Over the years I've been really interested in understanding the cost of winning in the NBA. That is to say, I'm only interested in the cost of the players' contracts because the data is publicly available and easy to scrutinize. Things like ticket sales, cost of running the team on a daily basis, coaches' salaries, front office salaries, etc. aren't readily available or are based on assumptions so I didn't bother with adding that into my findings. There are a few things that I did while my day at work was slow (NOTE: I'm not a statistician so I'm sure a lot of this is going to be rightfully scrutinized by some here): 1. Try to place a $ value on winning one NBA game 2. Try to place a value on a player relative to his performance 3. Draw some conclusions I used last season's statistics for this exercise. 1. Try to place a $ value on winning one NBA game Spoiler I took the total salaries paid out by each team in the NBA, bumped it up against the # of wins they had last season and get a general "cost" for each of those wins. NOTE: I did not include any luxury taxes that were paid out by teams like Golden State. I then took an average of the cost-per-win value and, league wide, a cost of winning one NBA game was a little bit less than $3 million at $2,903,554 This, more or less, gives us a baseline to work with. 2. Try to place a value on a player relative to his performance Spoiler One of the "fun" stats that I look at more than others is Win Shares which you can read about HERE. Of course, it's not perfect data but it's a quick and dirty way of seeing how much an individual player's performance contributes to his team's success. Using the average cost of 1 NBA win (~$3 million), I went to work evaluating our own team's data. James Harden led the league in total Win Shares last year at 15.4 WS. If we take this value and multiply it by the league average cost of 1 NBA win, Harden's value in 2017-2018 is $44.7 million, well above his contracted salary of $28.3 million. Below is a table of Houston's data: There aren't many things that jump out at me, however: 1. Some players were assigned a negative WS and those players actually were financial losses. Looking at those players I don't think anyone would argue that those players really didn't make any positive impact, overall, during the regular season 2. Ryan Anderson's Total Win Shares was a shocker. It almost made me throw this whole endeavor away but then I remember that were something like 44-13 last year when he was still a part of the rotation. At the end of the day, however, Anderson's production was no where near the $ value he was paid (e.g. Dude was overpaid...by a good amount). 3. Needless to say, Clint Capela's value last year was one of the best (not quite the best; see #3) in the league. Based on the WS he accumulated last year, his "value" was astronomically higher than what he was actually paid--$29.6 million vs. $2.3 million. We also saw this with players who passed the eye test last year and was deemed key contributors to the team's overall success. Luc ($8.1 million vs $1.47 million) and Gerald Green ($6.097 million vs. $872K) were monster values. 3. Draw some conclusions Spoiler So where does this leave us? I decided to pull the Top 20 Win Shares list and see if there's anything interesting on that end. Remember Clint's huge value? Jokic's performance last year gave him almost DOUBLE the value of what Clint provided when using their salaries. Additionally, any player that shows little to no variance between their salary and "value" pretty much means they played/lived up to their salary for that year (e.g. Derozan). The only player in the negative on this list is Steph Curry but there are a few reasons as to why this occurred but there's also one main reason why Golden State probably don't give 2 ***** and that's because they got to hoist the Larry O'Brien at the end of the year. I can only assume/conclude that there are NBA front offices that have more complex calculations like this that drives their decision-making when it comes to the players that they sign and to what offers they put out there. Maybe Ariza's agent used this crude model to justify that he's worth $15 million and Phoenix ran the same numbers and said "hell yea he is!"? Again, this isn't meant to convince anyone that Player A is overpaid or Player B is overpaid or Player C needs to be paid x amount. I thought it was a fun exercise to break away from the other constant chatters we see in the GARM.
Three things we already knew. 1. Superstars are usually underpaid because of the max contract restriction 2. Good players on rookie contracts are grossly underpaid because of the rookie pay scale. 3. Lulz, Grizzlies Spoiler
I really want to but 1. the data, so far, would have to include a few more assumptions/projections than what I would like to use (e.g. projected win totals) 2. we already know where some values are going to head for some of the players we're interested in (e.g. Chris Paul)
Of course if I were to try and apply what I did against our team then what Morey attempted to do with signing vet minimum contracts and/or buyouts "makes sense". Players on vet min contracts have more opportunity to outperform their contracts but that's always going to be a risky proposition.
