itself be a top-5 net rating ever) with the gap between the less effective role players. The exact calculation for each team looks a little different, but my instinctive reaction is that creating a gap substantial will be quite difficult.
The first thing we can do is shorten to the Celtics 3-5 rather than their 3-6. 12 players on 11 teams received 6MOY votes and not one of them was Al Horford. If we take seriously that awards are disaggregating individual player impact and are doing it roughly accurately, then we need to accept Al Horford is not a top third sixth man at this point in his career. While he may still give advantages in some matchups, particularly against teams like Denver or OKC who also don’t have such a player, he is not going to skew positive in our rough value calculation against a lot of teams.
This leaves us with Derrick White, Kristaps Porzingis, and Jrue Holiday to explain the gap between the Celtics and other historically good teams. Let’s start with Porzingis. He played 57 games this year and did not qualify for awards but led the team in WS/48. It’s possible that some could argue his impact in those 57 games was All-NBA level and elevated the team, but he was not selected for the All-Star Game, which he did qualify for. Additionally, the Celtics were 21-4 in games he sat, a 69-win pace. While I do believe Porzingis makes them a generally better team with fewer weaknesses, it’s hard to argue it was his dominance that led to this season. Perhaps we can argue he was something like a top-40 player for those 57 games but that still feels tricky for explaining how successful they were without him, but for the sake of argument we will give the benefit of the doubt; he is a very good but sub-All-Star level player.
Derrick White is another tricky piece to fit. He was heavily debated in the media as a potential All-Star but was again, ultimately left off the team. One of my previous posts about awards was his case for this team and I believe history says he should have made it, identifying him as a top-20 or top-30 player in the league. A third All-Star also could have gone a long way to bridging the strange awards gap we are currently looking at. However, his case was ultimately rejected, largely based on the idea that his individual counting stats simply were not All-Star level and team success shouldn’t be used to determine this honor. This once again leaves us in a tricky position of if we are being incredibly generous to the idea that awards are accurately assessing talent, they would have Derrick White somewhere in the top-40 as well.
Jrue Holiday took a major step back this season in his role. He has previously been named to All-Star teams, but this season he took the fewest shots since his rookie year. While Jrue may be able to produce at a top-30 level if fully unleashed, this would probably also relegate him to top-40 production this season, once again, being generous.
Next, we must account for the fact that two Eastern All-Stars did not play and replacements were selected that did not include any of the above players. Derozan also received All-NBA votes over all of these players, completing 15 Eastern players considered better than any of the Celtics back end starters. Given that the West is generally considered more talented, it seems fair to say all three Celtics players would be in the 35-40 range at best.
So now let’s ask about some of our historic teams that received more high-end awards than these Celtics. Starting with the other 64-win teams, the 97 Jazz, the 06 Pistons, the 94 Supersonics, and the 22 Suns. We can immediately eliminate the Pistons who had four All-Stars and three All-Defensive players; they both had better talent at the 3-5 and saw more high-end awards.
Going back through from the earliest starts us with the Supersonics. One year later, they would have three All-NBA players. Their third, Detlef Shrempf, would receive votes in 94 but not a single vote for MIP in 95. While MIP is a confusing award overall, this could indicate he wasn’t actually significantly better the next year or at least was within a normal range. The Sonics were also worse in 95 than in 94, suggesting he didn’t make a crazy leap outside the bounds of normal. This would place Shrempf somewhere in the top-25ish players. The Celtics are now in a deep talent hole against this team and the argument that Jrue Holiday and 57 games of Porzingis can make up that gap is tenuous.
The 97 Jazz had the MVP and another All-NBA player. No one else on their roster received All-NBA votes or played in the All-Star game. Perhaps, this is a team the Celtics outperformed on depth. I’m not going to deep-dive into trying to reconcile points, pace, and style across eras, so I will instead give the benefit of the doubt here and say the Celtics could possible have made up this gap with their 3-5. Of course, this is still tricky; Stockton and Malone absorbed huge amounts of possessions on their own and are often considered the best Pick and Roll duo ever. Based on awards, we would say that Malone was much better relative to the league than Tatum and Stockton was much better relative to Brown. While the team performances are close enough that perhaps one could talk themselves into the gap between 3-5 being equally significant, that is a tricky argument to make.