First off, I love this kind of thing There's of course a fine line between cost-efficiency and winning titles. KAT, Westbrook etc. are a bit iffy considering the (true or not) lack of chemistry they consistently have in their teams. Perhaps it's the team, or it's the player. OTOH Curry is arguably(b**** though we might) the best player on his position and more importantly combines this with mixing easily with other players. Sure he took a step back when Durant came (also had some physical malheurs), but even with such a relatively ball-intensive partner he still lights up the boards at an enviable rate. So yeah, the warriors look at his value and probably correctly value the guy as pretty much worth every cent they can direct to him.
absolutely love this, thanks for posting. if you're taking request - for analysis #3, can you add a column to show the diff? i'd also be curious who the worst contract is in the context of win shares
I have the Rockets' data available. Column D uses the league average "cost per win" value of ~$2.9 million per win. If we're comparing our players' vlaue against the league, then our team performed considerably better (Captain Obvious right?) than the league average last year. The values in red under Column E is the variance/difference I think you're looking for. The biggest dud here is, of course, Ryan Anderson. Taking this exercise a step further, if you were to use the "cost per win" value based SOLELY on the Rockets total salary / by # wins (getting you a "cost" of $1,819,560.78), you get data for Columns G and H. As you can see, the variance between Value (column G) and actual salary (column B) tightens up where there are more salaries in the red than when compared to variance using the league average "cost per win" value. Here I would consider any player between +/- $500k to have "lived up to their contract". Essentially we got what we paid for. The true values from last year's team are very evident here: 1. Clint Capela 2. Luc Mbah Moute 3. Gerald Green 4. Trevor Ariza 5. Nene All were fantastic values. Another quirky piece of data can be found in the sum row. If you compare the actual total salaries for last year's team (under column B) against the "value" of the team (under column G), you get a positive difference (under column H). This is a quick and dirty way of saying, yea the team last year not only was worth every penny in terms of regular season performance but also had a bit of positive gain as well. Moving forward, the Rockets will have to heavily rely on Harden and Capela to combine for 24-25 WS. Harden by himself has been stellar in this regard and even in his crappiest of seasons (Dwight's last year here) he still accumulated 13.3 WS. Capela was on a trajectory of eclipsing 10 WS this year before he went down with injury. I think he can still do it but this is especially important over the next few years as Chris Paul's production will undoubtedly decline. The hope is that he can provide around 6-7 WS until the end of his contract. I don't think it's unreasonable since Steve Nash was able to produce 6 WS at age 37, his last season with Phoenix and Jason Kidd netted 6.4 at the same age with Dallas. The difficult part is the rest of the equation. If you can confidently pencil in 30-31 WS among Harden/Paul/Capela, you're still needing 26-27 WS to get to 56-58 wins to be a true contender in the league. How/where are you going to get that? Assume that PJ and Gordon are still on the roster next season and they contribue 4 WS apiece, that's 38 wins from your 5 best players. Assuming MDA remains the head coach until Harden's contract ends (meaning his rotation is probably going to stay at 8-9 players deep), you're needing those 18-20 WS from 3-4 players, or roughly 5.5 WS per player. If we wanted to look at this season's crop of players, using WS/48 minutes would help forecast what kind of player could get you 5-6 WS a year. A WS/48 of about .11 will get you close to there. Another trend you want to see is how the player has performed throughout his career. If the guy is at around .11 WS/48 this year but was a total negative or low value in his other years in the league that, to me, is a red flag and should be considered an outlier. Below are a few names that have been at or above .7 WS/48 this year, who are free agents this summer and who, I think, could get you 5-6 WS within this team playing with Harden for the MLE (~$9 million) or less: Allen Crabbe Demarre Carroll Jeremy Lamb Marvin Williams Taj Gibson Bojan Bogdanovic Darren Collison Noah Vonleh Richaun Holmes Kostas Koufos Jeff Green