Finally, the 22 Suns. This is probably the first analogy that comes to many people’s minds for this type of team construction. However, I’d argue this isn’t at all comparable. The Suns had finalists for DPOY and 6MOY, creating sizable gaps with the Celtics comparable options. Even if you give the Suns bench a slight disadvantage, that team played exceptionally well that season, certainly they had a lot of talent and the starters would all have to be top-50 or maybe top-60. Despite this, awards voters saw fit to give Monty Williams the COTY trophy, acknowledging that having a historic season without top end talent is a tremendous feat. Joe Mazzulla was not even a finalist. While vote totals haven’t been released yet, this would indicate voters generally believe Mazzulla’s team was dramatically more talented, something that isn’t quite borne out by the rest of awards voting.
I’ve already addressed a lot of the weirdness on the high and low ends of team performances, but I’ll note again, teams without top-end awards winners tend to see either lower team success or more individual awards. In our 48 team sample, MVP has a correlation of r=0.28, All-NBA has r=0.51, and COTY has r=0.24, all skewing positive with wins. These numbers are all still positive but less pronounced with net rating. To reiterate, the Celtics would have had to have a one of one weird team construction to line up all the talent levels right historically. Maybe this can be sustained by arguing the league is changing and the gap between league average and league-best is shrinking, but this would also be a tough argument to make. Essentially, the argument that the Celtics were able to sustain what is at worst the 18th-best regular season ever without these major awards hinges entirely on plausibility rather than hard data. A person has to begin by believing that the awards were distributed correctly and then fit the justification to support it; the major thrust is that this year’s award slate was simply weird and a major historical outlier.
Changing Criteria:
The final argument, and the one I am most partial to, is simply that the criteria for awards have changed over time. I’ve made the case before that MVP historically has been associated with team success. From the start of the three-pont era in 1980 to 2017, only 4 players won MVP without winning a title. Since 2017, there have been three. While there is a straightforward argument here that these players are still playing and could therefore close the gap, every single prior MVP who has won the title later would win their first title within 5 years of their award, likely suggesting that Westbrook and Harden no longer have a good chance. Embiid still has a few years, but I’m not sure anyone seriously believes he has a chance to make a deep run. Similarly, from 1980 to 2000, 5 players won their first title the same year they won their first MVP. No one since has done so except for Stephen Curry in 2015. Setting aside first MVP, from 1980 to 2000 10 title teams had that year’s MVP. Since, only 4 have and no one has done so since Steph Curry once again. Incidentally, the first “advanced” stat, an attempt to disaggregate a player’s impact into a single number, was released in 2014. Since then, only Joel Embiid has won MVP without leading the league in either BPM or WS/48, most MVPs have led in both. The bulk of evidence suggests the narrative theme; awards used to select for teams who saw great success as a way to determine who the best players were, now, we believe these advanced metrics can successfully isolate impact, leading to awards given to players on second round exits with a high degree of consistency.
A similar trend can be seen in All-NBA votes. Over the last 3 years, 7 players have made an All-NBA team while missing the playoffs. This is the most of any stretch since the third team was added in 89 except for 97-99, also seeing 7. Similarly, 7 players made it this year from a team on a 6-seed or lower; once again tying the highest mark with 1993. We also see a strong positive correlation between the year and the average seed of an All-NBA player and a slight negative correlation, suggesting that as years have gone on it has become easier for lower seeds to make it. In fact, prior to 2023, no one had made the first team while missing the playoffs ever; 23 saw two such players. I don’t want to go too in-depth on this in this piece but we are once again left with the question: is the league talent curve flattening or is the criteria shifting? Either way, it would seem that the Celtics did get the short end of the stick by having one of their greatest seasons ever right as this change takes place. Maybe the curve is flattening and that is what allowed them to have the season they did, or maybe it is all just raw hatred for Boston sports. What do you think? Let me know, I’d love more explanation.